Article pubs.acs.org/JPCB
Untemplated Nonenzymatic Polymerization of 3′,5′cGMP: A Plausible Route to 3′,5′-Linked Oligonucleotides in Primordia Judit E. Šponer,*,†,‡ Jiří Šponer,†,‡ Alessandra Giorgi,§ Ernesto Di Mauro,*,∥ Samanta Pino,∥ and Giovanna Costanzo⊥ †
Institute of Biophysics, Academy of Sciences of the Czech Republic, Královopolská 135, CZ-61265 Brno, Czech Republic CEITEC - Central European Institute of Technology, Masaryk University, Campus Bohunice, Kamenice 5, CZ-62500 Brno, Czech Republic § Dipartimento di Scienze Biochimiche, “Sapienza” Università di Roma, P.le Aldo Moro, 5, Rome 00185, Italy ∥ Fondazione “Istituto Pasteur-Fondazione Cenci-Bolognetti” c/o Dipartimento di Biologia e Biotecnologie “Charles Darwin”, “Sapienza” Università di Roma, P.le Aldo Moro, 5, Rome 00185, Italy ⊥ Istituto di Biologia e Patologia Molecolari, CNR, P.le Aldo Moro, 5, Rome 00185, Italy ‡
S Supporting Information *
ABSTRACT: The high-energy 3′,5′ phosphodiester linkages conserved in 3′,5′ cyclic GMPs offer a genuine solution for monomer activation required by the transphosphorylation reactions that could lead to the emergence of the first simple oligonucleotide sequences on the early Earth. In this work we provide an in-depth characterization of the effect of the reaction conditions on the yield of the polymerization reaction of 3′,5′ cyclic GMPs both in aqueous environment as well as under dehydrating conditions. We show that the threshold temperature of the polymerization is about 30 °C lower under dehydrating conditions than in solution. In addition, we present a plausible exergonic reaction pathway for the polymerization reaction, which involves transient formation of anionic centers at the O3′ positions of the participating riboses. We suggest that excess Na+ cations inhibit the polymerization reaction because they block the anionic mechanism via neutralizing the negatively charged O3′. Our experimental findings are compatible with a prebiotic scenario, where gradual desiccation of the environment could induce polymerization of 3′,5′ cyclic GMPs synthesized in liquid.
■
INTRODUCTION The experiments presented here are part of an effort aiming to support the hypothesis that extant life originated by Darwinian evolution of a so far poorly characterized RNA world.1,2 The very first origin of RNA is fraught with uncertainty. RNA world might have been preceded by an “unknown-polymer” world,3−5 or it might have started its polymerization destiny from highly activated precursors as phosphoramidated nucleotides whose use has allowed unmatched progress in in vitro molecular evolution studies.6−16 It might even have taken origin outside this planet.17 Each one of these scenarios has its appeals and drawbacks. Alternatively, RNA might have appeared by autopolymerization of simple prebiotically plausible compounds, i.e., 3′,5′ cyclic nucleotides. Polymerization of acyclic nucleotides involves release of one water molecule for each chain extension step, which is thermodynamically unfavorable in an aqueous environment. In contrast, 3′,5′cGMPs possess a 6-membered ring with a preformed intramolecular 3′,5′ phosphodiester linkage. Thus, formally, their polymerization is equivalent with the replacement of these intramolecular linkages with intermolecular ones, without water condensation reactions. In addition, the 6-membered ring of 3′,5′cGMPs is rather strained, i.e. 3′,5′cGMPs are thermodynamically less stable, and more © XXXX American Chemical Society
reactive, than acyclic nucleotides. This makes 3′,5′cGMPs unique substrates for polymerization reactions both from thermodynamic and kinetic points of view. Polymerization of 3′,5′cGMP in water was previously reported18,19 and partially characterized. In particular, it was described that spontaneous polymerization of 3′,5′cGMP occurs in water, in formamide, and in dimethylformamide, and is stimulated (in water) by Brønsted bases such as 1,8diazabicycloundec-7-ene (DBU). The reaction requires neither template, nor enzymatic activities, is thermodynamically favored, and selectively yields 3′,5′-bonded oligoribonucleotides containing as many as 25 nucleotides. The reaction products were analyzed by denaturing PAGE, MALDI ToF MS, 31 P NMR analysis by specific RNases of the polymerized materials. Based on the measured stacking of the 3′,5′ cyclic monomers,19 on the activation of the reaction by Brønsted bases, and on the determination of the molecular species produced, a reaction pathway was proposed19 consisting of a simple process based on the formation of base stackingReceived: January 20, 2015 Revised: January 27, 2015
A
DOI: 10.1021/acs.jpcb.5b00601 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
