In search of the RNA world on Mars

Abstract Advances in origins of life research and prebiotic chemistry suggest that life as we know it may have emerged from an earlier RNA World. However, it has been difficult to reconcile the conditions used in laboratory experiments with real‐world geochemical environments that may have existed on the early Earth and hosted the origin(s) of life. This challenge is due to geologic resurfacing and recycling that have erased the overwhelming majority of the Earth's prebiotic history. We therefore propose that Mars, a planet frozen in time, comprised of many surfaces that have remained relatively unchanged since their formation > 4 Gya, is the best alternative to search for environments consistent with geochemical requirements imposed by the RNA world. In this study, we synthesize in situ and orbital observations of Mars and modeling of its early atmosphere into solutions containing a range of pHs and concentrations of prebiotically relevant metals (Fe2+, Mg2+, and Mn2+) spanning various candidate aqueous environments. We then experimentally determine RNA degradation kinetics due to metal‐catalyzed hydrolysis (cleavage) and evaluate whether early Mars could have been permissive toward the accumulation of long‐lived RNA polymers. Our results indicate that a Mg2+‐rich basalt sourcing metals to a slightly acidic (pH 5.4) environment mediates the slowest rates of RNA cleavage, though geologic evidence and basalt weathering models suggest aquifers on Mars would be near neutral (pH ~ 7). Moreover, the early onset of oxidizing conditions on Mars has major consequences regarding the availability of oxygen‐sensitive metals (i.e., Fe2+ and Mn2+) due to increased RNA degradation rates and precipitation. Overall, (a) low pH decreases RNA cleavage at high metal concentrations; (b) acidic to neutral pH environments with Fe2+ or Mn2+ cleave more RNA than Mg2+; and (c) alkaline environments with Mg2+ dramatically cleaves more RNA while precipitates were observed for Fe2+ and Mn2+.

