RNA Folding and Catalysis Mediated by Iron (II)

Abstract

Mg2+ shares a distinctive relationship with RNA, playing important and specific roles in the folding and function of essentially all large RNAs. Here we use theory and experiment to evaluate Fe2+ in the absence of free oxygen as a replacement for Mg2+ in RNA folding and catalysis. We describe both quantum mechanical calculations and experiments that suggest that the roles of Mg2+ in RNA folding and function can indeed be served by Fe2+. The results of quantum mechanical calculations show that the geometry of coordination of Fe2+ by RNA phosphates is similar to that of Mg2+. Chemical footprinting experiments suggest that the conformation of the Tetrahymena thermophila Group I intron P4–P6 domain RNA is conserved between complexes with Fe2+ or Mg2+. The catalytic activities of both the L1 ribozyme ligase, obtained previously by in vitro selection in the presence of Mg2+, and the hammerhead ribozyme are enhanced in the presence of Fe2+ compared to Mg2+. All chemical footprinting and ribozyme assays in the presence of Fe2+ were performed under anaerobic conditions. The primary motivation of this work is to understand RNA in plausible early earth conditions. Life originated during the early Archean Eon, characterized by a non-oxidative atmosphere and abundant soluble Fe2+. The combined biochemical and paleogeological data are consistent with a role for Fe2+ in an RNA World. RNA and Fe2+ could, in principle, support an array of RNA structures and catalytic functions more diverse than RNA with Mg2+ alone.

About the Authors:

Shreyas S. Athavale

Affiliations School of Biology, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Anton S. Petrov

Affiliations School of Biology, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Chiaolong Hsiao

Affiliations School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Derrick Watkins

Affiliations School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Caitlin D. Prickett

Affiliations School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

J. Jared Gossett

Affiliations School of Biology, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Lively Lie

Affiliations School of Biology, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Jessica C. Bowman

Affiliations School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Eric O'Neill

Affiliations School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Chad R. Bernier

Affiliations School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Nicholas V. Hud

Affiliations School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Roger M. Wartell

Affiliations School of Biology, Georgia Institute of Technology, Atlanta, Georgia, United States of America, School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Stephen C. Harvey

Affiliations School of Biology, Georgia Institute of Technology, Atlanta, Georgia, United States of America, School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Loren Dean Williams

* E-mail: [email protected]

Affiliations School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia, United States of America, NAI Center for Ribosomal Origins and Evolution, Georgia Institute of Technology, Atlanta, Georgia, United States of America

Introduction

When large RNAs fold into compact structures, negatively charged phosphate groups achieve close proximity. Folded RNAs are stabilized in part by inorganic cations that accumulate in and around the RNA envelope. ‘Diffuse’ cations remain hydrated and make primary contributions to global stability by mitigating electrostatic repulsion of the negatively charged backbone. Chelated ions are less frequent, but in some instances are essential for achieving specific local conformation of the RNA. A special importance of Mg2+ in tRNA folding was seen early on [1]–[3]. It is now known that Mg2+ plays important roles in folding of essentially all large RNAs [4]–[6]. In addition, Mg2+ ions assist directly in stabilizing transition states of some ribozymes [7], [8].

Here we use computation and experiment to address the question of whether Fe2+ can substitute for Mg2+ in RNA folding and catalysis. Mg2+ possesses important electronic and geometric properties that are key to its relationships with RNA. It is redox inactive, and does not cleave RNA via Fenton chemistry. The ionic radius of Mg2+ is small, the charge density is high, the coordination geometry is strictly octahedral, and the hydration enthalpy is large and negative [9]–[11]. In comparison with Group I cations, calcium, or polyamines, Mg2+ has a greater affinity for phosphate oxygens [12]. We find that Fe2+ is an excellent Mg2+ mimic in the absence of O2, readily substituting for Mg2+ in RNA folding and catalysis.

Our primary motivation is to study RNA under plausible early earth conditions. Understanding the influence of Fe2+ on RNA structure and function could provide important links between the geological record and the RNA world. It is believed that life originated with RNA-based genetic and metabolic systems, i.e. the RNA world [13], which apparently flourished in an anoxic environment in which iron was much more soluble and abundant than in our current oxidative environment. Life evolved for around a billion years before the rise of photosynthesis and the Great Oxidation Event [14], [15]. Fe2+, either instead of or in combination with Mg2+, seems to be a possible partner of RNA in the biology of the pre-photosynthesis anoxic earth.

With the rise in free oxygen, a product of photosynthesis, the Fe2+ of the early earth was oxidized and sequestered. Iron was deposited in banded iron formations (BIFs) [16], but BIF iron is seen by isotopic variations to have been a participant in ancient biological processes [17]. The transition from soluble to insoluble iron caused slow but dramatic shifts in biology and geology.

Results

Theory predicts that RNA conformation is conserved if Fe2+ substitutes for site-bound Mg2+

Quantum mechanical (QM) calculations show that RNA conformation and coordination geometry are conserved when Mg2+ is replaced by Fe2+ in first shell RNA-metal complexes. We focused here on an RNA-Mg2+ clamp [18], in which two adjacent RNA phosphates coordinate a common Mg2+ (Figure 1A). A complex with multiple first-shell RNA interactions with Mg2+ should provide a stringent test of the ability of Fe2+ to substitute. RNA-Mg2+ clamps are common in large RNAs [19]. One observes twenty-five RNA-Mg2+ clamps in the Haloarcula marisortui large ribosomal subunit [20], two in the P4-P6 domain of the Tetrahymena thermophila Group 1 intron [21], one in a self-splicing group II intron from Oceanobacillus iheyensis [22], and one in the L1 ribozyme ligase [23]. The folding and function of each of these RNAs is Mg2+ dependent.

[Figure omitted. See PDF.]

Figure 1. Conformations of RNA-Mg2+ and RNA-Fe2+ clamps are nearly identical.

A) RNA-Mg2+ clamp from the L1 ribozyme ligase (PDB 2OIU). B) RNA-Mg2+ clamp optimized by high level QM calculations. C) An optimized RNA-Fe2+ clamp. Each cation (Mg2+ or Fe2+) is hexacoordinate. Mg2+ is shown as a yellow sphere and Fe2+ is shown as a green sphere. Water molecules are omitted from the images for clarity. Distances are in Å.

https://doi.org/10.1371/journal.pone.0038024.g001

The conformations of an RNA-Mg2+ clamp and an RNA-Fe2+ clamp are nearly identical, as determined by Density Functional Theory (DFT) [24]. The RNA conformation and metal-oxygen distances and angles are very similar (Figures 1B and 1C). The calculations do indicate some subtle differences, however. Calculated interaction energies (energies of complex formation, Table S1) favor Fe2+ over Mg2+ by 1.3 kcal/mol in continuum solvent as indicated by DFT calculations. Natural Bond Orbital analysis [25] in the gas phase (Table S2) suggests that more charge is transferred from phosphate to Fe2+ (0.43 e-) than from phosphate to Mg2+ (0.29 e-). This difference implies that compared to Mg2+, Fe2+ better activates the phosphorous of RNA to nucleophilic attack. The increase in activation is attributable to the accessibility of the d-orbitals of Fe2+.