with 10 L of H2O to remove all residual Na+ ions. Afterward, the cGMP, H+ was applied again to the column to remove all remaining traces of Na+ ions. The resulting product fractions were pooled and filtered through a 0.2 μm membrane. The concentration was adjusted to 1 mM cGMP, H+ with deionized H2O and the solution was frozen at −26 °C. The purity of the final product was 99.67% (HPLC at 253 nm), as analyzed by the Provider. MALDI ToF MS analysis showed the total absence of oligomerized materials. DBU was from SigmaAldrich. Doubly distilled deionized Milli-Q water was used throughout. Enzymes. T4 Polynucleotide Kinase (PNK) (EC 2.7.1.78) from England Biolabs (Ipswich, MA, USA; No. M0201L) catalyzes the transfer and exchange of Pi from the γ-position of ATP to the 5′-OH terminus of polynucleotides, and the removal of the 3′-phosphoryl group from 3′-phosphoryl polynucleotides. One unit is defined as the amount of enzyme catalyzing the production of 1 nmol of phosphate to the 5′-OH end of an oligonucleotide from [γ-32P]ATP in 30 min at 37 °C. Methods. Polymerization of 3′,5′ cGMP. Polymerization of 3′,5′cGMP was performed essentially as described in refs 18, 19, and 23. 3′,5′cGMP Na+-free, H+ form (that was neither evaporated nor precipitated during the preparation steps) was concentrated from the initial 1 mM concentration in Milli-Q water by evaporation in Savant under vacuum and cooling mode until the desired concentration, or dryness, was obtained. Dryness was judged from the formation of a white semisolid aggregate. As analyzed in detail by AFM,24 evaporating RNA under vacuum in these sublimating conditions removes the majority of the water present but does not bring RNA to real dryness. Thus, residual hydration shells around the polymers are likely to remain. For an evaluation of the hydration shells on this semidry RNA see the Discussion in ref 24 and references therein. Polymerization in water was performed18,19,23 by incubating the 3′,5′cGMP solution at the concentration, temperature, and time span indicated where appropriate. The reaction was stopped by alcohol precipitation and analyzed by denaturing PAGE after 32P end-labeling, or by MALDI ToF MS. Polymerization under dehydrating conditions was performed using the modification introduced by Morasch et al. (ref 20), which essentially consists of carrying out the polymerization step in the course of desiccation. In this protocol each sample was prepared separately, starting from the volume of 1 mM 3′,5′ cGMP, H+ corresponding to the wanted final amount. Typically 150 μL of 1 mM solution (1.5 × 10−7 mol) was used. This material was reacted for polymerization in the conditions specified where appropriate. The product of polymerization was analyzed by denaturating PAGE or MALDI ToF. For gel electrophoresis, aliquots of the RNA samples were resuspended in 100% formamide and separated by electrophoresis on 16% polyacrylamide gels containing 7 M urea. DLUs (digital light units) is the measure of the autoradiographic signal on the films and is calibrated by using standards. Taking into account the signal intensity of radiographic spots produced by calibrated amounts of 32P-containing ATP whose specific radioactivity is known a priori, the amount of 32Pcontaining molecules can be calculated. Based on this calibration, 5 fmol of 32P-ATP correspond to 106 DLU. The quantities expressed in DLU in the plots can be transformed into molarities considering that each molecule carries only one 32 P-labeled group. Given the complexity of the population of molecules obtained, DLUs are thus the most convenient units to compare the results obtained in the various conditions.
supported pillared structures, followed by position-stimulated polymerization by transphosphorylation. The major limitations of this reaction are the following: (i) it has low yields in water. It was reported19 that 150 μmol of newly formed oligomers were typically obtained “when starting from 10 mmol of monomer (as determined by calculation of the integrals of signals areas in 31P NMR analysis)”, and as confirmed by calculations based on specific activity/autoradiographic signal intensity in denaturing PAGE analyses. (ii) Another limitation is its apparently erratic reproducibility. (iii) In addition, it was observed19 that the polymerization only occurs in the absence of Na+, but the reasons for the inhibitory activity of this cation were not established. A report of polymerization of 3′,5′cGMP has recently appeared20 showing that the reaction may also occur under dehydrating condition. In this case the product was analyzed by a posteriori intercalation of the fluorescent dye SYBR Gold. The detection of polymerization by this additional and different technique validates the reaction. However, numerous apparent discrepancies remain due to the lower sensitivity of fluorescent labeling, to its nondetection of shorter oligomers, and to the uncomplete exploration of reaction variables. Here we provide an extensive analysis of the reaction variables, the explanation of the reasons/conditions leading to the apparently erratic nature of the polymerization. This analysis has been made possible by the discovery20 that the reaction can be controlled when performed in dry, starting from a material that was devoid of Na+ and was never precipitated out of the solution before reaction.20 In addition, using state-of-the-art quantum chemical calculations, for the first time we present a kinetically and thermodynamically plausible mechanistic model of the reaction, and an explanation on the well-known inhibitory effect of Na+ ions. The 3′,5′cGMP polymerization reaction establishes the principle that nonenzymatic, untemplated polymerization of not-highly preactivated and prebiotically plausible precursors may occur in mild unsophisticated conditions. The prebiotically paramount relevance of spontaneous polymerization of prebiotically plausible21,22 RNA precursors lends interest to this in-depth analysis of an already reported18−20,23 reaction.
