Thermodynamic Potential for the Abiotic Synthesis of Adenine, Cytosine, Guanine, Thymine, Uracil, Ribose, and Deoxyribose in Hydrothermal Systems.

In recent years many theoretical and experimental studies have tested the hypothesis that ancient deep sea hydrothermal systems provided an environment conducive to the abiotic synthesis of biomolecules that are essential for the emergence of life (Bock and Goode 1996; Hazen et al. 2002; Holm 1992; Holm and Andersson 1995, 2005; Holm et al. 2006; McCollom and Seewald 2007; Nisbet and Fowler 1996; Shock 1990, 1992b, 1996; Shock et al. 2000; Simoneit 2004; Woese 1998; Woese et al. 1990). As a result, abiotic synthesis studies involving amino acids (Amend and Shock 1998, 2000; Hennet et al. 1992; Marshall 1994; Shock and Schulte 1990), lipid-like compounds (McCollom et al. 1999; Rushdi and Simoneit 2001), aldehydes (Shock and Schulte 1993), carboxylic acids (McCollom and Seewald 2001, 2003a, b), alcohols, ketones (Shock and Schulte 1995, 1998), polycyclic, alkylated and hydroxylated aromatic hydrocarbons (McCollom 2003; Williams et al. 2005; Zolotov and Shock 1999), thiols (Rushdi and Simoneit 2005; Schulte and Rogers 2004) and other organic compounds (Foustoukos and Seyfried 2004; McCollom and Simoneit 1999; Rushdi and Simoneit 2004; Shock and McKinnon 1993) have been carried out under hydrothermal conditions using experimental and/or computational–thermodynamic techniques. However, the common organic monomers constituting nucleic acids, that is, adenine, cytosine, guanine, cytosine, thymine, uracil, ribose, and deoxyribose, have received far less attention. A notable exception is the recent work carried out by Franiatte and coworkers in which the stability of aqueous adenine was measured at 300°C under controlled fugacities of H₂, CO₂, and N₂ (Franiatte et al. 2008). Exploring potential extraterrestrial sources of nucleobases, Saldino and coworkers have synthesized adenine cytosine, thymine and uracil in one-pot experiments from 0.12 mM formamide (CH₃NO) at 160°C in the presence of cosmic dust analogues (amorphous olivines), TiO₂ and montmorillonites (Saladino et al. 2003, 2004, 2005b) (for a review of experimental abiotic nucleobase synthesis see Saladino et al. 2005a). Similarly, several other experiments, carried out using high concentrations of precursor molecules, organic solvent extractions and a variety of catalysts under unspecified oxidation states, have produced nucleobases in gas, solid and aqueous phases (Ferris et al. 1968; Hayatsu et al. 1968; Hill and Orgel 2002; Miyakawa et al. 2000; Wakamatsu et al. 1966). But because the conditions under which these experiments were performed do not reflect those of any known current or past hydrothermal systems, their relevance to natural systems is unclear. The purpose of the present study is to explore the thermodynamic potential for the abiotic synthesis of the five common nucleobases (adenine, guanine, cytosine, thymine and uracil) and two sugars (ribose & deoxyribose) that make up nucleic acids (DNA & RNA) from two precursor molecules, formaldehyde (CH₂O) and hydrogen cyanide (HCN), over a range of pressures, temperatures, and bulk compositions that are characteristic of hydrothermal systems. This is accomplished by first quantifying the thermodynamic potential to synthesize the precursor molecules from N₂, H₂, CO₂ and CO, and then calculating the energetics of biomolecular synthesis from the two precursor compounds.