Chemical probing suggests that RNA conformation is conserved when Fe2+ substitutes for Mg2+

Selective 2′-hydroxyl acylation analyzed by primer extension (SHAPE) is a powerful RNA foot-printing technique that provides structural information at single-nucleotide resolution [26]–[28]. SHAPE has been used to accurately predict and confirm secondary structures of RNA ranging in length from tRNA to the HIV-1 genome [27], [29]. The method exploits the reactivity of the 2′-hydroxyl groups of RNA to electrophiles to form 2′-O-ribose adducts. Here we employed the SHAPE reagent benzoyl cyanide (BzCN). The relative reactivities of ribose 2′-hydroxyl groups to the electrophile are sensitive to local RNA flexibility, which depends primarily on whether or not a nucleotide is base-paired. Single-stranded nucleotides react preferentially over double-stranded nucleotides. Reverse transcription using fluorescently labeled primers gives products that are truncated at locations indicating 2′-O-ribose adducts. The fragments are resolved and visualized using capillary electrophoresis. Capillary electrophoresis data were processed as described [30].

The secondary structure of the T. thermophila Group I intron P4–P6 domain was assayed by SHAPE in the presence of Na+ alone, giving a reaction pattern consistent with the known secondary structure [21] (Figure 2A). For example, in the stem-loop formed by residues 143–160, the double-stranded nucleotides of the stem are unreactive while the GAAA nucleotides of the loop are reactive. Some of the most reactive nucleotides of the P4–P6 domain secondary structure are located within the A-rich bulge (nucleotides 182–188).

[Figure omitted. See PDF.]

Figure 2. Addition of Mg2+ or Fe2+ causes the same changes in the SHAPE reactivity of the P4–P6 domain of the T. thermophila Group 1 intron.

A) Shape profile in presence of 250 mM NaCl and no divalent cations. B) The addition of Mg2+ increases the reactivity at the sites indicated with the asterisks and decreases reactivity at other sites. This reaction contains 2.5 mM Mg2+ and 250 mM NaCl. C) The addition of Fe2+ causes the same changes in SHAPE reactivity as Mg2+. This reaction contains 2.5 mM Fe2+ and 250 mM NaCl.

https://doi.org/10.1371/journal.pone.0038024.g002

We probed the structure of the P4–P6 domain RNA in presence of Mg2+ (Figure 2B). The folding of RNAs from secondary structure to compact native states, containing long-range tertiary interactions, is known to be Mg2+-dependent [4]–[6]. The addition of 2.5 mM Mg2+ to the P4–P6 domain RNA causes pronounced changes in the SHAPE reactivity. SHAPE reactivity increases at nucleotides 122, 125, 177–179 and 198–200 (indicated by asterisks in Figure 2B). The Mg2+-dependence of SHAPE reactivities reflects (i) specific magnesium binding, (ii) diffuse interactions of Mg2+, and (iii) RNA-RNA tertiary interactions facilitated by Mg2+, as previously demonstrated for tRNA [27], RNase P [31], and Domain III of the ribosomal large subunit [30]. The pattern of SHAPE reactivity for P4–P6 domain RNA in the presence of Mg2+ observed here is nearly identical to that described previously for the same RNA in the presence of Mg2+ by Cech and coworkers (using N-methylisatoic anhydride instead of BzCN) [32].

The pattern of SHAPE reactivity for P4–P6 domain RNA is conserved when Mg2+ is replaced by Fe2+ under anaerobic conditions. Figure 2C shows that SHAPE reactivities in presence of 2.5 mM Fe2+ are identical, within the accuracy of the experiment, to those in presence of 2.5 mM Mg2+. These results suggest that tertiary interactions and even the ‘ion core’ of the P4–P6 domain are recapitulated by Fe2+ in the absence of oxygen. As expected, if Fe2+ is added to the RNA in the presence of atmospheric free oxygen, the RNA is quickly degraded (not shown).

Activity of two ribozymes is enhanced by Fe2+ compared to Mg2+

To investigate RNA function in presence of Fe2+, we tested the catalytic activity of the L1 ribozyme ligase in the presence of Mg2+ or Fe2+ (in the absence of oxygen). This ligase catalyzes formation of a phosphodiester linkage. The 3′-hydroxyl group of an RNA substrate attacks the α-phosphorus of the ribozyme 5′-triphosphate [23]. This ribozyme was selected in vitro in the presence of high [Mg2+] (60 mM) by Robertson and Ellington, and has been described as Mg2+-dependent [23]. The initial rate of ligation in 100 µM Mg2+ is 1.4×10−6 min−1, while the initial rate of ligation in 100 µM Fe2+ is 3.5×10−5 min−1, which is 25-fold higher (Figure 3A). A higher rate for Fe2+ over Mg2+ holds for essentially any reasonable equimolar concentration of the two cations. Achieving an equivalent rate of reaction requires around a 10-fold greater [Mg2+] than [Fe2+]. As expected, this ligase is inactive in Na+ alone. This control, along with the chemical footprinting of the P4–P6 domain in Na+ alone (Figure 2A), confirms the efficacy of our divalent cation extraction procedure using divalent cation chelating beads.

[Figure omitted. See PDF.]

Figure 3. Ribozyme activity is enhanced by Fe2+ compared to Mg2+.

A) L1 ribozyme ligase activity is enhanced in Fe2+ compared to Mg2+. Ligase reactions were performed under anaerobic conditions at room temperature and 250 mM Na+ in 100 µM [Fe2+] or 100 µM [Mg2+]. The reaction components were first annealed in 50 mM HEPES, pH 8.0, 200 mM sodium acetate by incubating at 90°C for 3 min and cooling to room temperature over 30 min. The ligation reaction was initiated by adding the appropriate cation salt. The Na+ only reaction gave no product. Reaction progress was monitored by gel electrophoresis. B) Hammerhead ribozyme activity is enhanced in Fe2+ compared to Mg2+. Hammerhead ribozyme cleavage reactions were performed under anaerobic conditions at room temperature in 50 mM HEPES, pH 7.5 and 25 µM [Fe2+] or 25 µM [Mg2+]. Substrate and ribozyme RNA strands were first annealed in 50 mM HEPES buffer by incubating at 90°C for 2 min and cooling to room temperature over 30 min. Cleavage reactions were initiated by addition of FeCl2 or MgCl2 from stock solutions. Reactions were monitored by both gel electrophoresis and capillary electrophoresis, which gave similar results.

https://doi.org/10.1371/journal.pone.0038024.g003

The hammerhead ribozyme was also assayed for activity in the presence of Fe2+. Hammerhead ribozyme sequences are widely distributed in the tree of life [33]. This ribozyme cleaves the RNA backbone via nucleophilic attack by a 2′-hydroxyl group on the 3′-phosphorous atom [34]. In these reactions the initial rate of hammerhead cleavage in 25 µM Mg2+ is 0.011 min−1, while the initial rate of cleavage in 25 µM Fe2+ is 0.035 min−1, which is 3-fold higher (Figure 3B). The maximum fraction of cleaved substrate was about 3-fold greater in Fe2+ versus Mg2+. When 100 µM of these two divalent cations were employed, Fe2+ again showed a higher initial rate of cleavage of ∼3.5 fold (data not shown).