■
MATERIALS AND METHODS Materials. Guanosine 3′,5′cyclic monophosphate (3′,5′cGMP) was obtained from BioLog LSI (Bremen, Germany), in acid form (3′,5′cGMP, H+) as 1 mM solution. The compound was custom-made and specially purified in order to guarantee (i) the maximal possible purity relative to the absence of adducts-forming cations (mostly Na+) and (ii) the absence of evaporation or precipitation steps during the course of the whole process. The process by the manufacturer consisted of the following: 735 μmol of cGMP, Na+ (BioLog LSI Catalog No. B 004), MW 367.2 g/mol was dissolved in 490 mL of deionized H2O (1.5 mM) and applied to a cation exchange column pre-equilibrated to the H+-form (5 L of 0.5 M HCI followed by 10 L of H2O until neutral) at a flow rate of 5 mL/min. The strong cation exchanger Toyopearl SP-650 M (TOSOH Bioscience, Stuttgart, Germany) (binding capacity: 0.15 mequiv/mL corresponding to a total of 73500 mequiv) was used. This translated to a 100-fold excess of hydrogen over sodium, if 735 μmol of cGMP, Na+ was applied to the cation exchanger. All product fractions with a concentration higher than 1 mM cGMP, H+ were pooled and stored at +4 °C. The column was regenerated with 5 L of 0.5 M HCI and washed B
DOI: 10.1021/acs.jpcb.5b00601 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
Figure 1. Polymerization of 3′,5′cGMP (H+ form) in water (pH 7.0). Denaturing PAGE electrophoretic analysis. The polymerization was performed on 3′,5′cGMP prepared by concentrating under vacuum the dilute solution (which experienced no precipitation nor concentration after the cationexchange chromatography preparative step) obtained by the Provider. Concentration went from an initial 1 mM (150 μL/sample) to a final 7.5 mM (15 μL). Markers M′ and M′′ are in dry polymerized 3′,5′cGMP (see Figure 2) and partially hydrolyzed 5′32P labeled G24 oligomer,19 respectively. Panels A−C: Analysis of the polymerization products obtained at 50, 75, and 85 °C upon reaction for the indicated times. Panels D, E, and F: Quantitative analysis of the polymerization products (in digital light units DLUs) for selected oligomers, as indicated, detailed as a function of temperature (panel D), time (panel E), and as growth of all the oligomers as a function of time at 85 °C (panel F) (the 6mer is omitted because of fragmentation). Panel G: Total amount of RNA synthesized as a function of time (abscissa) at 85 °C. Five femtomoles of 32P-ATP correspond to 106 DLU. The quantities expressed in DLU in the plots can be transformed into molarities considering that each molecule carries only one 32P-labeled group.
to the neo-polymerized RNA in 20 μL, followed by incubation at 37 °C for 30 min. This procedure typically provides a specific activity of 15 000 cpm/pmol. MALDI-ToF Mass Spectrometry. The product of polymerization (typically in the range of 100 ng in water) was mixed
5′-Terminal Phosphorylation of RNA Oligonucleotides. The standard procedure involves phosphorylation by kinase and [γ-32P] ATP. Phosphorylation was carried out by adding 1 μL of T4 polynucleotide kinase (PNK: 10 U/μL, New England Biolabs), 2 μL of 10 × PNK buffer, and 0.5 μL of [γ-32P]ATP C
DOI: 10.1021/acs.jpcb.5b00601 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
Figure 2. Polymerization of 3′,5′cGMP (H+ form) under dehydrating conditions. Panels A−C. Synthesis at 50, 75, and 85 °C for the indicated times. Panel D: Quantitative analysis of the polymerization products for selected oligomers, as indicated, (DLU) as a function of temperature (abscissa). Reactions run for 2 h. Panel E: Quantitative analysis of the polymerization products for selected oligomers, as indicated, as a function of time (abscissa) and of temperature (as specified). Panel F: Total amount of RNA synthesized as a function of time (abscissa) at 85 °C.