. Fundamentally, neither DNA, RNA, nor proteins can exist without the others as they do today. Nevertheless, this dilemma belies the fact that the capability of translating information between dissimilar polymers (e.g., polynucleotides to polypeptides) is mediated by the ribosome, an RNA enzyme (Cech, 2000). This is significant because the ribosome is arguably an evolutionary anachronism from a period where RNA polymers acted as both enzymes (protein) and information storage (DNA) (Petrov et al., 2015). Additional discoveries of structural and regulatory RNA molecules (Breaker, 2012) suggest that life may have emerged from an earlier RNA world dominated by ribozymes (e.g., the ribosome) (Gilbert, 1986) and ribonucleotide-containing molecules (e.g., adenosine triphosphate-ATP) (Hernández-Morales et al., 2019) catalyzing reactions and mediating a protometabolism.
Although strides in prebiotic chemistry have demonstrated the viability of an origin of life via the RNA world, a long-standing criticism is that RNA is inherently unstable due to the presence of a nucleophilic 2'-hydroxyl group which readily catalyzes cleavage of the 5',3'-phosphodiester bond (Li & Breaker, 1999). Because of this characteristic, RNA is deemed an ephemeral molecule that is unlikely to accumulate, functionalize, and precipitate life in a prebiotic world.
Researchers have therefore directed efforts toward determining particular conditions or cofactors which can stabilize RNA in real-world environments. Experimental work suggests the following: (a) RNA is most chemically stable between pH 4 -5 (Bernhardt & Tate, 2012;Oivanen et al., 1998) and near 0°C (Kua & Bada, 2011;Levy & Miller, 1998);(b) metal cofactors such as Fe 2+ , Mg 2+ , and Mn 2+ facilitate the folding of RNA polymers into stable secondary and tertiary structures Laing et al., 1994;Petrov et al., 2012); (c) copolymers such as polypeptides and polysaccharides can favor specific polynucleotide conformations, resulting in persistent structures and vice versa Runnels et al., 2018); (d) mutually stabilizing peptide-RNA conformations rapidly denature above 45°C ; and (e) folding of many RNA sequences decreases rapidly above 30°C (Moulton et al., 2000). Still, it has been difficult to confidently determine the dynamic environments that could have existed on the Hadean Earth and hosted the origin of life. This challenge is in part due to geologic resurfacing and recycling that have erased the overwhelming majority of the Earth's prebiotic history (Marchi et al., 2014). Nevertheless, we can speculate that the likeliest time interval for the origin of life on Earth can be constrained by the accretion of the first continents following a Moon-forming impact (Monteux et al., 2016) or possible late veneer impactor (Genda et al., 2017) ~4.5-4.4 Gya and the earliest unambiguous biological structures at ~ 3.5 Gya (Allwood et al., ,,,2006(Allwood et al., ,,, , 2009. Given this consideration, the best alternative is to search for environments consistent with RNA stability on Mars, a planet frozen in time, preserving primordial surfaces which have remained relatively unchanged since they formed > 4 Gya (Hartmann & Neukum, 2001).
Perhaps a window into the Hadean on Earth, the Noachian on Mars is characterized by meteoritic bombardment and punctuated aqueous activity resulting in extensive groundwater circulation (Ehlmann et al., 2011), valley networks (Fassett & Head, 2011), and long-lived lacustrine environments (Goudge et al., 2012(Goudge et al., , 2015Grotzinger et al. 2014). Some researchers would argue that life began on Mars and was transported to the Earth around the timing of the earliest putative biosignatures ~ 3.8 Gya found in metasedimentary rocks (Alleon & Summons, 2019;Benner & Kim, 2015;Hassenkam et al., 2017;Tashiro et al., 2017). That is, non-sterilizing lithological exchange between Mars and Earth from impact ejecta produced during the presumed Late-Heavy Bombardment period (Boehnke & Harrison, 2016;Gomes et al., 2005) may have transported viable microbes between planets resulting in ancestrally related life (Gladman et al., 1996;Weiss, 2000).
The case for an origin of life on Mars relies on prebiotic environments that are inferred to be analogous to early environments on Earth (Sasselov et al., 2020), common molecular feedstocks (including cometary sources) (Callahan et al., 2011), and plausible reactive pathways predicted on Earth that are applicable on Mars (e.g., Ritson et al., 2018) which may have resulted in parallel events in accordance with the RNA world hypothesis (Benner & Kim, 2015). This notion is further supported by in situ detection of boron (Gasda et al., 2017), which is considered crucial to stabilize ribose in the formose reaction (Furukawa & Kakegawa, 2017), experimental work that predicts higher phosphate bioavailability on Mars (Adcock et al., 2013), and the detection of clays (Ehlmann et al., 2011) that have been demonstrated to assist in non-enzymatic RNA polymerization (Ferris, 2006). This Mars origin of life hypothesis suggests that past or present Martian life may have utilized known building blocks (e.g., nucleic acids, sugars, amino acids) and closely resembled life as we know it. Moreover, if life exists on Mars today, it could theoretically be detected by means of nucleic acid (DNA and RNA) sequencing (Carr et al. 2017;Mojarro et al., 2019). Assuming that viable RNA was being delivered to Mars via unspecified sources (e.g., cometary or in situ synthesis) to UVshielded aqueous environments (Cockell et al., 2000), here we investigate whether early Mars was permissive toward the accumulation of long-lived RNA polymers. We anticipate our findings could provide insight into potential mechanisms, environments, and requirements necessary for sustaining an RNA world on the early Earth.

