Nicholas V Hud is at the Georgia Institute of Technology, Atlanta, USA and has over a decade of research in the area of nucleic acid-cation interactions.

Nucleic Acid–Metal Ion Interactions

By Nicholas V. Hud

The Royal Society of Chemistry

Copyright © 2009 Royal Society of Chemistry
All rights reserved.
ISBN: 978-0-85404-195-4

Contents

Chapter 1 Complexes of Nucleic Acids with Group I and II Cations Chiaolong Hsiao, Emmanuel Tannenbaum, Halena VanDeusen, Eli Hershkovitz, Ginger Perng, Allen R. Tannenbaum and Loren Dean Williams,
Chapter 2 Coordinative Bond Formation Between Metal Ions and Nucleic Acid Bases Bernhard Lippert,
Chapter 3 Sequence-specific DNA–Metal Ion Interactions Nicholas V. Hud and Aaron E. Engelhart,
Chapter 4 Metal Ion Interactions with G-Quadruplex Structures Aaron E. Engelhart, Janez Plavec, Özgül Persil and Nicholas V. Hud,
Chapter 5 Characterization of Nucleic Acid–Metal Ion Binding by Spectroscopic Techniques Victoria J. DeRose,
Chapter 6 Metal Ions and the Thermodynamics of RNA Folding David P. Giedroc and Nicholas E. Grossoehme,
Chapter 7 Metal Ions and RNA Folding Kinetics Somdeb Mitra and Michael Brenowitz,
Chapter 8 Metal Ions in RNA Catalysis John K. Frederiksen, Robert Fong and Joseph A. Piccirilli,
Chapter 9 Polyanion Models of Nucleic Acid–Metal Ion Interactions J. Michael Schurr,
Chapter 10 Metal Ion–Nucleic Acid Interactions in Disease and Medicine Ana M. Pizarro and Peter J. Sadler,
Subject Index, 417,

CHAPTER 1

Complexes of Nucleic Acids with Group I and II Cations

CHIAOLONG HSIAO, EMMANUEL TANNENBAUM, HALENA VanDEUSEN, ELI HERSHKOVITZ, GINGER PERNG, ALLEN R. TANNENBAUM AND LOREN DEAN WILLIAMS

1.1 Introduction

Recent structures of large RNAs, such as the P4–P6 domain of the Tetra -hymena ribozyme, and larger RNAs such as rRNAs, combined with a general increase over time in the sophistication of diffraction experiments, show cations in diverse and sometimes unexpected environments. The interactions of nucleic acids with cations follow basic principles of coordination chemistry. The effects of cations on RNA stability and conformation demonstrate the endurance of these relatively simple principles.

One focus of this chapter is the coordination of Na+, K+, Ca2+ and Mg2+ by phosphates and nucleic acids. We describe coordination chemistry, electrostatic forces/energetics, conformational effects and ion-selective binding. We explain the origins of the specific requirement for Mg2+ in RNA folding and the tight coupling between Mg2+ binding and RNA conformation. We deconstruct the two binding-mode formalism. We describe crystallographic methods for determining cation positions. We propose a model of RNA folding that is consistent with Mg2+ coordination properties of RNA. Previous reviews are available on roles of metals in biology, in polyelectrolyte theory (see Chapter 9), in DNA structure (see Chapter 3), in RNA folding (see Chapters 6 and 7) and in RNA catalysis (see Chapter 8).

1.1.1 Modern Treasure Troves of Structural Information: Large RNAs

Very large RNA assemblies are now available at high resolution. The largest and most accurate structures are used here in conjunction with smaller structures, down to the level of mononucleotides, to illustrate patterns of interaction of nucleic acids with cations. 23S-rRNAHM refers to the 23S rRNA from the archaeon H. marismortui (2.4 Å resolution, PDB entry 1JJ2), a halophile from the Dead Sea. 23S-rRNATT refers to the 23S rRNA from the eubacterium T. thermophilus (2.8 Å resolution, PDB entry 2J01), isolated from a thermal vent. The fractional sequence identity of the 23S rRNAs from HM and TT is around 60%. RNAP4–P6 refers the 160 nucleotide domain of the self-splicing Tetrahymena thermophila intron (2.3 Å resolution, PDB entry 1HR2, this ΔC209 mutant gives the best available resolution).