Discussion

The results here support a model of early evolution in which Fe2+ was an important metallo-cofactor for RNA. In this model, Fe2+ was replaced over time by Mg2+, by processes driven at least in part by photosynthesis and the Great Oxidation Event.

The RNA World – on Steroids

The RNA world is hypothesized to have occurred during the early Archean eon, prior to the Great Oxidation Event. Fe2+ in the early Archean would have been available, soluble and non-toxic. Our observations here of Fe2+-mediated RNA folding and catalysis, in combination with paleogeological information, suggest that RNA could have originated and evolved in association with Fe2+. The RNA-Fe2+ complexes recently observed in extant biology [35] could be molecular fossils from the RNA world, akin to the ribosome. The injection of Fe2+ into RNA World models opens broad new possibilities for ancient biochemistry. RNA and Fe2+ could, in principle, support an array of RNA structures and catalytic functions far more diverse than RNA with Mg2+ alone. Complexes of RNA with Fe2+ offer the prospect of redox chemistry and electron transfer reactions for ancient ribozymes.

Replacement of Fe2+ by Mg2+ in RNA is analogous to replacement of Fe2+ by Mn2+ in protein enzymes

The conversion of one metal to another is facile in some protein enzymes. In just one example, Mn2+ and Fe2+ are used as cofactors in a broad class of superoxide dismutases (Fe2+/Mn2+ SODs) [36]. The metal cofactors of these SODs can be interconverted between Mn2+ and Fe2+ while the coordination geometry, amino acid sequence and global fold of the protein are conserved. The discrimination between Fe2+ and Mn2+ in Fe2+/Mn2+ SODs in vivo is determined by species, organelle, protein isozyme, protein expression level and metal bioavailability. Metal substitution appears to be a useful biological strategy in nutrient-limited environments [37]–[39]. Falkowski has proposed that during the Great Oxidation Event, Mn2+ was appropriated into some Fe2+ dependent enzymes [37]. Here we suggest that the same strategy was employed with RNA, where Fe2+ was converted to Mg2+. It has been suggested, based on sites of Fenton cleavage (using iron/O2), that Fe2+ and Mg2+ compete for common sites in RNA [40]–[42] in vitro.

QM calculations suggest that an RNA fragment that forms multiple first-shell interactions with Mg2+ does not change conformation when Mg2 is replaced by Fe2+

The metal-oxygen distances and angles are nearly identical in the Mg2+ and Fe2+ complexes (Figures 1B and 1C). The QM calculations do indicate subtle differences between Mg2+ and Fe2+ complexes. It appears that more charge is transferred from phosphate to Fe2+ than from phosphate to Mg2+. This electronic difference, which activates the phosphorus atom to nucleophilic attack, is attributable to the accessibility of the d-orbitals of Fe2+.

Chemical probing experiments in solution, using the T. thermophila Group I intron P4–P6 domain RNA, demonstrate the ability of Fe2+ to substitute for Mg2+ during folding of large RNAs

The P4–P6 domain interacts with Mg2+ by a complex blend of diffuse and chelated modes [21], [43], [44]. In spite of this complexity, the changes in SHAPE reactivity of RNA induced by association with Mg2+ or Fe2+ in the absence of free oxygen are very similar (Figures 2B and 2C). SHAPE reports local RNA flexibility, which depends primarily on whether or not a nucleotide is base-paired in secondary or tertiary interactions. The results for the P4–P6 domain suggest that in the absence of free oxygen, Fe2+ can replace Mg2+ in compacting and folding large RNAs. Thus it appears that Fe2+ and Mg2+ are nearly interchangeable in their interactions with RNA.

Fe2+ can substitute for Mg2+ to support catalysis by ribozymes

At equimolar concentrations of Mg2+ or Fe2+, the initial rate of ligation observed for the L1 ribozyme ligase is 25-fold higher with Fe2+ than Mg2+ in the absence of free oxygen (Figure 3A). Similarly, at equimolar concentrations of Mg2+ or Fe2+, the initial rate of RNA cleavage observed for the hammerhead ribozyme is 3-fold higher with Fe2+ than Mg2+ (Figure 3B). In sum, we have looked at RNA folding in three independent experimental systems, by chemical footprinting (P4–P6 domain), and with two ribozyme assays (L1 ribozyme ligase and the hammerhead ribozyme). In each system examined, Fe2+ substitutes for Mg2+ in the absence of free oxygen. The increased activities of the ribozymes with Fe2+ over Mg2+ are consistent with our computational results that suggest Fe2+ is slightly better than Mg2+ at activating the phosphorous of RNA to nucleophilic attack. In the hammerhead, which is one of the best-characterized ribozymes, it has been shown that a Mg2+ ion interacts directly with the scissile phosphate before and during catalysis [45]. The results here suggest that Fe2+ is a superior substitute for Mg2+ in this catalytic role. A variety of effects such as differential affinity of Fe2+ and Mg2+ globally, or for various sites on RNA, could also contribute to differences in cleavage rates.

How much Fe2+ was available during the time of the RNA World?

It seems very likely that the early Archean earth provided a variety of Fe2+-rich microenvironments. On a global scale, the [Fe2+]marine in the early Archean is subject to debate, and is largely circumscribed in current models by PCO2 (atmospheric pressure of CO2) in the atmosphere (Fe2+ precipitates as siderite: Fe2+CO3−). Early earth PCO2 is inferred using estimates of greenhouse effects, sun luminosity, earth albedo and temperatures required to maintain liquid oceans. A variety of recent results challenge high PCO2 models [46], [47]. If PCO2 was low, [Fe2+]marine could have been as high as 100–1000 µM [14], [48], compared to 0.3–0.8 nM in the modern ocean [49]. PHS (atmospheric pressure of HS) would also have been an important influence on [Fe2+]marine due to precipitation of FeS [50].

How much Mg2+ was available during the RNA World?

The [Mg2+]marine in the early Archean is also uncertain. Although the models are tentative, it has been suggested that [Mg2+]marine during this time was attenuated by submarine hydrothermal systems associated with higher heat flow [51], [52], more vigorous seafloor spreading [53], [54], and by reduced Mg2+ delivery to the oceans by smaller continental landmasses or from the weathering of peridotite in the sea floor [55]. Both of these phenomena would tend to lower [Mg+2]marine of early oceans in comparison to today.