| Approach
The surface of Mars displays evidence for alternating climate regimes at regional-to-global magnitudes that have evolved on variable time scales not dissimilar to the Earth (McLennan et al., 2019). In general, early Mars contained a broad range of geochemical environments (e.g., acidic to alkaline) primarily influenced by redox chemistry. In this study, we synthesize in situ and orbital observations and modeling of the early Martian atmosphere in order to extrapolate representative solutions containing a range of pHs and metals analogous to various candidate aqueous environments on Mars. Below we detail our experimental design, which involves incubating a hybrid RNA-DNA oligomer (simply referred to as the RNA-containing oligomer) to quantify the hydrolysis rate of the 5',3'-phosphodiester bond at a single ribonucleotide within the aforementioned solutions.
The goal of this study is to understand the influence of bedrock composition (e.g., mafic-ultramafic, iron-rich, and magnesium-rich) and subsequent weathering of prebiotically relevant metals (i.e., Fe 2+ , Mg 2+ , and Mn 2+ ) which have been demonstrated to catalyze hydrolysis (Fedor, 2002), folding (Laing et al., 1994), non-enzymatic replication (Adamala & Szostak, 2013), translation (Bray et al., 2018), and impart novel catalytic function (Hsiao et al., 2013) on RNA stability. Furthermore, pH is simultaneously adjusted to reflect the composition of a hypothetical anoxic and CO 2 -dominated atmosphere at variable pressures in equilibrium with surface waters and pHs found at an average acid vent and an average alkaline vent (Kua & Bada, 2011). The end result is an analysis of single-stranded RNA stability and degradation kinetics in an array of simulated prebiotic geochemical spaces on Mars.

| Relevant observations-Hadean Earth and the last universal common ancestor (LUCA)
Little-to-no record exists of Earth's prebiotic history during the Hadean. Analysis of isotopic signatures and inclusions preserved within zircons (e.g., the oldest-surviving crustal material) suggest that a global ocean (Mojzsis et al., 2001) and the first continents (Wilde et al., 2001)  Phylogenetic reconstructions of the last universal common ancestor (LUCA), however, have since indicated that LUCA was most likely a mesophilic surface-dweller capable of UV repair (e.g., Cantine & Fournier, 2018). Ribosomal RNA (rRNA) ancestral state reconstructions show a GC-content consistent with mesophilic optimal growth temperatures (Galtier, 1999;Groussin et al., 2013) while protein reconstructions demonstrate a depletion of thermostable amino acids (Boussau et al., 2008;Zeldovich et al., 2007) and a divergence between informational and metabolic families inconsistent with a thermophilic origin (Berkemer & McGlynn, 2020). In addition, from a building block perspective, prior work by Kua & Bada, 2011 has demonstrated that ribose, cytosine (i.e., the least stable nucleobase), and the phosphodiester linkage are most stable at 0º C while Levy & Miller, 1998 concluded a high-temperature origin could not involve the canonical genetic code. Altogether, these results indicate thermophily may have been an evolutionary adaptation in response to a thermophilic bottleneck (e.g., the late-heavy bombardment) (Boussau et al., 2008).
Overall, phylogenetics and experimental work on RNA synthesis, stability, and function strongly indicates a planetary surface origin driven by UV photochemistry (Patel et al., 2015), common molecular F I G U R E 1 RNA-containing oligonucleotide cleavage assay. A hybrid RNA-DNA oligomer, 5'-Cy3-TTT-TTTrCTT-TTT-TTT-3', was designed to contain a single ribonucleotide (r) in between a chain of deoxyribonucleotides which could allow us to quantify cleavage at a single site. Representative gel scan displays two fluorescent bands belonging to either the intact 15-mer (Band 1) or the residual 7-mer (Band 2) cleaved at the single ribonucleotide site feedstocks (e.g., HCN) (Parkos et al., 2018;Patel et al., 2015;Toner & Catling, 2019), cool temperatures Kua & Bada, 2011;Levy & Miller, 1998;Moulton et al., 2000), and wet-dry cycling (Sasselov et al., 2020). Therefore, while the Hadean prebiotic record has been lost, we may be able to elucidate candidate environments by studying the surface of Mars.

| Relevant observations-Mars
Below we list relevant observations from rover/orbiter mission and modeling that have been utilized to synthesize candidate aqueous environments and are relevant for our discussion on the viability of an RNA world on Mars ( Figure 2).