D2, but the resulting geometries did not differ significantly from those obtained with the D2 approximation. Nonetheless, the convergence of the geometry optimization was much slower with the D3 approach. Therefore, the TPSS-D2 method was selected as the best compromise between accuracy and computational expenses for the purposes of the current study. Note that the studied systems contained ca. 200 atoms, so that the QM calculations, mainly the transition state searches and harmonic vibrational analyses, are at the limits of contemporary hardware. Wherever it was possible, we performed geometry optimizations without any structural constraints, i.e. with full relaxation of all parameters, using the Gaussian09 suite of programs.33 In some specific cases (discussed in the Results and Discussion part in detail), however, we had to use geometrical constraints to find intermediates and transition states due to the
(1:1, 1:3, 1:5 ratios) with an aqueous 3-hydroxypicolinic acid matrix solution (20 mg/mL) and analyzed with an AutoFlex II instrument (Bruker Daltonics, Bremen, Germany), equipped with a 337 nm nitrogen laser and operating in reflector positive mode. Quantum Chemical Calculations (QM). Geometries were optimized at the DFT-D (i.e., dispersion-corrected density functional theory) level of theory, using the TPSS density functional25 combined with the D2 empirical dispersion correction introduced by Grimme26 (hereafter abbreviated as TPSS-D2 level). All atoms were described with the TZVP basis set.27,28 Dispersion-corrected TPSS has been shown to be an optimal and balanced method for larger fragments of nucleic acids with a balanced description of backbone conformations and stacking interactions. 29,30 We also performed test calculations using the D3 dispersion correction31,32 instead of D
DOI: 10.1021/acs.jpcb.5b00601 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B flatness of the potential energy surface. The C-PCM continuum solvent approximation34,35 (ε = 78.4) along with the atomic radii of Klamt and Schüürmann36 was used throughout the calculations. Cartesian coordinates of all optimized geometries are listed in the Supporting Information. The models are described in detail at the respective places of the Results and Discussion part. Free energies of the studied compounds (G) were calculated from the total electronic energy (Etot) and from the thermal and entropic correction terms to the Gibbs free energy (δG) via harmonic approximation from frequency calculations at 298 K:
■
G = Etot + δG
The yield of the polymerization was calculated for 3′,5′cGMP reacted in water at pH 7.0 without the addition of Brønsted bases, at 85 °C for 18 h at 15 mM concentration. This reaction transformed the initial 60 nmol in 9.6 pmol of oligomers which, corrected for the Navg length of the products (=8 nt) gives a yield of 0.13%. Navg was calculated as described in ref 18. The yields of the reactions vary extensively depending on the conditions, as described in the following sections. The catalytic constant was determined by standard graphical methods to be (at 15 mM and 85 °C in water) 3.2 × 10−11 M h−1. The optimum pH for the polymerization was previously determined18,19 to be 9.0, in agreement with the observed stimulation by Brønsted bases19 and with the polymerization occurring in formamide.19 Polymerization under Dehydrating Conditions. The polymerization under dehydrating conditions was analyzed by drying under vacuum 150 μL/sample of the starting 1 mM 3′,5′cGMP solution. Samples were treated for the indicated times (from 0 to 24 h) at the indicated temperatures, then analyzed. Figure 2 shows polymerization at 50, 75, and 85 °C (panels A to C) for time periods up to 24 h. Data were also collected at 37 and 60 °C (see panel D and Figure 3). Panels D to F of Figure 2 describe the polymerization products as follows: Panel D shows a quantitative analysis of the polymerization products as a function of temperature, given an amount (in DLU) of selected representative oligomers synthesized at the time point 2 h. After 2 h the kinetics at the higher temperatures is essentially plateauing (data plot not shown), as is evident from visual inspection of the gel patterns in panels A−C. The plateauing effect observed at high temperature, both as amount and as fragment length, is not due to temperature-dependent degradation (which is marginally observed in water, see Figure 1) because degradation would result in double bands caused by the two-step cleavage of the phosphodiester bonds.37−39 The mechanism of RNA hydrolysis is fully characterized.37−39 Hydrolysis of the phosphodiester linkage involves participation of the 2′-OH group as an internal nucleophile38 in a two-step process catalyzed by protons, hydroxide, nitrogen derivatives, and metal ions. The first step is a transesterification reaction that leads to the formation of a 2′,3′ cyclic monophosphate, which undergoes a hydrolysis in the second step yielding a mixture of 3′- and 2′-phosphate monoesters. In the electrophoretic images this two-step mechanism manifests itself as a double-banded profile.38 The absence of these double bands in RNA polymerized under dehydrating conditions shows that lower signal directly means lower synthesis. Panel E of Figure 2 shows a plot of selected representative oligomers (5-, 9-, 12mers), given as a function of time at different temperatures. Inspection shows that a critical transition temperature exists resulting in a time lag at 50 °C (arrowed). At or below 50 °C synthesis is largely limited to fragments