Fe2+ is the double-edged sword

The early earth's abundant Fe2+ has been oxidized and sequestered to the extent that current biomass and species diversity in many ecosystems is limited by Fe2+ availability [56], [57]. Iron in the presence of oxygen is rare, toxic, and biologically expensive to manage [58], [59]. Yet living systems are dependent on and must acquire and utilize iron. The concentration of iron in cells is on the order of 100 µM, with iron largely constrained to heme, iron-sulfur clusters, and di-iron or mono-iron centers, transporters, carriers, exporters, and concentrators such as ferritins [60]. Because the solubility of ferric iron in water or plasma is so low (10−18 M), cells must combat a massive concentration gradient. The transition from benign and abundant iron to scarce and toxic iron would have caused a slow but dramatic shift in biology that required transformations in biochemical mechanisms and metabolic pathways.

Materials and Methods

QM calculations

The initial atomic coordinates of a Mg2+-RNA clamp were extracted from the X-ray structure of the H. marismortui large ribosomal subunit (PDB entry 1JJ2) [20] as previously described [19]. The 5′ and 3′ phosphates were capped with methyl groups in lieu of the remainder of the RNA polymer and hydrogen atoms were added, where appropriate. The Fe2+-RNA clamp was constructed by replacing the magnesium ion with an iron as described [18].

The binding of a Mg2+ or Fe2+ ion to an RNA fragment was described by the following reactions:

RNA2- + Me2+(H2O)6 → RNA2--Me2+.•(H2O)4 complex + 2H2O,

where Me2+  =  (Mg2+, Fe2+)

The reactants and products were fully optimized using density functional theory (DFT) at the B3LYP level, which combines the GGA exchange three-parameter hybrid functional developed by Becke [61] and the correlation functional of Lee-Yang-Parr [62] and the 6–311G(d,p)++ basis set and multiplicity = 1 as implemented in Gaussian 09 [63]. The Fe2+-RNA clamp and the Fe2+(H2O)6 were optimized at the unrestricted B3LYP/6–31G(d,p) level of theory with spin of iron = 2 and multiplicity = 5. Single point energies for these complexes were further obtained at the UB3LYP/6–311++G(d,p) level of theory using SCF options DIIS, NOVARACC, VTL, MaxCyc = 1000.

The interaction energies were calculated in water using the gas phase optimized geometries within the framework of the Integral Equation Formalism of Polarized Continuum Model [24]. The basis set superposition error (BSSE) in the dimer-centered basis set was obtained by applying the counterpoise procedure of Boys and Bernardi [64]. The corrected interaction energies were calculated by taking into account deformational energies of monomers according to the scheme proposed by van Duijneveldt-van de Rijdt and van Duijneveldt [65]. The IEFPCM approach was used to account for the effect of a polar solvent.

Natural Bond Order (NBO) [25] and Natural Energy Decomposition Analysis (NEDA) [66], [67] calculations were performed on the optimized complexes at the (U)B3LYP/6–31G(d,p) level of theory using the GAMESS package [68].

DNA and RNA synthesis

The genes and RNA transcripts for the L1 ribozyme ligase and the P4–P6 domain the Tetrahymena thermophila Group 1 intron were synthesized and purified as described in Text S1. After transcription, Mg2+ was removed from the RNA by heat annealing in the presence of divalent cation chelating beads (Hampton Research). SHAPE reactivity and ribozyme reactions confirm that the divalent cations are removed by chelating beads.

SHAPE probing of P4–P6 RNA

All manipulations of RNA with Fe2+ were conducted in a Coy chamber with an atmosphere of 85% N2, 10% CO2, 5% H2. P4–P6 domain RNA (11.25 µg) was lyophilized, transferred to the anaerobic chamber, left open for several hours, and resuspended in 240 µL of 50 mM HEPES, pH 8.0, 200 mM sodium acetate (final concentration) that had been previously deoxygenated by bubbling with argon for several hours. The RNA was denatured and renatured using a thermal cycler, by heating to 90°C for 3 min and then quickly cooling to 4°C. Eighty µL of the RNA solution was aliquoted into three tubes. To the first tube, 10 µL of 25 mM FeCl2 (Avantor Performance Materials) solution was added. To the second tube, 10 µL of 25 mM MgCl2 was added. To the third tube, 10 µL water was added. The tubes were incubated at room temperature for 5 minutes. The RNA from each tube was then divided equally between two additional tubes. To one tube of each pair, 5 µL of 800 mM benzoyl cyanide (BzCN) in anhydrous DMSO was added. The other tube of the pair served as a negative control, to which 5 µL neat DMSO was added. The benzoyl cyanide reactions are complete in a few seconds at room temperature [31]. The Fe2+ was removed by treatment with chelating beads. The beads and the associated divalent cations were removed with a 0.2 micron filtration spin column. Denaturing SHAPE experiments were performed in 20 mM HEPES pH 8.0 (final concentration) for 4 min at 90°C using 130 mM N-methylisatoic anhydride (NMIA) in anhydrous DMSO. Modified RNA was purified using RNeasy Mini Kit (Qiagen) and re-suspended in 20 µL 1× TE. The recovery after purification was 65–75%.

A 20-nt long DNA oligomer, 5′- GAACTGCATCCATATCAACA -3′, that anneals to the 3′-end of the P4-P6 domain, was used to prime reverse transcription. The primer was labeled with 6-FAM at its 5′-end (Eurofins MWG Operon). Reverse transcription, capillary electrophoresis and data processing were performed as described [30].

L1 ribozyme ligation reactions

As noted above, all manipulations in which RNA was in contact with Fe2+ were conducted in a Coy chamber. The substrate RNA (5′-UGCACU-3′) labeled with Cy3 at its 5′-end and DNA enhancer (5′-GCGACTGGACATCACGAG-3′) were purchased from Eurofins MWG Operon. Aliquoted reaction components (L1 ligase RNA, substrate RNA and DNA enhancer; typical molar ratio used was 1∶0.1∶2, respectively) were lyophilized separately, transferred to the anaerobic chamber, left open for several hours, and resuspended in 50 mM HEPES, pH 8.0, 200 mM sodium acetate that had been previously deoxygenated by bubbling with argon for several hours. The reaction components were annealed inside the anaerobic chamber by incubating at 90°C for 3 min and cooling to room temperature over 30 min. FeCl2 was weighed and dissolved in water inside the anaerobic chamber. The final volumes of the reaction mixtures were generally 270 µL.

Ligation reactions were initiated by addition of the appropriate cation salt to the dissolved reaction components. At predetermined time points, 30 µL aliquots were withdrawn and quenched by treatment with chelating beads. The beads and divalent cations were removed with a spin column and the samples were frozen and stored at −80°C. The L1 ligase RNA was stable for days in 10 mM Fe2+ in the anaerobic chamber, but degraded quickly upon exposure to atmospheric oxygen. After the Fe2+ was removed with chelating beads, the L1 ligase RNA was stable to exposure to atmospheric oxygen.

