Mass Spectrometry of Nucleic Acid Noncovalent Complexes.

Nucleic acids have been among the first targets for antitumor drugs and antibiotics. With the unveiling of new biological roles in regulation of gene expression, specific DNA and RNA structures have become very attractive targets, especially when the corresponding proteins are undruggable. Biophysical assays to assess target structure as well as ligand binding stoichiometry, affinity, specificity, and binding modes are part of the drug development process. Mass spectrometry offers unique advantages as a biophysical method owing to its ability to distinguish each stoichiometry present in a mixture. In addition, advanced mass spectrometry approaches (reactive probing, fragmentation techniques, ion mobility spectrometry, ion spectroscopy) provide more detailed information on the complexes. Here, we review the fundamentals of mass spectrometry and all its particularities when studying noncovalent nucleic acid structures, and then review what has been learned thanks to mass spectrometry on nucleic acid structures, self-assemblies (e.g., duplexes or G-quadruplexes), and their complexes with ligands.

Native Electrospray Ionization of Multi-Domain Proteins via a Bead Ejection Mechanism.

Native ion mobility mass spectrometry is potentially useful for the biophysical characterization of proteins, as the electrospray charge state distribution and the collision cross section distribution depend on their solution conformation. We examine here the charging and gas-phase conformation of multi-domain therapeutic proteins comprising globular domains tethered by disordered linkers. The charge and collision cross section distributions are multimodal, suggesting several conformations in solution, as confirmed by solution hydrogen/deuterium exchange. The most intriguing question is the ionization mechanism of these structures: a fraction of the population does not follow the charged residue mechanism but cannot ionize by pure chain ejection either. We deduce that a hybrid mechanism is possible, wherein globular domains are ejected one at a time from a parent droplet. The charge vs solvent accessible surface area correlations of denatured and intrinsically disordered proteins are also compatible with this "bead ejection mechanism", which we propose as a general tenet of biomolecule electrospray.

High-Affinity Hybridization of Complementary Aromatic Oligoamide Strands in Water.

We prepared a series of water-soluble aromatic oligoamide sequences all composed of a segment prone to form a single helix and a segment prone to dimerize into a double helix. These sequences exclusively assemble as antiparallel duplexes. The modification of the duplex inner rim by varying the nature of the substituents borne by the aromatic monomers allowed us to identify sequences that can hybridize by combining two chemically different strands, with high affinity and complete selectivity in water. X-ray crystallography confirmed the expected antiparallel configuration of the duplexes whereas NMR spectroscopy and mass spectrometry allowed us to assess precisely the extent of the hybridization. The hybridization kinetics of the aromatic strands was shown to depend on both the nature of the substituents responsible for strand complementarity and the length of the aromatic strand. These results highlight the great potential of aromatic hetero-duplex as a tool to construct non-symmetrical dynamic supramolecular assemblies.

Adaptive Binding of Alkyl Glycosides by Nonpeptidic Helix Bundles in Water: Toward Artificial Glycolipid Binding Proteins.

Amphipathic water-soluble helices formed from synthetic peptides or foldamers are promising building blocks for the creation of self-assembled architectures with non-natural shapes and functions. While rationally designed artificial quaternary structures such as helix bundles have been shown to contain preformed cavities suitable for guest binding, there are no examples of adaptive binding of guest molecules by such assemblies in aqueous conditions. We have previously reported a foldamer 6-helix bundle that contains an internal nonpolar cavity able to bind primary alcohols as guest molecules. Here, we show that this 6-helix bundle can also interact with larger, more complex guests such as n-alkyl glycosides. X-ray diffraction analysis of co-crystals using a diverse set of guests together with solution and gas-phase studies reveals an adaptive binding mode whereby the apo form of the 6-helix bundle undergoes substantial conformational change to accommodate the hydrocarbon chain in a manner reminiscent of glycolipid transfer proteins in which the cavity forms upon lipid uptake. The dynamic nature of the self-assembling and molecular recognition processes reported here marks a step forward in the design of functional proteomimetic molecular assemblies.