Untemplated Nonenzymatic Polymerization of 3′,5′cGMP: A Plausible Route to 3′,5′-Linked Oligonucleotides in Primordia Judit E. Šponer,*,†,‡ Jiří Šponer,†,‡ Alessandra Giorgi,§ Ernesto Di Mauro,*,∥ Samanta Pino,∥ and Giovanna Costanzo⊥ †
Institute of Biophysics, Academy of Sciences of the Czech Republic, Královopolská 135, CZ-61265 Brno, Czech Republic CEITEC - Central European Institute of Technology, Masaryk University, Campus Bohunice, Kamenice 5, CZ-62500 Brno, Czech Republic § Dipartimento di Scienze Biochimiche, “Sapienza” Università di Roma, P.le Aldo Moro, 5, Rome 00185, Italy ∥ Fondazione “Istituto Pasteur-Fondazione Cenci-Bolognetti” c/o Dipartimento di Biologia e Biotecnologie “Charles Darwin”, “Sapienza” Università di Roma, P.le Aldo Moro, 5, Rome 00185, Italy ⊥ Istituto di Biologia e Patologia Molecolari, CNR, P.le Aldo Moro, 5, Rome 00185, Italy ‡
S Supporting Information *
ABSTRACT: The high-energy 3′,5′ phosphodiester linkages conserved in 3′,5′ cyclic GMPs offer a genuine solution for monomer activation required by the transphosphorylation reactions that could lead to the emergence of the first simple oligonucleotide sequences on the early Earth. In this work we provide an in-depth characterization of the effect of the reaction conditions on the yield of the polymerization reaction of 3′,5′ cyclic GMPs both in aqueous environment as well as under dehydrating conditions. We show that the threshold temperature of the polymerization is about 30 °C lower under dehydrating conditions than in solution. In addition, we present a plausible exergonic reaction pathway for the polymerization reaction, which involves transient formation of anionic centers at the O3′ positions of the participating riboses. We suggest that excess Na+ cations inhibit the polymerization reaction because they block the anionic mechanism via neutralizing the negatively charged O3′. Our experimental findings are compatible with a prebiotic scenario, where gradual desiccation of the environment could induce polymerization of 3′,5′ cyclic GMPs synthesized in liquid.
■
INTRODUCTION The experiments presented here are part of an effort aiming to support the hypothesis that extant life originated by Darwinian evolution of a so far poorly characterized RNA world.1,2 The very first origin of RNA is fraught with uncertainty. RNA world might have been preceded by an “unknown-polymer” world,3−5 or it might have started its polymerization destiny from highly activated precursors as phosphoramidated nucleotides whose use has allowed unmatched progress in in vitro molecular evolution studies.6−16 It might even have taken origin outside this planet.17 Each one of these scenarios has its appeals and drawbacks. Alternatively, RNA might have appeared by autopolymerization of simple prebiotically plausible compounds, i.e., 3′,5′ cyclic nucleotides. Polymerization of acyclic nucleotides involves release of one water molecule for each chain extension step, which is thermodynamically unfavorable in an aqueous environment. In contrast, 3′,5′cGMPs possess a 6-membered ring with a preformed intramolecular 3′,5′ phosphodiester linkage. Thus, formally, their polymerization is equivalent with the replacement of these intramolecular linkages with intermolecular ones, without water condensation reactions. In addition, the 6-membered ring of 3′,5′cGMPs is rather strained, i.e. 3′,5′cGMPs are thermodynamically less stable, and more © XXXX American Chemical Society
reactive, than acyclic nucleotides. This makes 3′,5′cGMPs unique substrates for polymerization reactions both from thermodynamic and kinetic points of view. Polymerization of 3′,5′cGMP in water was previously reported18,19 and partially characterized. In particular, it was described that spontaneous polymerization of 3′,5′cGMP occurs in water, in formamide, and in dimethylformamide, and is stimulated (in water) by Brønsted bases such as 1,8diazabicycloundec-7-ene (DBU). The reaction requires neither template, nor enzymatic activities, is thermodynamically favored, and selectively yields 3′,5′-bonded oligoribonucleotides containing as many as 25 nucleotides. The reaction products were analyzed by denaturing PAGE, MALDI ToF MS, 31 P NMR analysis by specific RNases of the polymerized materials. Based on the measured stacking of the 3′,5′ cyclic monomers,19 on the activation of the reaction by Brønsted bases, and on the determination of the molecular species produced, a reaction pathway was proposed19 consisting of a simple process based on the formation of base stackingReceived: January 20, 2015 Revised: January 27, 2015
A
DOI: 10.1021/acs.jpcb.5b00601 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