| Mars Atmosphere and Volatile Evolution (MAVEN) Orbiter
Isotopic evidence indicates a continuous loss of a ≥ 0.5 bar CO 2dominated atmosphere (Jakosky et al., 2017) since the early Noachian ~ 4.1 Gya due to erosion by solar wind when the Martian dynamo is thought to have shut down (Lillis et al. 2013).

| Atmospheric Modeling
Various models have suggested a range of atmospheric compositions (e.g., H 2 , CO 2 , H 2 O, SO 2 , H 2 S); however, many have not been able to resolve counteracting cooling effects of atmospheric density and albedo due to aerosol and cloud formation (e.g., Tian et al., 2010). Primarily due to lower solar luminosity values in early history, atmospheric general circulation models (e.g., Forget et al., 2013;Wordsworth et al., 2013) have found it difficult to maintain the > 273 K mean annual temperature (MAT) seemingly required to support a "warm and wet" or "warm and arid" early Mars and valley network formation (e.g., Craddock et al., 2003). Most recently, work by Ramirez et al. 2020 has proposed that CO 2 -H 2 collision-induced absorption could have raised mean surface temperatures above 273 K. That is, assuming the presence of a hypothetical northern lowland ocean (Chan et al., 2018), atmospheric pressures as low as 0.55 bar CO 2 and 1% H 2 may have sustained a relatively warm and wet early Mars (Ramirez, 2017).
Instead, these models suggest a "cold and icy" early Mars climate (Head & Marchant, 2014) in which snow and ice were deposited

| Mars prebiotic geochemical solutions
Given the aforementioned observations (Figure 2) (Mittlefehldt, 1994). inside an anaerobic glove box (Coy). The atmosphere inside the glove box was N 2 with 2.5% -3% H 2 and internal circulation through a platinum catalyst maintained residual oxygen levels below 10 ppm.
All RNA reactions occurred inside the glove box on a miniPCR mini16 thermocycler (Amplyus, QP-1016-01) kept at 75°C in order to facilitate rapid RNA degradation.

| RNA degradation quantification
All RNA degradation experiments were quantified via urea polyacrylamide gel electrophoresis (National Diagnostics, EC-830 & EC-840) followed by imaging on a Typhoon 9,410 (GE Healthcare).

| RNA degradation kinetics
Of 5
Normally, these metal aquo complexes interact with secondary and tertiary RNA structures to neutralize the electrostatic repulsion of negatively charged phosphate groups brought into close proximity, or to increase local rigidity and join distal RNA structures by incorporating phosphate groups into their first coordination shell . However, in our experiments, we sought to quantify the effect of metal-catalyzed hydrolysis (i.e., RNA cleavage by transesterification), which is thought to be analogous to how certain ribozymes (e.g., the hammerhead self-cleaving ribozyme) utilize  (Fedor, 2002;Hampel & Cowan, 1997;Johnson-Buck et al., 2011). Namely, 1) acid/base interactions with water result in the activation of the ribose 2'-hydroxyl nucleophile,