Negative Electrospray Supercharging Mechanisms of Nucleic Acid Structures.

When sprayed from physiological ionic strength, nucleic acids typically end up with low levels of charging and in compact conformations. Increasing the electrospray negative charging of nucleic acids while preserving the native noncovalent interactions can help distinguish solution folds by ion mobility mass spectrometry. To get fundamental insight into the supercharging mechanisms of nucleic acids in the negative mode, we studied model G-quadruplex structures and single-strand controls in 100 mM ammonium acetate. We found that adding 0.4% of propylene carbonate, 0.4% of sulfolane, or 0.1% of m-NBA induces native supercharging. However, although 0.4% of m-NBA shows the highest supercharging ability, it induces unwanted unfolding of solution-folded G-quadruplexes. The supercharging effect resembles the effect of lowering the ionic strength, and this could be explained by partial neutralization of the ampholytes when droplets become more concentrated in their nonaqueous components. The supercharging ability ranks PC < sulfolane < m-NBA. m-NBA adducts to G-quadruplexes with high-charge states confirm that the supercharging agent interacts directly with DNA. Surprisingly, in the presence of supercharging agents, the most negatively-charged states also bear more alkali metal ion adducts. Larger droplets are known to result in more counterion adduction, so our results are consistent with native supercharging conditions producing larger droplets evaporating to a charged residue. However, when negative charge carriers from the electrolyte become too rare, chain ejection accompanied by denaturation, and hence non-native supercharging, can become predominant.

Lennard-Jones interaction parameters of Mo and W in He and N2 from collision cross-sections of Lindqvist and Keggin polyoxometalate anions.

Drift tube ion mobility spectrometry (DTIMS) coupled with mass spectrometry was used to determine the collision cross-sections (DTCCS) of polyoxometalate anions in helium and nitrogen. As the geometry of the ion, more than its mass, determines the collision cross-section with a given drift gas molecule, we found that both Lindqvist ions Mo6O192- and W6O192- had a DTCCSHe value of 103 ± 2 Å2, and both Keggin ions PMo12O403- and PW12O403- had a DTCCSHe value of 170 ± 2 Å2. Similarly, ion mobility experiments in N2 led to DTCCSN2 values of 223 ± 2 Å2 and 339 ± 4 Å2 for Lindqvist and Keggin anions, respectively. Using optimized structures and partial charges determined from density functional theory calculations, followed by CCS calculations via the trajectory method, we determined Lennard-Jones 6-12 potential parameters ε, σ of 5.60 meV, 3.50 Å and 3.75 meV, 4.40 Å for both Mo and W atoms interacting with He and N2, respectively. These parameters reproduced the CCS of polyoxometalates within 2% accuracy.

Design, synthesis, and antiproliferative effect of 2,9-bis[4-(pyridinylalkylaminomethyl)phenyl]-1,10-phenanthroline derivatives on human leukemic cells by targeting G-quadruplex.

Current multiagent chemotherapy regimens have improved the cure rate in acute leukemia patients, but they are highly toxic and poorly efficient in relapsed patients. To improve the treatment approaches, new specific molecules are needed. The G-quadruplexes (G4s), which are noncanonical nucleic acid structures found in specific guanine-rich DNA or RNA, are involved in many cellular events, including control of gene expression. G4s are considered as targets for the development of anticancer agents. Heterocyclic molecules are well known to target and stabilize G4 structures. Thus, a new series of 2,9-bis[(substituted-aminomethyl)phenyl]-1,10-phenanthroline derivatives (1a-i) was designed, synthesized, and evaluated against five human myeloid leukemia cell lines (K562, KU812, MV4-11, HL60, and U937). Their ability to stabilize various oncogene promoter G4 structures (c-MYC, BCL-2, and K-RAS) as well as the telomeric G4 was also determined through the fluorescence resonance energy transfer melting assay and native mass spectrometry. In addition, the more bioactive ligands 1g-i were tested for telomerase activity in HuT78 and MV4-11 protein extracts.

Large-Amplitude Conformational Changes in Self-Assembled Multi-Stranded Aromatic Sheets.