with 10 L of H2O to remove all residual Na+ ions. Afterward, the cGMP, H+ was applied again to the column to remove all remaining traces of Na+ ions. The resulting product fractions were pooled and filtered through a 0.2 μm membrane. The concentration was adjusted to 1 mM cGMP, H+ with deionized H2O and the solution was frozen at −26 °C. The purity of the final product was 99.67% (HPLC at 253 nm), as analyzed by the Provider. MALDI ToF MS analysis showed the total absence of oligomerized materials. DBU was from SigmaAldrich. Doubly distilled deionized Milli-Q water was used throughout. Enzymes. T4 Polynucleotide Kinase (PNK) (EC 2.7.1.78) from England Biolabs (Ipswich, MA, USA; No. M0201L) catalyzes the transfer and exchange of Pi from the γ-position of ATP to the 5′-OH terminus of polynucleotides, and the removal of the 3′-phosphoryl group from 3′-phosphoryl polynucleotides. One unit is defined as the amount of enzyme catalyzing the production of 1 nmol of phosphate to the 5′-OH end of an oligonucleotide from [γ-32P]ATP in 30 min at 37 °C. Methods. Polymerization of 3′,5′ cGMP. Polymerization of 3′,5′cGMP was performed essentially as described in refs 18, 19, and 23. 3′,5′cGMP Na+-free, H+ form (that was neither evaporated nor precipitated during the preparation steps) was concentrated from the initial 1 mM concentration in Milli-Q water by evaporation in Savant under vacuum and cooling mode until the desired concentration, or dryness, was obtained. Dryness was judged from the formation of a white semisolid aggregate. As analyzed in detail by AFM,24 evaporating RNA under vacuum in these sublimating conditions removes the majority of the water present but does not bring RNA to real dryness. Thus, residual hydration shells around the polymers are likely to remain. For an evaluation of the hydration shells on this semidry RNA see the Discussion in ref 24 and references therein. Polymerization in water was performed18,19,23 by incubating the 3′,5′cGMP solution at the concentration, temperature, and time span indicated where appropriate. The reaction was stopped by alcohol precipitation and analyzed by denaturing PAGE after 32P end-labeling, or by MALDI ToF MS. Polymerization under dehydrating conditions was performed using the modification introduced by Morasch et al. (ref 20), which essentially consists of carrying out the polymerization step in the course of desiccation. In this protocol each sample was prepared separately, starting from the volume of 1 mM 3′,5′ cGMP, H+ corresponding to the wanted final amount. Typically 150 μL of 1 mM solution (1.5 × 10−7 mol) was used. This material was reacted for polymerization in the conditions specified where appropriate. The product of polymerization was analyzed by denaturating PAGE or MALDI ToF. For gel electrophoresis, aliquots of the RNA samples were resuspended in 100% formamide and separated by electrophoresis on 16% polyacrylamide gels containing 7 M urea. DLUs (digital light units) is the measure of the autoradiographic signal on the films and is calibrated by using standards. Taking into account the signal intensity of radiographic spots produced by calibrated amounts of 32P-containing ATP whose specific radioactivity is known a priori, the amount of 32Pcontaining molecules can be calculated. Based on this calibration, 5 fmol of 32P-ATP correspond to 106 DLU. The quantities expressed in DLU in the plots can be transformed into molarities considering that each molecule carries only one 32 P-labeled group. Given the complexity of the population of molecules obtained, DLUs are thus the most convenient units to compare the results obtained in the various conditions.
supported pillared structures, followed by position-stimulated polymerization by transphosphorylation. The major limitations of this reaction are the following: (i) it has low yields in water. It was reported19 that 150 μmol of newly formed oligomers were typically obtained “when starting from 10 mmol of monomer (as determined by calculation of the integrals of signals areas in 31P NMR analysis)”, and as confirmed by calculations based on specific activity/autoradiographic signal intensity in denaturing PAGE analyses. (ii) Another limitation is its apparently erratic reproducibility. (iii) In addition, it was observed19 that the polymerization only occurs in the absence of Na+, but the reasons for the inhibitory activity of this cation were not established. A report of polymerization of 3′,5′cGMP has recently appeared20 showing that the reaction may also occur under dehydrating condition. In this case the product was analyzed by a posteriori intercalation of the fluorescent dye SYBR Gold. The detection of polymerization by this additional and different technique validates the reaction. However, numerous apparent discrepancies remain due to the lower sensitivity of fluorescent labeling, to its nondetection of shorter oligomers, and to the uncomplete exploration of reaction variables. Here we provide an extensive analysis of the reaction variables, the explanation of the reasons/conditions leading to the apparently erratic nature of the polymerization. This analysis has been made possible by the discovery20 that the reaction can be controlled when performed in dry, starting from a material that was devoid of Na+ and was never precipitated out of the solution before reaction.20 In addition, using state-of-the-art quantum chemical calculations, for the first time we present a kinetically and thermodynamically plausible mechanistic model of the reaction, and an explanation on the well-known inhibitory effect of Na+ ions. The 3′,5′cGMP polymerization reaction establishes the principle that nonenzymatic, untemplated polymerization of not-highly preactivated and prebiotically plausible precursors may occur in mild unsophisticated conditions. The prebiotically paramount relevance of spontaneous polymerization of prebiotically plausible21,22 RNA precursors lends interest to this in-depth analysis of an already reported18−20,23 reaction.