The successful experimental synthesis of organic compounds from the condensation of CH₂O and HCN, usually referred to as Strecker synthesis, has lead to the hypothesis that reactions involving these precursor molecules were responsible for the abiotic synthesis of biomolecules on the early Earth (Miller 1957; Ferris et al. 1978; Schulte and Shock 1993, 1995). Oró and coworkers were among the first to contribute to this idea by synthesizing adenine, C₅H₅N₅, from HCN (Oró 1960; Oró and Kimball 1961). In addition, it has long been established that carbohydrates of varying carbon number, such as ribose, C₅H₁₀O₅, can be made abiotically from formaldehyde according to the formose reaction (Butlerov 1861). More recently, Hennet et al. (1992) synthesized several amino acids from CH₂O and HCN at 150°C. The prevalence of CH₂O and HCN in ancient or modern hydrothermal systems is not known, but the amount of these compounds that could have existed in prebiotic hydrothermal systems can be estimated by taking into account reactions among simple sources of N, C, H and O in these environments that can be relatively well-constrained.

Hydrogen Cyanide

HCN in hydrothermal systems could come from reactions between H₂, N₂, and CO₂ and CO (Schulte and Shock 1995; Shock 1992a), all of which are considered to have been common species on the early Earth (Kasting 1993). The activities of aqueous HCN in equilibrium with N₂, H₂ and CO and N₂, H₂ and CO₂ from 0°C to 350°C, at pressures that are either equal to those corresponding to the liquid/vapor saturation curve for H₂O (Psat) or at 500 bars are shown in Fig. 1a,b, respectively. The phase diagram for H₂O shown in Fig. 2 displays the values of pressure used in the Psat calculations as a function of temperature. It can be seen in Fig. 1a,b that the activity of HCN in equilibrium with N₂, H₂ and CO or CO₂ decreases substantially as temperature increases and that pressure has little effect on the magnitude of this variation.

The curves in Fig. 1a,b were generated by first calculating the equilibrium constant, K, for

and

Statements of the law of mass action for these reactions, e.g.,

were then rearranged to solve for the equilibrium activity of the species of interest:

Similarly, the logarithmic equivalent for Reaction (2) can be written:

where K ₁ and K ₂ refer to the equilibrium constants of Reactions (1) and (2), respectively. Values of K for these and the following reactions were calculated using

where R stands for the gas constant, T represents the absolute temperature and ΔG0rΔGr0 denotes the standard molal Gibbs energy of reaction. Values of ΔG0rΔGr0 were computed with the SUPCRT92 software package (Johnson et al. 1992) which relies on the revised Helgeson–Kirkham–Flowers (HKF) equations of state (Shock and Helgeson 1990; Tanger and Helgeson 1988) to calculate the thermodynamic properties of chemical reactions at elevated temperatures and pressures. The standard state adopted in the present study for aqueous species other than H₂O corresponds to unit activity of the species in a hypothetical one molal solution referenced to infinite dilution at any pressure and temperature (see LaRowe and Helgeson 2007). The activity of liquid H₂O was taken to be 1. The activities of N₂, H₂, CO and CO₂ were specified by multiplying the measured concentrations of these species (Table 1) in the presently-active Rainbow hydrothermal system, located at 36°14′ N on the Mid-Atlantic Ridge at a depth of 2,300 m (Charlou and colleagues reported fluids emanating at 365°C at pH=2.8 (Charlou et al. 2002)), with activity coefficients, γ, that were computed using the HCh software package (Shvarov and Bastrakov 1999). Because the ionic strength used in the activity coefficient calculations is that of modern seawater, values of γ for neutral species were generally slightly less than one (∼0.975). This approach was used principally because the composition of ancient hydrothermal systems is unknown, and, at the very least, the modern concentrations represent a proxy for characterizing geochemical solutions that may have prevailed in the past. In addition, this is the rare hydrothermal systems for which concentrations of CO, CO₂, H₂ and N₂ have been reported in the literature. To specify the composition of ancient hydrothermal fluids, Schulte and Shock (1995) followed a different approach that used instead fugacities of CO, CO₂, and N₂ constrained from estimated ancient atmospheric concentrations of the same species made by Kasting (1993). However, a cursory examination of modern hydrothermal vent compositions (Charlou et al. 2002; Von Damm 1995) reveals that the concentrations of CO, CO₂, and N₂ are highly variable from one system to another and do not correlate with their concentrations in the modern atmosphere. The assumption that the concentrations of the species in ancient and modern hydrothermal fluids are similar is reasonable because the composition of hydrothermal fluids are primarily determined by their interactions with mantle-derived rocks, whose composition (i.e., oxidation state) has not changed significantly in the last four billion years (Delano 2001).