1.2 Nucleic Acid Folding

1.2.1 Cations

During protein folding, water molecules in contact with hydrophobic surfaces are released to bulk solvent. During nucleic acid folding, cations are sequestered from bulk solvent and held in close proximity to the polymer. Protein side-chains are multifarious, with a variety of shapes and chemical properties. The nucleic acid backbone is intricate, with many accessible rotameric states, and carries charge.

Functional nucleic acids generally fold into compact and stable states of given conformation. DNA can form quadruplexes, triplexes, i-motifs, etc. Structured RNAs range in size from aptamers and tRNAs to ribosomes. However some functional nucleic acids, such as riboswitches, are conformationally polymorphic. For our purposes, folded nucleic acid structures fall into three general classes: (i) helical structures such as A-form, B-form and triplexes, (ii) quasi-globular structures such as tRNA, with base–base tertiary interactions but no buried phosphates, and quadruplexes, and (iii) true globular structures such as the Tetrahymena ribozyme and its P4–P6 domain, with base–base tertiary interactions plus buried OP atoms (OP indicates a non-bridging phosphate oxygen). True globular structures have distinct ‘insides’ and ‘outsides’. Folding of helices, quasi-globular structures and true globular structures increases proximities of phosphate groups and the electrostatic repulsion among them. Therefore, folding is intrinsically linked to association with cations. Phosphate-phosphate repulsion must be offset by attraction between phosphates and cations. Cations most strongly associate with regions of DNA and RNA in which phosphate groups assume greatest ‘density’.

As will become clear in the following sections, Mg2+ stabilizes distinctive conformational and energetic states of nucleic acids. Mg2+ shares a special geometric and electrostatic complementarity with phosphate, with a specific coordination and thermodynamic fingerprint. These states are simply not accessible in the absence of Mg2+ (or Mn2+), even when other cations are at high concentration. The thermodynamic and conformational consequences of first-shell OP interactions with Mg2+ are different to those for neutral ligands or for other cations, with lesser charge or greater size.

1.2.2 The RNA Folding Hierarchy

RNA folding is hierarchical. Folding progresses through a series of intermediates that are commonly characterized by extents and types of base–base hydrogen bonding and stacking interactions. The unfolded state, the random coil, is a conformationally polymorphic and fluctuating ensemble with few local or long-range base–base interactions. Early intermediates contain double-stranded stems and hairpin loops, interspersed by single-stranded regions. These stems and loops are known collectively as secondary structural units. Late intermediates and the final folded state are stabilized by base–base tertiary interactions, between residues that are remote in the secondary structure. To a first approximation, secondary structure can be conceptually and experimentally separated from tertiary structure. Secondary structure forms before tertiary structure and is favorable in a broad range of ionic conditions. Tertiary structure is favored by divalent cations. Although compact structures with base–base tertiary interactions can be achieved at very high concentrations of other cations, for true globular structures, the fully folded state is absolutely dependent on Mg2+. It can be useful to make a distinction between the tertiary structure of an RNA, which is a description of short- and-long range base–base interactions, and a folded RNA, which is a description of three-dimensional positions of all atoms, including of course the phosphate groups.

A hierarchical model that focuses exclusively on base–base interactions is a useful but somewhat limited approximation. In true globular structures, ground states are stabilized by specifically associated Mg2+ ions, each with up to four first-shell OP ligands. The importance of Mg2+ (OP)3 and Mg2+ (OP)4 coordination complexes is discussed in later sections. Multidentate interactions of OP atoms with Mg2+ are generally local along the RNA backbone. A small and important subset of OP ligands of a common Mg2+ are remote in the secondary and primary structures, thus forming ‘electrostatic tertiary interactions’ (Mg2+ mediated linkages between remote OP groups). Extensive base-base tertiary interactions during folding do not necessary imply the formation of electrostatic tertiary interactions. At least some globular RNAs fold into compact (but non-native) structures, with extensive base-base tertiary interactions – in the absence of Mg2+. We do not know at present if the converse is true (i.e. are electro -static tertiary interactions fully dependent on correct base-base tertiary interactions?). At any rate, to understand and describe fully the structure of a globular RNA, one can extend a conventional tertiary description of base–base interactions to include electrostatic tertiary interactions.

1.2.3 Alternative RNA Folding Hierarchies

To illuminate the underlying dependence of folding on cations, one can re-state the hierarchy of RNA folding using ‘phosphate density’. In early folding steps, a subset of phosphate–phosphate distances decreases from > 7Å (P to P) in random coil to around 5.8–6.2Å (in A-form helical regions and loops). In subsequent steps, a subset of P to P distances decreases further, to 5.0–4.6 Å. Associated with this group of short P to P distances are tightly packed anionic OP atoms, which are in van der Waals contact with each other (dOP–OP = 2.8–3.2 Å). This tight packing of anionic oxygen atoms is dependent on multidentate chelation of Mg2+ by OP atoms. Neither monovalent cations nor polyamines can substitute for Mg2+ in stabilizing structures with such short OP–OP contacts. During folding, some phosphates and associated Mg2+ ions become buried in the globular interior.