■
MATERIALS AND METHODS Materials. Guanosine 3′,5′cyclic monophosphate (3′,5′cGMP) was obtained from BioLog LSI (Bremen, Germany), in acid form (3′,5′cGMP, H+) as 1 mM solution. The compound was custom-made and specially purified in order to guarantee (i) the maximal possible purity relative to the absence of adducts-forming cations (mostly Na+) and (ii) the absence of evaporation or precipitation steps during the course of the whole process. The process by the manufacturer consisted of the following: 735 μmol of cGMP, Na+ (BioLog LSI Catalog No. B 004), MW 367.2 g/mol was dissolved in 490 mL of deionized H2O (1.5 mM) and applied to a cation exchange column pre-equilibrated to the H+-form (5 L of 0.5 M HCI followed by 10 L of H2O until neutral) at a flow rate of 5 mL/min. The strong cation exchanger Toyopearl SP-650 M (TOSOH Bioscience, Stuttgart, Germany) (binding capacity: 0.15 mequiv/mL corresponding to a total of 73500 mequiv) was used. This translated to a 100-fold excess of hydrogen over sodium, if 735 μmol of cGMP, Na+ was applied to the cation exchanger. All product fractions with a concentration higher than 1 mM cGMP, H+ were pooled and stored at +4 °C. The column was regenerated with 5 L of 0.5 M HCI and washed B
DOI: 10.1021/acs.jpcb.5b00601 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
Figure 1. Polymerization of 3′,5′cGMP (H+ form) in water (pH 7.0). Denaturing PAGE electrophoretic analysis. The polymerization was performed on 3′,5′cGMP prepared by concentrating under vacuum the dilute solution (which experienced no precipitation nor concentration after the cationexchange chromatography preparative step) obtained by the Provider. Concentration went from an initial 1 mM (150 μL/sample) to a final 7.5 mM (15 μL). Markers M′ and M′′ are in dry polymerized 3′,5′cGMP (see Figure 2) and partially hydrolyzed 5′32P labeled G24 oligomer,19 respectively. Panels A−C: Analysis of the polymerization products obtained at 50, 75, and 85 °C upon reaction for the indicated times. Panels D, E, and F: Quantitative analysis of the polymerization products (in digital light units DLUs) for selected oligomers, as indicated, detailed as a function of temperature (panel D), time (panel E), and as growth of all the oligomers as a function of time at 85 °C (panel F) (the 6mer is omitted because of fragmentation). Panel G: Total amount of RNA synthesized as a function of time (abscissa) at 85 °C. Five femtomoles of 32P-ATP correspond to 106 DLU. The quantities expressed in DLU in the plots can be transformed into molarities considering that each molecule carries only one 32P-labeled group.
to the neo-polymerized RNA in 20 μL, followed by incubation at 37 °C for 30 min. This procedure typically provides a specific activity of 15 000 cpm/pmol. MALDI-ToF Mass Spectrometry. The product of polymerization (typically in the range of 100 ng in water) was mixed
5′-Terminal Phosphorylation of RNA Oligonucleotides. The standard procedure involves phosphorylation by kinase and [γ-32P] ATP. Phosphorylation was carried out by adding 1 μL of T4 polynucleotide kinase (PNK: 10 U/μL, New England Biolabs), 2 μL of 10 × PNK buffer, and 0.5 μL of [γ-32P]ATP C
DOI: 10.1021/acs.jpcb.5b00601 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B
Figure 2. Polymerization of 3′,5′cGMP (H+ form) under dehydrating conditions. Panels A−C. Synthesis at 50, 75, and 85 °C for the indicated times. Panel D: Quantitative analysis of the polymerization products for selected oligomers, as indicated, (DLU) as a function of temperature (abscissa). Reactions run for 2 h. Panel E: Quantitative analysis of the polymerization products for selected oligomers, as indicated, as a function of time (abscissa) and of temperature (as specified). Panel F: Total amount of RNA synthesized as a function of time (abscissa) at 85 °C.