For gel analysis, 5 µL of reaction mixture was mixed with 15 µL loading buffer (8 M urea, 1× TTE, 10% glycerol) and denatured by heating to 90°C for 2 min. The reaction components were then resolved on 8% denaturing PAGE gels and visualized on a Typhoon Trio variable mode imager. Some representative gels are shown in Figure S1. The band intensities were quantified using Fujifilm MultiGauge 2.0 software.

Hammerhead ribozyme cleavage reactions

As noted above, all manipulations of the hammerhead RNA in the presence of Fe2+ were carried out in a Coy chamber. The hammerhead ribozyme-substrate was based on the unmodified HHα1 RNA described by Stage-Zimmermann and Uhlenbeck [69]. A 31 nucleotide substrate strand (5′-GGCAAUCGAAACGCGAAAGCGUCUAGCGGGC-3′), labeled at the 3′-end with FAM, and the 21 nucleotide ribozyme strand (5′-CCCGCUACUGAUGAGAUUGCC-3′) were purchased from IDT. Substrate and ribozyme strands (typical molar ratio used was 1∶1000) were lyophilized separately, transferred to the anaerobic chamber, left open for several hours, and resuspended in 50 mM HEPES, pH 7.5 (pH adjusted with KOH). The buffer had previously been deoxygenated by bubbling with argon for several hours. The strands were annealed inside the anaerobic chamber by incubating at 90°C for 2 min and cooling to room temperature over 30 min.

Reactions (150 µL final volume) were initiated by addition of 1.5 µL of cation solution (Fe2+ or Mg2+). At predetermined time points, 20 µL aliquots were withdrawn and quenched by treatment with divalent cation chelating beads. The beads and the associated divalent cations were removed with a spin column, and the samples were frozen and stored at −80°C. For gel analysis, 1 µL of reaction mixture was mixed with 9 µL loading buffer (8 M urea, 1× TTE, 10% glycerol) and denatured by heating to 90°C for 2 minutes. The intact 31 nucleotide substrate and 7 nucleotide product were resolved on 15% denaturing PAGE gels and visualized on a Typhoon Trio variable mode imager, or separated by capillary electrophoresis and quantified as described previously [30].

Supporting Information

[Figure omitted. See PDF.]

Figure S1.

8% polyacryalamide – 8 M urea denaturing gels showing L1 Ribozyme Ligase reaction progress. Only species tagged with 5′-Cy3 dye (substrate and product) are visible. The L1 Ribozyme Ligase is visible when the gel is stained with cyber gold or ethidium. The reaction rate increases when [Fe2+] is increased from 100 µM (LH panel, reaction product observable at 4 hours) to 625 µM (center panel, reaction product observable at first time point, 30 min). The rate of the reaction in 1 mM Mg2+ (RH panel) is roughly equivalent to that in 100 µM Fe2+ (LH panel).

https://doi.org/10.1371/journal.pone.0038024.s001

(TIF)

Table S1.

Electronic energies, interaction energies and the corresponding counterpoise-corrected interaction energies calculated at the (U)B3LYP/6–311++G(d,p) level of theory within the framework of IEFPCM in water.

https://doi.org/10.1371/journal.pone.0038024.s002

(DOCX)

Table S2.

Electronic configurations of Mg2+ and Fe2+ in the RNA2− -Mg2+(H2O)4 and RNA2− -Fe2+(H2O)4 complexes as revealed by the NBO at the (U)B3LYP/6–31G(d,p) level of theory.

https://doi.org/10.1371/journal.pone.0038024.s003

(DOCX)

Text S1.

RNA Synthesis and Purification.

https://doi.org/10.1371/journal.pone.0038024.s004

(DOCX)

Acknowledgments

The authors thank Drs. Clark Johnson, Jeffrey Bada, John Peters, Kent Barefield and Joseph Sadighi for helpful discussions.

Author Contributions

Conceived and designed the experiments: SSA ASP CH DW CDP LL RMW SCH NVH LDW. Performed the experiments: SSA ASP DW CDP LL. Analyzed the data: SSA ASP DW CDP LL JJG NVH SCH RMW LDW. Contributed reagents/materials/analysis tools: JCB EO CRB JJG. Wrote the paper: SSA LDW ASP RMW NVH SCH.

Citation: Athavale SS, Petrov AS, Hsiao C, Watkins D, Prickett CD, Gossett JJ, et al. (2012) RNA Folding and Catalysis Mediated by Iron (II). PLoS ONE7(5): e38024. https://doi.org/10.1371/journal.pone.0038024

References

1. Stein A, Crothers DM (1976) Conformational changes of transfer RNA. The role of magnesium (II) equilibrium binding of magnesium (II) by Escherichia coli tRNAfMet. Biochemistry 15: 160–168.A. SteinDM Crothers1976Conformational changes of transfer RNA. The role of magnesium (II) equilibrium binding of magnesium (II) by Escherichia coli tRNAfMet.Biochemistry15160168

2. Lynch DC, Schimmel PR (1974) Cooperative binding of magnesium to transfer ribonucleic acid studied by a fluorescent probe. Biochemistry 13: 1841–1852.DC LynchPR Schimmel1974Cooperative binding of magnesium to transfer ribonucleic acid studied by a fluorescent probe.Biochemistry1318411852

3. Lindahl T, Adams A, Fresco JR (1966) Renaturation of transfer ribonucleic acids through site binding of magnesium. Proc Natl Acad Sci U S A 55: 941–948.T. LindahlA. AdamsJR Fresco1966Renaturation of transfer ribonucleic acids through site binding of magnesium.Proc Natl Acad Sci U S A55941948

4. Brion P, Westhof E (1997) Hierarchy and dynamics of RNA folding. Annu Rev Biophys Biomol Struct 26: 113–137.P. BrionE. Westhof1997Hierarchy and dynamics of RNA folding.Annu Rev Biophys Biomol Struct26113137

5. Draper DE (2008) RNA folding: Thermodynamic and molecular descriptions of the roles of ions. Biophys J 95: 5489–5495.DE Draper2008RNA folding: Thermodynamic and molecular descriptions of the roles of ions.Biophys J9554895495

6. Auffinger P, Grover N, Westhof E (2011) Metal ion binding to RNA. Met Ions Life Sci 9: 1–35.P. AuffingerN. GroverE. Westhof2011Metal ion binding to RNA.Met Ions Life Sci9135

7. Butcher SE (2011) The spliceosome and its metal ions. Met Ions Life Sci 9: 235–251.SE Butcher2011The spliceosome and its metal ions.Met Ions Life Sci9235251

8. Johnson-Buck AE, McDowell SE, Walter NG (2011) Metal ions: Supporting actors in the playbook of small ribozymes. Met Ions Life Sci 9: 175–196.AE Johnson-BuckSE McDowellNG Walter2011Metal ions: Supporting actors in the playbook of small ribozymes.Met Ions Life Sci9175196

9. Brown ID (1992) Chemical and steric constraints in inorganic solids. Acta Crystallogr Sect B 48: 553–572.ID Brown1992Chemical and steric constraints in inorganic solids.Acta Crystallogr Sect B48553572