The orchestration of ever larger conformational changes is made possible by the development of increasingly complex foldamers. Aromatic sheets, a rare motif in synthetic foldamer structures, have been designed so as to form discrete stacks of intercalated aromatic strands through the self-assembly of two identical subunits. Ion-mobility ESI-MS confirms the formation of compact dimers. X-ray crystallography reveals the existence of two distinct conformational dimeric states that require large changes to interconvert. Molecular dynamics simulation validates the stability of the two conformations and the possibility of their interconversion.

Recommendations for reporting ion mobility Mass Spectrometry measurements.

Here we present a guide to ion mobility mass spectrometry experiments, which covers both linear and nonlinear methods: what is measured, how the measurements are done, and how to report the results, including the uncertainties of mobility and collision cross section values. The guide aims to clarify some possibly confusing concepts, and the reporting recommendations should help researchers, authors and reviewers to contribute comprehensive reports, so that the ion mobility data can be reused more confidently. Starting from the concept of the definition of the measurand, we emphasize that (i) mobility values (K0 ) depend intrinsically on ion structure, the nature of the bath gas, temperature, and E/N; (ii) ion mobility does not measure molecular surfaces directly, but collision cross section (CCS) values are derived from mobility values using a physical model; (iii) methods relying on calibration are empirical (and thus may provide method-dependent results) only if the gas nature, temperature or E/N cannot match those of the primary method. Our analysis highlights the urgency of a community effort toward establishing primary standards and reference materials for ion mobility, and provides recommendations to do so. © 2019 The Authors. Mass Spectrometry Reviews Published by Wiley Periodicals, Inc.

Symmetric and dissymmetric carbohydrate-phenyl ditriazole derivatives as DNA G-quadruplex ligands: Synthesis, biophysical studies and antiproliferative activity.

Here we present a novel G4-binding family of compounds based on a central core of phenyl ditriazole (PDTZ) modified with carbohydrates and phenyl pyrrolidinyl side-chains. Their synthesis was achieved using controlled click chemistry conditions to obtain both, symmetric and dissymmetric carb-PDTZ derivatives without any intermediate protecting steps through an optimized methodology. Binding of the new carb-PDTZ to a variety of G-quadruplex motifs was examined using different biophysical techniques. The symmetric carb-PDTZ derivatives were not able to stabilize G4, but the dissymmetric ones (containing one sugar and one phenyl pyrrolidinyl side-chain) did. Interestingly, the dissymmetric carb-PDTZ derivatives showed much higher G4 vs duplex DNA selectivity than the control compound PDTZ 1, which contains two phenyl pyrrodilinyl side-chains and no carbohydrates. Their potential antitumoral activity was also investigated by in vitro cytotoxicity measurements on different cancerous cell lines. All carb-PDTZ derivatives showed higher IC50 values than the control PDTZ 1, probably due to the lack of compound stability of some derivatives and to lower cellular uptake.

Design and Structure Determination of a Composite Zinc Finger Containing a Nonpeptide Foldamer Helical Domain.

A number of foldamer backbones have been described as useful mimics of protein secondary structure elements, enabling for example the design of synthetic oligomers with the ability to engage specific protein surfaces. Synthetic folded backbones can also be used to create artificial proteins in which a folded peptide segment (e.g., an α-helix, a loop) is replaced by its unnatural counterpart, with the expectation that the resulting molecule would maintain its ability to fold while manifesting new exploitable features. The similarities in screw sense, pitch, and polarity between peptide α-helices and oligourea 2.5-helices suggest that a tertiary structure could be retained when swapping the two backbones in a protein sequence. In the present work, we move a step toward the creation of such composite proteins by replacing the 10-residue long original α-helical segment in the Cys2His2 zinc finger 3 of transcription factor Egr1 (also known as Zif268) by an oligourea sequence bearing two appropriately spaced imidazole side chains for zinc coordination. We show by spectroscopic techniques and mass spectrometry analysis under native conditions that the ability of the peptide/oligourea hybrid to coordinate the zinc ion is not affected by the foldamer replacement. Moreover, detailed NMR analysis provides evidence that the engineered zinc finger motif adopts a folded structure in which the native ß-sheet arrangement of the peptide region and global arrangement of DNA-binding side chains are preserved. Titration in the presence of the Egr1 target DNA sequence supports binding to GC bases as reported for the wild-type zinc finger.