D2, but the resulting geometries did not differ significantly from those obtained with the D2 approximation. Nonetheless, the convergence of the geometry optimization was much slower with the D3 approach. Therefore, the TPSS-D2 method was selected as the best compromise between accuracy and computational expenses for the purposes of the current study. Note that the studied systems contained ca. 200 atoms, so that the QM calculations, mainly the transition state searches and harmonic vibrational analyses, are at the limits of contemporary hardware. Wherever it was possible, we performed geometry optimizations without any structural constraints, i.e. with full relaxation of all parameters, using the Gaussian09 suite of programs.33 In some specific cases (discussed in the Results and Discussion part in detail), however, we had to use geometrical constraints to find intermediates and transition states due to the
(1:1, 1:3, 1:5 ratios) with an aqueous 3-hydroxypicolinic acid matrix solution (20 mg/mL) and analyzed with an AutoFlex II instrument (Bruker Daltonics, Bremen, Germany), equipped with a 337 nm nitrogen laser and operating in reflector positive mode. Quantum Chemical Calculations (QM). Geometries were optimized at the DFT-D (i.e., dispersion-corrected density functional theory) level of theory, using the TPSS density functional25 combined with the D2 empirical dispersion correction introduced by Grimme26 (hereafter abbreviated as TPSS-D2 level). All atoms were described with the TZVP basis set.27,28 Dispersion-corrected TPSS has been shown to be an optimal and balanced method for larger fragments of nucleic acids with a balanced description of backbone conformations and stacking interactions. 29,30 We also performed test calculations using the D3 dispersion correction31,32 instead of D
DOI: 10.1021/acs.jpcb.5b00601 J. Phys. Chem. B XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry B flatness of the potential energy surface. The C-PCM continuum solvent approximation34,35 (ε = 78.4) along with the atomic radii of Klamt and Schüürmann36 was used throughout the calculations. Cartesian coordinates of all optimized geometries are listed in the Supporting Information. The models are described in detail at the respective places of the Results and Discussion part. Free energies of the studied compounds (G) were calculated from the total electronic energy (Etot) and from the thermal and entropic correction terms to the Gibbs free energy (δG) via harmonic approximation from frequency calculations at 298 K:
■
G = Etot + δG
The yield of the polymerization was calculated for 3′,5′cGMP reacted in water at pH 7.0 without the addition of Brønsted bases, at 85 °C for 18 h at 15 mM concentration. This reaction transformed the initial 60 nmol in 9.6 pmol of oligomers which, corrected for the Navg length of the products (=8 nt) gives a yield of 0.13%. Navg was calculated as described in ref 18. The yields of the reactions vary extensively depending on the conditions, as described in the following sections. The catalytic constant was determined by standard graphical methods to be (at 15 mM and 85 °C in water) 3.2 × 10−11 M h−1. The optimum pH for the polymerization was previously determined18,19 to be 9.0, in agreement with the observed stimulation by Brønsted bases19 and with the polymerization occurring in formamide.19 Polymerization under Dehydrating Conditions. The polymerization under dehydrating conditions was analyzed by drying under vacuum 150 μL/sample of the starting 1 mM 3′,5′cGMP solution. Samples were treated for the indicated times (from 0 to 24 h) at the indicated temperatures, then analyzed. Figure 2 shows polymerization at 50, 75, and 85 °C (panels A to C) for time periods up to 24 h. Data were also collected at 37 and 60 °C (see panel D and Figure 3). Panels D to F of Figure 2 describe the polymerization products as follows: Panel D shows a quantitative analysis of the polymerization products as a function of temperature, given an amount (in DLU) of selected representative oligomers synthesized at the time point 2 h. After 2 h the kinetics at the higher temperatures is essentially plateauing (data plot not shown), as is evident from visual inspection of the gel patterns in panels A−C. The plateauing effect observed at high temperature, both as amount and as fragment length, is not due to temperature-dependent degradation (which is marginally observed in water, see Figure 1) because degradation would result in double bands caused by the two-step cleavage of the phosphodiester bonds.37−39 The mechanism of RNA hydrolysis is fully characterized.37−39 Hydrolysis of the phosphodiester linkage involves participation of the 2′-OH group as an internal nucleophile38 in a two-step process catalyzed by protons, hydroxide, nitrogen derivatives, and metal ions. The first step is a transesterification reaction that leads to the formation of a 2′,3′ cyclic monophosphate, which undergoes a hydrolysis in the second step yielding a mixture of 3′- and 2′-phosphate monoesters. In the electrophoretic images this two-step mechanism manifests itself as a double-banded profile.38 The absence of these double bands in RNA polymerized under dehydrating conditions shows that lower signal directly means lower synthesis. Panel E of Figure 2 shows a plot of selected representative oligomers (5-, 9-, 12mers), given as a function of time at different temperatures. Inspection shows that a critical transition temperature exists resulting in a time lag at 50 °C (arrowed). At or below 50 °C synthesis is largely limited to fragments