F I G U R E 3
and 2) first shell phosphate ligands draw electron density and expose phosphorous to nucleophilic attack. Attack of the 2'-hydroxyl on the adjacent phosphate results in formation of a 2'-3' cyclic phosphate terminated oligonucleotide plus a second oligonucleotide product that begins with a 5'-hydroxyl.
Our results appear to reproduce the enhanced RNA degradation rates expected to be associated with each metal's respective acid dissociation constant (pK a ) in solution. Between pH 5.4 and 8, slower rates of metal-catalyzed hydrolysis occur in the presence of Mg 2+ (pK a = 11.4) followed by Fe 2+ (pK a = 9.6) and then Mn 2+ (pK a = 10.6) ( Figure 4, likely compound the rate of RNA cleavage due to mechanism 2) described above (Hampel & Cowan, 1997). That is, Fe 2+ is able to accept and retain electrons with greater affinity after its hydrated hexa aquo species loses a proton. At pH 9, it would appear that slower degradation rates occur in the presence of Fe 2+ (B max (h -1 ) = 47.4 x 10 -2 ) rather than Mg 2+ (B max (h -1 ) = 96.1 x 10 -2 ) contrary to our interpretation for pH 5.4 -8 ( Figure 5). Nonetheless, it is known that alkaline Fe 2+ solutions will begin to form insoluble species such as Fe(OH) 2 around pH ~ 9 (Gayer & Woontner, 1956 (Jin et al., 2018). Results for pH 3.2 were unexpected as increasing concentrations of prebiotically relevant metals decreased the rate of degradation (Figure 4), and additional work is required to understand the precise preservation mechanism.
However, researchers have proposed that perhaps metal ions can act as Lewis acids which stabilize the 2'-hydroxyl group and prevent nucleophilic attack of phosphorous (Fedor, 2002). Results for the basalt analogs demonstrate greater degradation rates with increasing concentrations of Fe 2+ relative to Mg 2+ (Figure 6). This is best observed at pH 6.7 and 8 where 80:20 (20 mM for all basalt analogs is inferred to be primarily Mg 2+ -dominated as Fe 2+ was observed to precipitate out of solution.

| RNA on Mars-polymerization, stability, and redox environments
In order to fully evaluate the viability of an origin of life via the RNA world on Mars, we must consider synthesis in addition to stability. In other words, replication of RNA must be faster than degradation to effectively explore the fitness landscape. Work on non-enzymatic RNA replication has demonstrated that metal catalysis actually increases the rate of polymerization by facilitating the deprotonation of the 3'-hydroxyl group in 2-methylimidazole nucleotides activating the 3'-hydroxyl as a nucleophile (Li et al., 2017 Athavale et al., 2012). This is because some models predict reducing, circumneutral, and cool conditions throughout the Hadean (Charnay et al., 2017;Kadoya et al., 2020). Specifically, studies have demonstrated that Fe 2+ may have preceded life's transition to Mg 2+ due to its versatility in catalyzing polymerization (e.g., Jin et al., 2018), ribozyme folding (Athavale et al., 2012), translation (Petrov et al., 2015), and imparting novel catalytic activity (Guth-Metzler et al., 2020;Hsiao et al., 2013;Okafor et al., 2017). It is therefore conceivable that Fe 2+ could have had a comparable influence on the stability, polymerization, and function of RNA on early Mars and should accordingly be discussed below.
In situ observations of Mars suggest pH regimes ranging from acidic (pH 2 -4) (e.g., Squyres & Knoll, 2005) to circumneutral (pH ~ 7) (e.g., Grotzinger et al. 2014) have existed on various locations under evolving redox conditions ( Figure 2). Namely, pre-to early Noachian Mars is inferred to have been primarily reducing and neutral and progressively became oxidizing by the late Noachian resulting in pervasive acidic surface conditions ( Figure 2) (Hurowitz et al., 2010;McLennan et al., 2019). Observations by MAVEN constrain the likely composition of an early atmosphere to primarily CO 2 -dominated at ≥ 0.5 bars, which accordingly would not acidify surface environments below pH ~ 6.7 (Kua & Bada, 2011) pre-to early Noachian. Modeling of continental weathering of early Earth basalts additionally suggests that waters with high alkalinity would have stabilized pH between 6.6 and 7 under a ~ 1 bar CO 2 atmosphere (Halevy & Bachan, 2017;Krissansen-Totton et al., 2018). The global acidification of Mars is therefore presumed to be the result of increasing atmospheric O 2 and/or photo-oxidation of Fe 2+ to Fe 3+ as its atmosphere was lost (Hurowitz et al., 2010) which would consequently be highly detrimental to the prospects of an Fe 2+ -mediated RNA world.
Our results indicate that a Mg 2+ -rich basalt (e.g., McSween, 2002;Mustard et al., 2005) sourcing metals to a slightly acidic (pH 5.4) aqueous environment on Mars would have best supported long-lived single-stranded RNA polymers through the mid-Noachian (Figures 2 and 5). Notwithstanding this observation, CO 2 pressures (10 bar) required to sufficiently acidify surface waters are not supported by atmospheric models (e.g., Forget et al., 2013;Tian et al., 2010) as buffering from basaltic aquifers would neutralize pH as indicated above. Low pH values on Mars would have most likely been prevalent by the late Noachian (Figure 2) or perhaps locally at natural acid springs (e.g., Varekamp et al., 2009).
Results from our experiments at pH 6.7 therefore represent the most accurate interpretation of potentially global conditions on Mars since the pre-Noachian (Bibring et al., 2006). Assuming a cool (e.g., as indicated by atmospheric models), reducing, and UV- rates due to differences in pH optimums ( Figure 6). Depending