Electronic spectroscopy of isolated DNA polyanions.

In solution, UV-vis spectroscopy is often used to investigate structural changes in biomolecules (e.g., nucleic acids), owing to changes in the environment of their chromophores (e.g., the nucleobases). Here we address whether action spectroscopy could achieve the same for gas-phase ions, while taking advantage of the additional spectrometric separation of complex mixtures. We systematically studied the action spectroscopy of homo-base 6-mer DNA strands (dG6, dA6, dC6, dT6) and discuss the results in light of gas-phase structures validated by ion mobility spectrometry and infrared ion spectroscopy, of electron binding energies measured by photoelectron spectroscopy, and of calculated electronic photo-absorption spectra. When UV photons interact with oligonucleotide polyanions, two main actions can take place: (1) fragmentation and (2) electron detachment. The action spectra reconstructed from fragmentation follow the absorption spectra well, and result from multiple cycles of photon absorption and internal conversion. In contrast, the action spectra reconstructed from the electron photodetachment (ePD) efficiency reveal interesting phenomena. First, ePD depends on the charge state because it depends on electron binding energies. We illustrate with the G-quadruplex [dTG4T]4 that the ePD action spectrum shifts with the charge state, pointing to possible caveats when comparing the spectra of systems having different charge densities to deduce structural parameters. Second, ePD is particularly efficient for purines but not pyrimidines. ePD thus reflects not only absorption, but also particular relaxation pathways of the electronic excited states. As these pathways lead to photo-oxidation, their investigation in model gas-phase systems may prove useful to elucidating mechanisms of photo-oxidative damage, which are linked to mutations and cancers.

Probing ligand and cation binding sites in G-quadruplex nucleic acids by mass spectrometry and electron photodetachment dissociation sequencing.

Mass spectrometry provides exquisite details on ligand and cation binding stoichiometries with a DNA target. The next important step is to develop reliable methods to determine the cation and ligand binding sites in each complex separated by using a mass spectrometer. To circumvent the caveat of ligand derivatization for cross-linking, which may alter the ligand binding mode, we explored a tandem mass spectrometry (MS/MS) method that does not require ligand derivatization, and is therefore also applicable to localize metal cations. By putting more negative charge states on the complexes using supercharging agents, and by creating radical ions by electron photodetachment, oligonucleotide bonds become weaker than the DNA-cation or DNA-ligand noncovalent bonds upon collision-induced dissociation of the radicals. This electron photodetachment (EPD) method allows one to locate the binding regions of cations and ligands by top-down sequencing of the oligonucleotide target. The very potent G-quadruplex ligands 360A and PhenDC3 were found to replace a potassium cation and bind close to the central loop of 4-repeat human telomeric sequences.

Thermal Denaturation of DNA G-Quadruplexes and Their Complexes with Ligands: Thermodynamic Analysis of the Multiple States Revealed by Mass Spectrometry.

Designing ligands targeting G-quadruplex nucleic acid structures and affecting cellular processes is complicated because there are multiple target sequences and some are polymorphic. Further, structure alone does not reveal the driving forces for ligand binding. To know why a ligand binds, the thermodynamics of binding must be characterized. Electrospray mass spectrometry enables one to detect and quantify each specific stoichiometry (number of strands, cations, and ligands) and thus to simultaneously determine the equilibrium constants for each complex. Using a temperature-controlled nanoelectrospray source, we determined the temperature dependence of the equilibrium constants, and thus the enthalpic and entropic contributions to the formation of each stoichiometry. Enthalpy drives the formation of each quartet-K+-quartet unit, whereas entropy drives the formation of quartet-K+-triplet units. Consequently, slip-stranded structures can become more abundant as the temperature increases. In the presence of ligands (Phen-DC3, TrisQ, TMPyP4, Cu-ttpy), we observed that, even when only a 1:1 (ligand/quadruplex) complex is observed at room temperature, new states are populated at intermediate temperatures, including 2:1 complexes. In most cases, ligand-G4-quadruplex binding is entropically driven, and we discuss that this may have resulted from biases when ranking ligand potency using melting experiments. Other thermodynamic profiles could be linked to topology changes in terms of number of G-quartets (reflected in the number of specific K+ ions in the complex). The thermodynamics of ligand binding to each form, one ligand at a time, provides unprecedented detail on the interplay between ligand binding and topology changes in terms of number of G-quartets.