 

Table 1 Concentrations of selected species from Rainbow hydrothermal field (Charlou et al. 2002)

Formaldehyde

Formaldehyde, CH₂O, is often cited as a common prebiotic molecule (Orgel 2004; Shapiro 1988), but the concentration of this compound on the early earth is not known. In order to quantify possible concentrations of aqueous CH₂O in hydrothermal systems, as with the HCN, equilibrium activities of this species have been calculated using the H₂, CO and CO₂ concentrations from the Rainbow site (Fig. 1c,d). Although the equilibrium activity of CH₂O follows the same qualitative trend with temperature as the activity of HCN shown in Fig. 1a,b, these calculations show that higher pressure (500 bars) favors higher equilibrium activities of CH₂O relative to lower pressure conditions (Psat). However, even under the most favorable conditions of low temperatures and high pressure, the equilibrium activity of CH₂O, with respect to H₂ and CO and CO₂, is much less than HCN and 10⁻⁵. These results reflect the conditions in one modern hydrothermal system only so if the concentrations of H₂, CO and CO₂ would have been higher in ancient hydrothermal systems, the activity of CH₂O in equilibrium with these reactants could also have been greater. Kasting (1993) estimated that during the first several hundred million years of Earth’s history the atmosphere contained CO + CO₂ equal to 10 bars, a result which suggests that the concentration of precursor carbon molecules in hydrothermal systems could have been different from those today.

The curves in Fig. 1c,d were generated following the strategy outlined above for HCN, but in accordance with

and

The equilibrium activities of HCN and CH₂O shown in Fig. 1 are used in the following section to calculate the thermodynamic drive for the abiotic synthesis of nucleobases, ribose and deoxyribose.

In this section, the thermodynamic drive to synthesize the five common nucleobases, ribose and deoxyribose is quantified by calculating the Gibbs energy of reaction for each of these compounds from the precursor molecules HCN and CH₂O as a function of temperature and temperature.

Nucleobases

The nucleobases considered in this study are the common purines (adenine and guanine) and pyrimidines (cytosine, thymine, and uracil) found in DNA and RNA. Experimental attempts to synthesize these compounds abiotically have been less successful than for amino acids, especially guanine and uracil (Saladino et al. 2005a).

The overall Gibbs energy of reaction, ΔG _r, for the synthesis of the five common nucleobases from aqueous HCN and CH₂O are calculated for

and

and are shown as a function of temperature at Psat and 500 bars in Fig. 3a–e. It can be seen in this figure that the values of ΔG _r are negative for Reactions 9–13 from 0°C to between 175°C and 240°C, depending on the nucleobase. Above these temperatures, Reactions 9–13 are no longer favored. The small difference in values of ΔG _r between the Psat and 500-bar curves indicate that pressure has little effect on these reactions. The dashed lines in Fig. 3a–e represent values of ΔG0rΔGr0 for the respective reactions. These lines are shown to illustrate the quantitative difference between the energetics of chemical reactions when the activities of the reactants and products are (ΔG _r) and are not (ΔG0rΔGr0) taken into account.

Values of the Gibbs energy of reaction were calculated using

where Q refers to the activity product, which is defined by

where the symbols a _i and ν _i denote the activity and stoichiometric reaction coefficient of the ith species, respectively. The standard molal thermodynamic properties and revised HKF equation of state parameters for the nucleobases required to calculate values of K were taken from LaRowe and Helgeson (2006). The values of a _i for HCN and CH₂O were taken from the results shown in Fig. 1b,d. The ΔG _r curves, calculated using CO as a carbon source, are not included here because they are nearly identical to the CO₂-derived calculations. The activities of H₂ required to evaluate Eq. 15 for reactions 10 & 13 were taken from the Rainbow hydrothermal system (Table 1). The activities of the nucleobases were taken to be 10⁻¹², after (McCollom and Amend 2005).