1.3 Coordination Chemistry

The binding of ligands to Group I and II cations is dictated by the chemical properties of the cations and of the ligands and, to a significant extent, by interactions between ligands. Chelators, with covalently linked ligands, create cavities for ions and bind with greater affinity and selectivity than monomeric ligands. The length of the chelator linker is a critical component of stability. As the linker length increases, the entropic cost of assembling the ligands for joint coordination increases. Hud and Polak previously noted the chelation properties of DNA, calling it an ionophore. Here we illustrate how the phosphate groups of nucleic acids commonly act as chelators of cations.

1.3.1 Group I

Group I cations prefer hard neutral ligands or one singly charged ligand plus additional neutral ligands. In their associations with nucleic acids, Group I cations are most commonly associated with non-anionic oxygens (i.e. oxygen atoms other than OP) as inner shell ligands.

The monovalent cations [sodium (Na+), potassium (K+), rubidium (Rb+), caesium (Cs+), thallium (Tl+) and ammonium (NH4+), excluding lithium (Li+)] are characterized by relatively large ionic radii, low charge density and modest enthalpies of hydration (Table 1.1). The coordination chemistry of Li+, with its small atomic radius and high charge density, is distinct from that of other Group I metals. Tl+ and NH4+ are listed here along with the Group I metals because they are well-developed K substitutes with useful spectroscopic and crystallographic signals. NH4+ positions are indicated by NOEs in solution. Tl+ positions are indicated in solution by NMR and in crystals by a distinctive X-ray scattering signal (anomalous scattering). Tl+, K+ and NH4+ have similar ionic radii and enthalpies of hydration (Table 1.1).

These monovalent ions, except Li+, display irregular and variable coordination geometry. The variability in coordination geometry is associated with non-covalency of interaction, weak ligand–ligand interactions and loose ligand–ligand packing. For a given monovalent ion, the number of first-shell ligands can vary from four to over 10. These properties are quantitated by ‘average observed coordination numbers’ (AOCN, Table 1.1) over a large number of structures within the Cambridge Structural Database as reported by Brown.

1.3.1.1 Na+

The ideal Na+ to oxygen distance is 2.4 Å (Figure 1.1a). The distance between first-shell ligands of Na+ is variable, depending on coordination number and coordination geometry. An octahedral arrangement of first shell oxygen ligands is loosely packed. The O to O distance is 3.4 Å, which is significantly greater than twice the van der Waals radius of oxygen (oxygen radius = 1.4 Å). Therefore, inner shell ligands of Na+ are not crowded and the geometry of the Na+ inner sphere is not determined by ligand–ligand interactions. An Na+ ion with ideal octahedral geometry in association with the O6 position of a guanine of DNA with five water molecule inner-shell ligands is shown in Figure 1.1a. As can be seen, the Na+ to O distances average around 2.4 Å, whereas the distance between cis oxygen atoms (adjacent oxygen ligands) averages around 3.4 Å. DNA-cation interactions are discussed in Chapter 3.

13.1.2 K+

The ideal K+ to oxygen distance is around 2.7 Å. For an octahedral arrangement of first-shell oxygen ligands, the average O to O distance is over 4.0 Å. Hence the inner-shell ligands of K+ are of less defined geometry than those of Na+. Specific K+ binders, that exclude Na+, are generally composed of stacked, planar arrangements of keto oxygens. In these K+-selective structures, the positions of the keto oxygens (Oketo) are fixed such that the enthalpy of dehydration is compensated by Oketo–K+ interactions but not by Oketo–Na+ interactions, which are too long.

1.3.2 Group II

1.3.2.1 Ca2+

Ca2+ often shows irregular coordination geometry and coordination numbers greater than six. The ionic radius of Ca2+ is ca. 0.99 Å. Compared with Mg2+, Ca2+ is large, its charge density is low, its inner-shell ligands are loosely packed and the magnitude of is hydration enthalpy is small (Table 1.1). However, like Mg2+, Ca2+ prefers a mix of anionic and neutral ligands.