10. Rashin AA, Honig B (1985) Reevaluation of the Born model of ion hydration. J Phys Chem 89: 5588–5593.AA RashinB. Honig1985Reevaluation of the Born model of ion hydration.J Phys Chem8955885593

11. Bock CW, Markham GD, Katz AK, Glusker JP (2006) The arrangement of first- and second-shell water molecules around metal ions: Effects of charge and size. Theor Chem Acc 115: 100–112.CW BockGD MarkhamAK KatzJP Glusker2006The arrangement of first- and second-shell water molecules around metal ions: Effects of charge and size.Theor Chem Acc115100112

12. Hsiao C, Tannenbaum M, VanDeusen H, Hershkovitz E, Perng G, Tannenbaum A, Williams LD (2008) Complexes of nucleic acids with group I and II cations. In: Hud N, editor. editor. Nucleic acid metal ion interactions. London: The Royal Society of Chemistry. pp. 1–35.C. HsiaoM. TannenbaumH. VanDeusenE. HershkovitzG. PerngA. TannenbaumLD Williams2008Complexes of nucleic acids with group I and II cations.N. HudeditorNucleic acid metal ion interactions. LondonThe Royal Society of Chemistry135

13. RNA worlds: From life's origins to diversity in gene regulation (2011) Atkins JF, Gesteland RF, Cech TR, editors: Cold Spring Harbor Laboratory Press. regulation RNA worlds: From life's origins to diversity in gene2011Atkins JF, Gesteland RF, Cech TR, editors: Cold Spring Harbor Laboratory Press

14. Anbar AD (2008) Oceans. Elements and evolution. Science 322: 1481–1483.AD Anbar2008Oceans. Elements and evolution.Science32214811483

15. Hazen RM, Ferry JM (2010) Mineral evolution: Mineralogy in the fourth dimension. Elements 6: 9–12.RM HazenJM Ferry2010Mineral evolution: Mineralogy in the fourth dimension.Elements6912

16. Klein C (2005) Some Precambrian banded iron-formations (BIFs) from around the world: Their age, geologic setting, mineralogy, metamorphism, geochemistry, and origin. Am Mineral 90: 1473–1499.C. Klein2005Some Precambrian banded iron-formations (BIFs) from around the world: Their age, geologic setting, mineralogy, metamorphism, geochemistry, and origin.Am Mineral9014731499

17. Johnson CM, Beard BL, Roden EE (2008) The iron isotope fingerprints of redox and biogeochemical cycling in modern and ancient earth. Annual Review of Earth and Planetary Sciences 36: 457–493.CM JohnsonBL BeardEE Roden2008The iron isotope fingerprints of redox and biogeochemical cycling in modern and ancient earth.Annual Review of Earth and Planetary Sciences36457493

18. Petrov AS, Bowman JC, Harvey SC, Williams LD (2011) Bidentate RNA-magnesium clamps: On the origin of the special role of magnesium in RNA folding. RNA 17: 291–297.AS PetrovJC BowmanSC HarveyLD Williams2011Bidentate RNA-magnesium clamps: On the origin of the special role of magnesium in RNA folding.RNA17291297

19. Hsiao C, Williams LD (2009) A recurrent magnesium-binding motif provides a framework for the ribosomal peptidyl transferase center. Nucleic Acids Res 37: 3134–3142.C. HsiaoLD Williams2009A recurrent magnesium-binding motif provides a framework for the ribosomal peptidyl transferase center.Nucleic Acids Res3731343142

20. Ban N, Nissen P, Hansen J, Moore PB, Steitz TA (2000) The complete atomic structure of the large ribosomal subunit at 2.4 Å resolution. Science 289: 905–920.N. BanP. NissenJ. HansenPB MooreTA Steitz2000The complete atomic structure of the large ribosomal subunit at 2.4 Å resolution.Science289905920

21. Cate JH, Hanna RL, Doudna JA (1997) A magnesium ion core at the heart of a ribozyme domain. Nat Struct Biol 4: 553–558.JH CateRL HannaJA Doudna1997A magnesium ion core at the heart of a ribozyme domain.Nat Struct Biol4553558

22. Toor N, Keating KS, Taylor SD, Pyle AM (2008) Crystal structure of a self-spliced group II intron. Science 320: 77–82.N. ToorKS KeatingSD TaylorAM Pyle2008Crystal structure of a self-spliced group II intron.Science3207782

23. Robertson MP, Scott WG (2007) The structural basis of ribozyme-catalyzed RNA assembly. Science 315: 1549–1553.MP RobertsonWG Scott2007The structural basis of ribozyme-catalyzed RNA assembly.Science31515491553

24. Tomasi J, Mennucci B, Cances E (1999) The IEF version of the PCM solvation method: An overview of a new method addressed to study molecular solutes at the QM ab initio level. Journal of Molecular Structure-Theochem 464: 211–226.J. TomasiB. MennucciE. Cances1999The IEF version of the PCM solvation method: An overview of a new method addressed to study molecular solutes at the QM ab initio level.Journal of Molecular Structure-Theochem464211226

25. Glendening ED, Streitwieser A (1994) Natural energy decomposition analysis-an energy partitioning procedure for molecular-interactions with application to weak hydrogen-bonding, strong ionic, and moderate donor-acceptor interactions. J Chem Phys 100: 2900–2909.ED GlendeningA. Streitwieser1994Natural energy decomposition analysis-an energy partitioning procedure for molecular-interactions with application to weak hydrogen-bonding, strong ionic, and moderate donor-acceptor interactions.J Chem Phys10029002909

26. Merino EJ, Wilkinson KA, Coughlan JL, Weeks KM (2005) RNA structure analysis at single nucleotide resolution by selective 2′-hydroxyl acylation and primer extension (SHAPE). J Am Chem Soc 127: 4223–4231.EJ MerinoKA WilkinsonJL CoughlanKM Weeks2005RNA structure analysis at single nucleotide resolution by selective 2′-hydroxyl acylation and primer extension (SHAPE).J Am Chem Soc12742234231

27. Wilkinson KA, Merino EJ, Weeks KM (2005) RNA SHAPE chemistry reveals nonhierarchical interactions dominate equilibrium structural transitions in tRNA(asp) transcripts. J Am Chem Soc 127: 4659–4667.KA WilkinsonEJ MerinoKM Weeks2005RNA SHAPE chemistry reveals nonhierarchical interactions dominate equilibrium structural transitions in tRNA(asp) transcripts.J Am Chem Soc12746594667

28. Mortimer SA, Weeks KM (2007) A fast-acting reagent for accurate analysis of RNA secondary and tertiary structure by SHAPE chemistry. J Am Chem Soc 129: 4144–4145.SA MortimerKM Weeks2007A fast-acting reagent for accurate analysis of RNA secondary and tertiary structure by SHAPE chemistry.J Am Chem Soc12941444145