Influence of the metals and ligands in dinuclear complexes on phosphopeptide sequencing by electron-transfer dissociation tandem mass spectrometry.

Phosphorylation is one of the most important protein modifications, and electron-transfer dissociation tandem mass spectrometry (ETD-MS/MS) is a potentially useful method for the sequencing of phosphopeptides, including determination of the phosphorylation site. Notably, ETD-MS/MS typically provides useful information when the precursor contains more than three positive charges. It is not yet used as an analysis method for large-scale phosphopeptide production due to difficulties occurring in the production of acidic phosphopeptides having more than three positive charges. To increase the charge state of phosphopeptides, we used dinuclear metal complexes, which selectively bind to the phosphate group in phosphopeptides with the addition of positive charge(s). Dinuclear copper, zinc, and gallium complexes were tested and it was found that the type of metal present in the complex strongly affected the affinity of the phosphorylated compounds and their ETD fragmentation. The dinuclear copper complex interacted weakly with the phosphate groups and ETD-induced peptide fragmentation was largely suppressed by the presence of Cu2+, which worked as an electron trap. The dinuclear gallium complex was strongly bound to a phosphate group. However, the ligand binding to gallium acted as an electron trap and the presence of dinuclear gallium complex in the precursor for ETD-MS/MS hampered the sequencing of the phosphopeptides, as in the case of dinuclear copper complexes. In contrast, dinuclear zinc complexes efficiently bind to phosphopeptides with an increase in the charge state, facilitating phosphopeptide sequencing by ETD-MS/MS. The fragmentation of the ligand and peptide backbone in the dinuclear zinc-phosphopeptide complex were competitively induced by ETD. These processes are influenced by the ligand structure and so the detailed ETD fragmentation pathways were investigated using density functional theory calculations.

Unexpected Position-Dependent Effects of Ribose G-Quartets in G-Quadruplexes.

To understand the role of ribose G-quartets and how they affect the properties of G-quadruplex structures, we studied three systems in which one, two, three, or four deoxyribose G-quartets were substituted with ribose G-quartets. These systems were a parallel DNA intramolecular G-quadruplex, d(TTGGGTGGGTTGGGTGGGTT), and two tetramolecular G-quadruplexes, d(TGGGT) and d(TGGGGT). Thermal denaturation experiments revealed that ribose G-quartets have position-dependent and cumulative effects on G-quadruplex stability. An unexpected destabilization was observed when rG quartets were presented at the 5'-end of the G stack. This observation challenges the general belief that RNA residues stabilize G-quadruplexes. Furthermore, in contrast to past proposals, hydration is not the main factor determining the stability of our RNA/DNA chimeric G-quadruplexes. Interestingly, the presence of rG residues in a central G-quartet facilitated the formation of additional tetramolecular G-quadruplex topologies showing positive circular dichroism signals at 295 nm. 2D NMR analysis of the tetramolecular TGgGGT (lowercase letter indicates ribose) indicates that Gs in the 5'-most G-quartet adopt the syn conformation. These analyses highlight several new aspects of the role of ribose G-quartets on G-quadruplex structure and stability, and demonstrate that the positions of ribose residues are critical for tuning G-quadruplex properties.

Drift Tube Ion Mobility: How to Reconstruct Collision Cross Section Distributions from Arrival Time Distributions?