1.3.2.2 Mg2+

Mg2+, from life’s beginning, has been closely associated with some of the central players in biological systems – phosphates and phosphate esters. shares a special geometric, electrostatic and thermodynamic relationship with phosphates and phosphate esters. In comparison with group I ions, Ca2+ or polyamines, Mg2+ has a greater affinity for OP atoms and binds to OP with well-defined geometry. Unlike other cations, Mg2+ brings OP atoms into direct contact with each other.

The ionic radius of Mg2+ is small (ca. 0.65 Å), the charge density is high, the six ligands of an octahedral inner first shell are tightly packed and the magnitude of the hydration enthalpy is large (Table 1.1). The heat of hydration of Mg2+ is much greater than for other biological cations (Table 1.1). Mg2+ prefers two to four oxyanions (along with a complement of water molecules) over uncharged oxygens and nitrogens as inner-shell ligands.

The first coordination sphere of Mg2+ assumes octahedral geometry, as shown in Figure 1.1b. The AOCN of Mg2+ is 5.98. Because Mg2+ is small and highly charged, ligand–ligand crowding is one of the hallmarks of Mg2+ complexes, leading to highly restrained ligand–Mg2+–ligand geometry and strong ligand–ligand repulsive forces. Although probably not germane to nucleic acid structure, four-coordinate Mg2+ is observed at high temperature in the gas phase.

1.3.2.2.1 Hexaaqua Mg2+ Complexes. In hexaaqua complexes [Mg2+(H2O) 6 or Mg2+aq], oxygen–Mg2+ distances are 2.07 Å. The cis O ··· Mg2+ ··· O angle is 90° and the cis O ··· O distance is 2.93 Å. The trans O ··· Mg ··· O angle is 180°. Adjacent oxygen atoms in Mg2+aq are in van der Waals contact. First-shell water molecules are strictly oriented such that their dipole moments are directed in towards the metal, with Mg2+ ··· O–H angles of 120–128°. This orientation prevents hydrogen bonding between water molecules in the Mg2+ first coordination shell.

1.3.2.2.2 ADP–Mg2+ Complexes. The nature of complexes formed by Mg2+ with ADP and ATP are highly predictive of Mg2+ interactions with other multidentate OP ligands such as RNA. Some of the ADP–Mg2+ and ATP–Mg2+ complexes identified in the Protein Data Bank are shown in Figure 1.2 and here we discuss the implications of these and related structures provided by X-ray crystallography. One can immediately appreciate that ADP and ATP are mono- and multidentate chelators of Mg2+, contributing OP ligands to the inner coordination sphere. In this section, we focus on structural aspects of Mg2+ complexes with nucleotides. Thermodynamic aspects are discussed in a subsequent section.

Mg2+ interacts with non-bridging OP atoms, but not with bridging oxygens, base or sugar atoms of ADP/ATP. The OP atoms of ADP bind to Mg2+ by either monodentate interactions (40% of structures surveyed; Figure 1.2a and b, Table 1.2 and Table 1.3) or by bidentate chelation (60%, Figure 1.2c–e). Monodentate interactions occur exclusively by OP whereas bidentate chelation involves OαP and OβP.

Chelation ring size is an important factor in modulating stability (see Section 1.3). The bidentate chelation complexes of Mg2+ with ADP/ATP are composed of six-membered rings consisting of atoms Mg2+–OαP –P–O–P–OβP–Mg2+ or (Mg2+–OβP–P–O–P–Oγ P–Mg2+). Bidentate chelation by two OP atoms bound to a common phosphorus atom would require a chelation ring size of four and is not observed in ADP/ATP complexes. When Mg2+ is monochelated by ADP, at least one protein ligand is also found in the Mg2+ first shell. The protein ligand is invariably oxygen, but may be charged or neutral. Similarly, most bidentate ADP–Mg2+ complexes contain protein first-shell ligands. OP ligands are adjacent to each other within the octahedron of first-shell ligands of Mg2+. All bidentate Mg2+–ADP complexes assume cis (OP–Mg2+–OP angle 90°), cis–cis or cis–cis–cis orientation of the ligands surrounding the Mg2+ depending on the number of first shell protein ligands. Trans bidentate Mg2+– ADP complexes (OP–Mg2+–OP angle 180°) are not observed and appear to be stereochemically prohibited.

(Continues…)Excerpted from Nucleic Acid–Metal Ion Interactions by Nicholas V. Hud. Copyright © 2009 Royal Society of Chemistry. Excerpted by permission of The Royal Society of Chemistry.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.