29. Wilkinson KA, Gorelick RJ, Vasa SM, Guex N, Rein A, Mathews DH, Giddings MC, Weeks KM (2008) High-throughput SHAPE analysis reveals structures in hiv-1 genomic RNA strongly conserved across distinct biological states. PLoS Biol 6: 883–899.KA WilkinsonRJ GorelickSM VasaN. GuexA. ReinDH MathewsMC GiddingsKM Weeks2008High-throughput SHAPE analysis reveals structures in hiv-1 genomic RNA strongly conserved across distinct biological states.PLoS Biol6883899

30. Athavale SS, Gossett JJ, Hsiao C, Bowman JC, O'Neill E, Hershkovitz E, Preeprem T, Hud NV, Wartell RM, Harvey SC, Williams LD (2012) Domain III of the T. thermophilus 23S rRNA folds independently to a near-native state. RNA 18: 752–758.SS AthavaleJJ GossettC. HsiaoJC BowmanE. O'NeillE. HershkovitzT. PreepremNV HudRM WartellSC HarveyLD Williams2012Domain III of the T. thermophilus 23S rRNA folds independently to a near-native state.RNA18752758

31. Mortimer SA, Weeks KM (2008) Time-resolved RNA SHAPE chemistry. J Am Chem Soc 130: 16178–16180.SA MortimerKM Weeks2008Time-resolved RNA SHAPE chemistry.J Am Chem Soc1301617816180

32. Vicens Q, Gooding AR, Laederach A, Cech TR (2007) Local RNA structural changes induced by crystallization are revealed by SHAPE. RNA 13: 536–548.Q. VicensAR GoodingA. LaederachTR Cech2007Local RNA structural changes induced by crystallization are revealed by SHAPE.RNA13536548

33. Seehafer C, Kalweit A, Steger G, Graf S, Hammann C (2011) From alpaca to zebrafish: Hammerhead ribozymes wherever you look. RNA 17: 21–26.C. SeehaferA. KalweitG. StegerS. GrafC. Hammann2011From alpaca to zebrafish: Hammerhead ribozymes wherever you look.RNA172126

34. Scott WG (2007) Ribozymes. Curr Opin Struct Biol 17: 280–286.WG Scott2007Ribozymes.Curr Opin Struct Biol17280286

35. Goss DJ, Theil EC (2011) Iron responsive mRNAs: A family of Fe2+ sensitive riboregulators. Acc Chem Res 44: 1320–1328.DJ GossEC Theil2011Iron responsive mRNAs: A family of Fe2+ sensitive riboregulators.Acc Chem Res4413201328

36. Aguirre JD, Culotta VC (2012) Battles with iron: Manganese in oxidative stress protection. J Biol Chem 287: 13541–13548.JD AguirreVC Culotta2012Battles with iron: Manganese in oxidative stress protection.J Biol Chem2871354113548

37. Wolfe-Simon F, Starovoytov V, Reinfelder JR, Schofield O, Falkowski PG (2006) Localization and role of manganese superoxide dismutase in a marine diatom. Plant physiology 142: 1701–1709.F. Wolfe-SimonV. StarovoytovJR ReinfelderO. SchofieldPG Falkowski2006Localization and role of manganese superoxide dismutase in a marine diatom.Plant physiology14217011709

38. Peers G, Price NM (2004) A role for manganese in superoxide dismutases and growth of iron-deficient diatoms. Limnol Oceanogr 49: 1774–1783.G. PeersNM Price2004A role for manganese in superoxide dismutases and growth of iron-deficient diatoms.Limnol Oceanogr4917741783

39. Whitfield M (2001) Interactions between phytoplankton and trace metals in the ocean. Adv Mar Biol: Academic Press. pp. 1–128.M. Whitfield2001Interactions between phytoplankton and trace metals in the ocean.Adv Mar BiolAcademic Press1128

40. Vary CP, Vournakis JN (1984) RNA structure analysis using methidiumpropyl-EDTA-Fe(II): A base-pair-specific RNA structure probe. Proc Natl Acad Sci U S A 81: 6978–6982.CP VaryJN Vournakis1984RNA structure analysis using methidiumpropyl-EDTA-Fe(II): A base-pair-specific RNA structure probe.Proc Natl Acad Sci U S A8169786982

41. Hanson S, Bauer G, Fink B, Suess B (2005) Molecular analysis of a synthetic tetracycline-binding riboswitch. RNA 11: 503–511.S. HansonG. BauerB. FinkB. Suess2005Molecular analysis of a synthetic tetracycline-binding riboswitch.RNA11503511

42. Wang JF, Cech TR (1992) Tertiary structure around the guanosine-binding site of the tetrahymena ribozyme. Science 256: 526–529.JF WangTR Cech1992Tertiary structure around the guanosine-binding site of the tetrahymena ribozyme.Science256526529

43. Takamoto K, Das R, He Q, Doniach S, Brenowitz M, et al. (2004) Principles of RNA compaction: Insights from the equilibrium folding pathway of the P4–P6 RNA domain in monovalent cations. J Mol Biol 343: 1195–1206.K. TakamotoR. DasQ. HeS. DoniachM. Brenowitz2004Principles of RNA compaction: Insights from the equilibrium folding pathway of the P4–P6 RNA domain in monovalent cations.J Mol Biol34311951206

44. Misra VK, Draper DE (2001) A thermodynamic framework for Mg2+ binding to RNA. Proc Natl Acad Sci U S A 98: 12456–12461.VK MisraDE Draper2001A thermodynamic framework for Mg2+ binding to RNA.Proc Natl Acad Sci U S A981245612461

45. Ward WL, Derose VJ (2012) Ground-state coordination of a catalytic metal to the scissile phosphate of a tertiary-stabilized hammerhead ribozyme. RNA 18: 16–23.WL WardVJ Derose2012Ground-state coordination of a catalytic metal to the scissile phosphate of a tertiary-stabilized hammerhead ribozyme.RNA181623

46. Rosing MT, Bird DK, Sleep NH, Bjerrum CJ (2010) No climate paradox under the faint early sun. Nature 464: 744–747.MT RosingDK BirdNH SleepCJ Bjerrum2010No climate paradox under the faint early sun.Nature464744747

47. Som SM, Catling DC, Harnmeijer JP, Polivka PM, Buick R (2012) Air density 2.7 billion years ago limited to less than twice modern levels by fossil raindrop imprints. Nature 484: 359–362.SM SomDC CatlingJP HarnmeijerPM PolivkaR. Buick2012Air density 2.7 billion years ago limited to less than twice modern levels by fossil raindrop imprints.Nature484359362

48. Holland H (1984) The chemical evolution of the atmosphere and oceans. Princeton, New Jersey: Princeton University Press. 582 p.H. Holland1984The chemical evolution of the atmosphere and oceans.Princeton, New JerseyPrinceton University Press582

49. Moore JK, Braucher O (2007) Observations of dissolved iron concentrations in the world ocean: Implications and constraints for ocean biogeochemical models. Biogeosciences Discuss 4: 1241–1277.JK MooreO. Braucher2007Observations of dissolved iron concentrations in the world ocean: Implications and constraints for ocean biogeochemical models.Biogeosciences Discuss412411277