Ion mobility spectrometry allows one to determine ion collision cross sections, which are related to ion size and shape. Collision cross sections (CCS) are usually discussed based on the peak center, yet the width of each peak contains further information on the distribution of collision cross sections of each conformational ensemble. Here, we analyze how to convert arrival time distributions (ATD) to CCS distributions (CCSD). With a calibration curve taking into account the CCS dependence of the time spent outside the mobility region, one can reconstruct CCS distributions with correct peak center values. However, the peak widths are incorrectly rendered because ion diffusion, which affects the peak width in the time domain, is irrelevant to collision cross sections. For drift tube ion mobility, we describe a new method, coined "FWHMstep", using a step-field experiment and processing the peak's full width at half-maximum to reconstruct CCSDs. The width of the CCS distribution helps to characterize the analyte's structural heterogeneity, and/or its flexibility (i.e., the variety of ways the analyte ions can rearrange following electrospray into kinetically stable gas-phase conformations).

Specific Stabilization of c-MYC and c-KIT G-Quadruplex DNA Structures by Indolylmethyleneindanone Scaffolds.

Stabilization of G-quadruplex DNA structures by small molecules has emerged as a promising strategy for the development of anticancer drugs. Since G-quadruplex structures can adopt various topologies, attaining specific stabilization of a G-quadruplex topology to halt a particular biological process is daunting. To achieve this, we have designed and synthesized simple structural scaffolds based on an indolylmethyleneindanone pharmacophore, which can specifically stabilize the parallel topology of promoter quadruplex DNAs (c-MYC, c-KIT1, and c-KIT2), when compared to various topologies of telomeric and duplex DNAs. The lead ligands (InEt2 and InPr2) are water-soluble and meet a number of desirable criteria for a small molecule drug. Highly specific induction and stabilization of the c-MYC and c-KIT quadruplex DNAs (ΔT1/2 up to 24 °C) over telomeric and duplex DNAs (ΔT1/2 ∼ 3.2 °C) by these ligands were further validated by isothermal titration calorimetry and electrospray ionization mass spectrometry experiments (Ka ∼ 10(5) to 10(6) M(-1)). Low IC50 (∼2 µM) values were emerged for these ligands from a Taq DNA polymerase stop assay with the c-MYC quadruplex forming template, whereas the telomeric DNA template showed IC50 values >120 µM. Molecular modeling and dynamics studies demonstrated the 5'- and 3'-end stacking modes for these ligands. Overall, these results demonstrate that among the >1000 quadruplex stabilizing ligands reported so far, the indolylmethyleneindanone scaffolds stand out in terms of target specificity and structural simplicity and therefore offer a new paradigm in topology specific G-quadruplex targeting for potential therapeutic and diagnostic applications.

Anatomy of an Oligourea Six-Helix Bundle.

Non-natural synthetic oligomers that adopt well-defined secondary structures (i.e., foldamers) represent appealing components for the fabrication of bioinspired self-assembled architectures at the nanometer scale. Recently, peptidomimetic N,N'-linked oligourea helices have been designed de novo with the ability to fold into discrete helix bundles in aqueous conditions. In order to gain better insight into the determinants of oligourea helix bundle formation, we have investigated the sequence-to-structure relationship of an 11-mer oligourea previously shown to assemble into a six-helix bundle. Using circular dichroism, NMR spectroscopy, native mass-spectrometry and X-ray crystallography, we studied how bundle formation was affected by systematic replacement of the hydrophobic surface of the oligourea helix with either polar or different hydrophobic side chains. The molecular information gathered here has revealed several key requirements for foldamer bundle formation in aqueous conditions, and provides valuable insight toward the development of foldamer quaternary assemblies with improved (bio)physical properties and divergent topologies.

Assembly of chemically modified G-rich sequences into tetramolecular DNA G-quadruplexes and higher order structures.

In this review, we introduce the biophysical and biochemical methods currently used to investigate the structures and stabilities of tetramolecular DNA G-quadruplexes containing chemical modifications. We hope this paper will guide others as they perform similar experiments leading to more information about the effects of chemical modifications on G-quadruplex formation. The structures of tetramolecular quadruplexes and some higher order structures based on tetramolecular quadruplexes are also described.