| CON CLUS IONS
Discoveries of structural and regulatory RNA molecules suggest that life as we know it may have emerged from an earlier RNA world (Bernhardt, 2012). However, due to global resurfacing and recycling (Marchi et al., 2014), it has been challenging to reconstruct the types of real-world environments that may have existed on the Hadean Earth and hosted the origin(s) of life. We believe that Mars is the next best alternative to search for environments consistent with requirements imposed by the RNA world. In this study, we investigated the influence of bedrock composition (e.g., mafic-ultramafic, ironrich, and magnesium-rich) and subsequent weathering of prebiotically relevant metals (Fe 2+ , Mg 2+ , and Mn 2+ ) on RNA stability. These metals have been demonstrated to catalyze hydrolysis (cleavage), folding, polymerization, and impart RNA with novel catalytic properties. In addition, we simultaneously adjusted pH to reflect the composition of hypothetical CO 2 -dominated atmospheres in equilibrium with surface waters and waters at an acidic and alkaline vent. We determined that RNA stability depends on metal concentration and pH. Our results reproduce the enhanced RNA cleavage rates associated with each metal's respective acid dissociation constant (pK a ) and an increase in metal concentration ( Figure 4, Table 1). Degradation rates unexpectedly decreased with increasing metal concentration via an unknown preservation mechanism at pH 3.2 ( Figure 4) though so do rates of polymerization (e.g., Jin et al., 2018). At pH 9, we encountered Fe 2+ and Mn 2+ precipitation which artificially decreased cleavage rates (Figure 4). We conclude that a Mg 2+ -rich basalt sourcing metals to slightly acidic (pH 5.4) waters would therefore be the stability optimum (as determined here) for RNA on Mars. However, it is important to note RNA replication chemistry with Mg 2+ as the metal cofactor requires mildly alkaline pH values in order to result in net accumulation (Jin et al. 2018). Geologic evidence and modeling of basalt weathering otherwise indicate that early Mars pore and (pre-oxidizing) surface waters would have been near-neutral pH ~ 7.
Our experiments at pH 6.7 therefore represent the most accurate interpretation of potentially global conditions on Mars. Results from Fe 2+ at this pH and prior work on iron catalysis suggest that while high cleavage rates decrease RNA stability, catalysis may result in net accumulation (e.g., in the presence of template strands and monomers). However, global oxidizing conditions (due to the lack of a dynamo) on the surface of Mars may have led to significant RNA instability due to the precipitation of RNA-Fe 3+ complexes in Fe 2+rich environments possibly as early as ~ 4.1 Gya. We therefore presume the non-redox-sensitive Mg 2+ would have been the principal catalyst on Mars as hypothesized on Earth after the great oxidation event ~ 2.6 Gya (e.g., Athavale et al., 2012).

AUTH O R D I S CLOS U R E S TATE M E NT
No competing financial interests exist.