50. Hazen RM, Papineau D, Bleeker W, Downs RT, Ferry JM, McCoy TJ, Sverjensky DA, Yang H (2008) Mineral evolution. Am Mineral 93: 1693–1720.RM HazenD. PapineauW. BleekerRT DownsJM FerryTJ McCoyDA SverjenskyH. Yang2008Mineral evolution.Am Mineral9316931720

51. Condie KC (2007) The distribution of paleoarchean crust. Van Kranendonk MJ, Smithies RH, Bennett VC, editors. Earth's oldest rocks. Elsevier Publishers. KC Condie2007The distribution of paleoarchean crust.Van Kranendonk MJ, Smithies RH, Bennett VC, editors. Earth's oldest rocks. Elsevier Publishers

52. Tajika E, Matsui T (1992) Evolution of terrestrial proto-CO2 atmosphere coupled with thermal history of the earth. Earth and Planetary Science Letters 113: 251–266.E. TajikaT. Matsui1992Evolution of terrestrial proto-CO2 atmosphere coupled with thermal history of the earth.Earth and Planetary Science Letters113251266

53. Condie KC, Kroner A (2008) When did plate tectonics begin? Evidence from the geologic record. Condie KC, Pease V, editors. Geological society of America special papers. pp. 281–294.KC CondieA. Kroner2008When did plate tectonics begin?Evidence from the geologic record. Condie KC, Pease V, editors. Geological society of America special papers281294

54. Davies GF (1992) On the emergence of plate-tectonics. Geology 20: 963–966.GF Davies1992On the emergence of plate-tectonics.Geology20963966

55. Snow JE, Dick HJB (1995) Pervasive magnesium loss by marine weathering of peridotite. Geochim Cosmochim Acta 59: 4219–4235.JE SnowHJB Dick1995Pervasive magnesium loss by marine weathering of peridotite.Geochim Cosmochim Acta5942194235

56. Sunda WG, Huntsman SA (1995) Iron uptake and growth limitation in oceanic and coastal phytoplankton. Mar Chem 50: 189–206.WG SundaSA Huntsman1995Iron uptake and growth limitation in oceanic and coastal phytoplankton.Mar Chem50189206

57. Coale KH, Johnson KS, Fitzwater SE, Gordon RM, Tanner S, et al. (1996) A massive phytoplankton bloom induced by an ecosystem-scale iron fertilization experiment in the equatorial pacific ocean. Nature 383: 495–501.KH CoaleKS JohnsonSE FitzwaterRM GordonS. Tanner1996A massive phytoplankton bloom induced by an ecosystem-scale iron fertilization experiment in the equatorial pacific ocean.Nature383495501

58. Kell DB (2009) Iron behaving badly: Inappropriate iron chelation as a major contributor to the aetiology of vascular and other progressive inflammatory and degenerative diseases. BMC Med Genomics 2: 2.DB Kell2009Iron behaving badly: Inappropriate iron chelation as a major contributor to the aetiology of vascular and other progressive inflammatory and degenerative diseases.BMC Med Genomics22

59. Chiancone E, Ceci P, Ilari A, Ribacchi F, Stefanini S (2004) Iron and proteins for iron storage and detoxification. BioMetals 17: 197–202.E. ChianconeP. CeciA. IlariF. RibacchiS. Stefanini2004Iron and proteins for iron storage and detoxification.BioMetals17197202

60. Theil EC, Goss DJ (2009) Living with iron (and oxygen): Questions and answers about iron homeostasis. Chem Rev 109: 4568–4579.EC TheilDJ Goss2009Living with iron (and oxygen): Questions and answers about iron homeostasis.Chem Rev10945684579

61. Becke AD (1988) Density-functional exchange-energy approximation with correct asymptotic behavior. Physical Review A 38: 3098–3100.AD Becke1988Density-functional exchange-energy approximation with correct asymptotic behavior.Physical Review A3830983100

62. Lee CT, Yang WT, Parr RG (1988) Development of the Colle-Salvetti correlation energy formula into a functional of the electron density. Physical Review B 37: 785–789.CT LeeWT YangRG Parr1988Development of the Colle-Salvetti correlation energy formula into a functional of the electron density.Physical Review B37785789

63. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, et al. (2009) Gaussian 09, revision A.01. Wallingford, CT: Gaussian, Inc. MJ FrischGW TrucksHB SchlegelGE ScuseriaMA Robb2009Gaussian 09, revision A.01.Wallingford, CTGaussian, Inc

64. Boys SF, Bernardi F (1970) Calculation of small molecular interactions by differences of separate total energies-some procedures with reduced errors. Mol Phys 19: 553–566.SF BoysF. Bernardi1970Calculation of small molecular interactions by differences of separate total energies-some procedures with reduced errors.Mol Phys19553566

65. van Duijneveldt-van de Rijdt JGCM, van Duijneveldt FB (1997) Ab initio methods applied to hydrogen bonded systems. In: Hadzi D, editor. editor. Theoretical treatments of hydrogen bonding. Chichester, UK: Wiley. pp. 13–47.JGCM van Duijneveldt-van de RijdtFB van Duijneveldt1997Ab initio methods applied to hydrogen bonded systems.D. HadzieditorTheoretical treatments of hydrogen bonding. Chichester, UKWiley1347

66. Glendening ED (1996) Natural energy decomposition analysis: Explicit evaluation of electrostatic and polarization effects with application to aqueous clusters of alkali metal cations and neutrals. J Am Chem Soc 118: 2473–2482.ED Glendening1996Natural energy decomposition analysis: Explicit evaluation of electrostatic and polarization effects with application to aqueous clusters of alkali metal cations and neutrals.J Am Chem Soc11824732482

67. Schenter GK, Glendening ED (1996) Natural energy decomposition analysis: The linear response electrical self energy. J Phys Chem 100: 17152–17156.GK SchenterED Glendening1996Natural energy decomposition analysis: The linear response electrical self energy.J Phys Chem1001715217156

68. Gordon MS, Schmidt MW (2005) Advances in electronic structure theory: Gamess a decade later. In: Dykstra CE, Frenking G, Kim KS, Scuseria GE, editors. Theory and applications of computational chemistry: The first forty years. Amsterdam: Elsevier. MS GordonMW Schmidt2005Advances in electronic structure theory: Gamess a decade later.CE DykstraG. FrenkingKS KimGE ScuseriaTheory and applications of computational chemistry: The first forty yearsAmsterdamElsevier

69. Stage-Zimmermann TK, Uhlenbeck OC (2001) A covalent crosslink converts the hammerhead ribozyme from a ribonuclease to an RNA ligase. Nat Struct Biol 8: 863–867.TK Stage-ZimmermannOC Uhlenbeck2001A covalent crosslink converts the hammerhead ribozyme from a ribonuclease to an RNA ligase.Nat Struct Biol8863867