Ribose and Deoxyribose

Values of ΔG _r for the synthesis of aqueous ribose, C₅H₁₀O_{5 (aq)}, and deoxyribose, C₅H₁₀O_{4 (aq)}, from CO₂-derived formaldehyde,

and

are shown as a function of temperature at Psat and 500 bars in Fig. 4a,b, respectively. For reference, the dashed lines in these figures stand for values of ΔG0rΔGr0 for the respective reactions. The overall Gibbs energy of reaction for Reactions 16 and 17 are negative from 0°C to ∼140°C and 0°C to ∼175°C, respectively. The temperatures at which these reactions become endergonic at Psat and 500 bars differ by 5°C or less. These calculations were carried out using Eqs. 14 and 15 and the thermodynamic properties of ribose and deoxyribose (Amend and Plyasunov 2001; LaRowe and Helgeson 2006). The values of a _i for CH₂O required to evaluate Eq. 15 were taken from the equilibrium activities of CO₂-derived formaldehyde shown in Fig. 1d. The activities of ribose and deoxyribose were set equal to 10⁻¹². The values of oxygen fugacity, fO2fO2, used to compute Q for Reaction 17 were determined by calculating the oxygen fugacity that is in equilibrium with the fayalite, magnetite, quartz (FMQ) buffer for the temperatures and pressures considered here, a common proxy for the oxidation state of submarine hydrothermal systems (Schulte and Shock 1995). These values are shown in Fig. 5.

The feasibility of Reaction 16 representing a likely pathway for the abiotic synthesis of ribose has been doubted (Shapiro 1988, 1995). However, due to recent progress in determining the chemical environments that promote the abiotic synthesis of ribose (Orgel 2004; Ferris 2005; Holm et al. 2006), the likelihood that Reaction 16 represents a plausible origin of this fundamental sugar has been substantiated. Initially, a number of studies showed that, under undiscriminating laboratory conditions, the formose reaction produces dozens of aldoses, ketoses, and sugar alcohols from CH₂O, i.e., very little ribose (Shapiro 1988). However, it has recently been found that in the presence of Pb²⁺, aldopentoses are preferentially formed during the formose reaction and some of these reaction products are isomerized to ribose (Zubay 1998; Zubay and Mui 2001). Also, Ricardo et al. (2004) showed that borate minerals preferentially bind pentoses, ribose most so, thus stabilizing and potentially concentrating them over other pentoses. Furthermore, when phosphate is present, ribose-phosphate compounds, the repetitive-unit backbone of RNA, are formed and do not readily convert into other molecules (Müller et al. 1990).

 

Because the concentrations of N₂, H₂, CO, and CO₂ used to calculate the activities of CH₂O and HCN in this study are highly variable in modern systems (Von Damm 1995) and, likely, ancient hydrothermal systems, the activities of the precursor molecules produced in them may have also been quite variable. In order to quantitatively assess the impact that variable precursor molecule concentrations would have on the production of nucleobases, ribose and deoxyribose, the activities of these species in equilibrium with varying activities of CH₂O and HCN and, where relevant, H₂ and O₂, were calculated under temperature and pressure conditions that favor their formation.

The activities of adenine and ribose in equilibrium with varying activities of HCN and CH₂O, respectively, at 150°C and 500 bars are shown in Fig. 6a,b. Activities of adenine between 10⁻²– 10⁻⁶ are in equilibrium with those of HCN equal to from 10⁻⁶ to 10⁻⁷. Similarly, ribose can achieve equilibrium activities between 10⁻²– 10⁻⁶ with CH₂O activities ranging from 10^−4.5 to 10^−5.2. The activities of the other biomolecules considered in this study depend on more than one chemical variable and are thus represented on the plane intersecting this parameter space. For example, Fig. 7a,b show contoured activities of guanine and deoxyribose as a function of the activities of HCN and CH₂O, respectively, and the oxidation state. In the case of guanine, the oxidation state is described by the activity of H₂, aH2aH2, while that for deoxyribose is the fugacity of oxygen. It can be seen in both Fig. 7a,b that large activities of guanine and deoxyribose (10⁻²– 10⁻⁶) are in equilibrium with relatively small activities of their respective precursor molecules (log a HCN = −3.8 to −6.6 in the case of guanine and log a CH₂O = −1 to −9 for deoxyribose). Also, in both cases, for a given activity of precursor molecule, more reducing conditions favor higher activities of guanine and deoxyribose.

Contoured equilibrium activities of cytosine, thymine and uracil are shown as a function of the activities of CH₂O and HCN in Fig. 8a–c. All three of these pyrimidines can coexist at high activities (10⁻²– 10⁻⁶) with relatively low activities of the precursor molecules (10⁻¹⁰ or lower) at 150°C and 500 bars.

Thermodynamic calculations have revealed that adenine, guanine, cytosine, thymine, uracil, ribose, and deoxyribose can be synthesized from the precursor molecules CH₂O and HCN under temperatures, pressures, and fluid compositions that are characteristic of hydrothermal systems. However, nucleobase and sugar synthesis is thermodynamically favored only at the lower end of the temperature range considered here. This finding corroborates previous studies which have suggested that organic synthesis in submarine hydrothermal systems most likely did not occur in black smoker vent sites where the temperatures can exceed 400°C, but more plausibly on the distal, off-axis, portions of hydrothermal systems where the temperature is lower (Shock 1990, 1992a, b). Greater activities of the precursor molecules, CH₂O and HCN, can co-exist in equilibrium with CO, CO₂, H₂, and N₂ as measured in modern hydrothermal systems at cooler temperatures. Activities of HCN corresponding to concentrations in the millimolar range have been calculated to exist at the lower end of the temperature spectrum considered in this study regardless of whether it was generated from CO or CO₂. However, the concentrations of CO, CO₂, H₂, and N₂ from a modern hydrothermal system may not be representative of past geologic conditions. If the concentrations of these building block molecules would be higher, as hypothesized for the early-Earth atmosphere (Kasting 1993), then higher activities of CH₂O and HCN could then coexist in equilibrium with CO, CO₂, H₂, and N₂. It follows that the synthetic potential for the nucleobases, ribose and deoxyribose would also be greater than what is reported in the present study. Therefore, we have explored the range of CH₂O and HCN activities on the equilibrium activities of nucleobase, ribose, and deoxyribose at what can be considered one particular hydrothermal flank condition (150°C and 500 bars). Under these conditions, activities of all of the biomolecules considered here are in equilibrium with the precursor molecules at concentrations comparable to those in modern living organisms (Voet et al. 1999).

Complementing other studies that have shown that various biological and organic molecules such as amino acids (Amend and Shock 2000), carboxylic acids, alcohols, and ketones (Shock and Schulte 1998) can be synthesized in hydrothermal systems, the results of this study support the hypothesis that hydrothermal systems could have served as efficient anabolic reactors for building the molecules that are essential to living organisms. The thermodynamic calculations shown here also provide constraints on the temperature, pressure and bulk composition necessary for the synthesis of these fundamental biomolecules in hydrothermal systems. Our results provide a useful theoretical framework for the design of experiments aimed at synthesizing nucleobases and sugars from inorganic precursors.

This work is supported by the Netherlands Organization for Scientific Research (NWO) grant number 815.01.008.

Open Access

This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.

Affiliations

1. Department of Earth Sciences—Geochemistry, Faculty of Geosciences, Utrecht University, P.O. Box 80.021, TA 3508, Utrecht, Netherlands

   Douglas E. LaRowe & Pierre Regnier

Corresponding author

Correspondence to Douglas E. LaRowe.

Open Access This is an open access article distributed under the terms of the Creative Commons Attribution Noncommercial License (https://creativecommons.org/licenses/by-nc/2.0), which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.

Reprints and Permissions