Constraints on the emergence of RNA through non-templated primer extension with mixtures of potentially prebiotic nucleotides

Abstract The emergence of RNA on the early Earth is likely to have been influenced by chemical and physical processes that acted to filter out various alternative nucleic acids. For example, UV photostability is thought to have favored the survival of the canonical nucleotides. In a recent proposal for the prebiotic synthesis of the building blocks of RNA, ribonucleotides share a common pathway with arabino- and threo-nucleotides. We have therefore investigated non-templated primer extension with 2-aminoimidazole-activated forms of these alternative nucleotides to see if the synthesis of the first oligonucleotides might have been biased in favor of RNA. We show that non-templated primer extension occurs predominantly through 5′-5′ imidazolium-bridged dinucleotides, echoing the mechanism of template-directed primer extension. Ribo- and arabino-nucleotides exhibited comparable rates and yields of non-templated primer extension, whereas threo-nucleotides showed lower reactivity. Competition experiments confirmed the bias against the incorporation of threo-nucleotides. The incorporation of an arabino-nucleotide at the end of the primer acts as a chain terminator and blocks subsequent extension. These biases, coupled with potentially selective prebiotic synthesis, and the templated copying that is known to favour the incorporation of ribonucleotides, provide a plausible model for the effective exclusion of arabino- and threo-nucleotides from primordial oligonucleotides.


Oligonucleotide Synthesis
Oligonucleotides were either purchased from IDT or synthesized in-house by solid phase synthesis on the Expedite 8909 DNA/RNA synthesizer.Phosphoramidites and reagents were purchased from ChemGenes (Wilmington, MA) and Glen Research (Sterling, MA).The synthesized oligonucleotides were deprotected and purified by Glen-Pak RNA purification cartridges (Sterling, MA).

Figure S2.
Mechanism of non-templated primer extension.(A) Schematic representation of a nontemplated primer extension reaction using an Alexa488-labeled RNA 12-mer with bridged dinucleotides (rG*rG) or activated mononucleotides (*rG).2AI was added to reduce the formation of bridged dinucleotides.(B) Gel electrophoresis images of non-templated primer extension reactions using bridged dinucleotides or activated mononucleotides.(C) Kinetic analysis with observed pseudo-first-order reaction rates (kobs).Reaction conditions: 1 μM primer (XJ-Alexa-12mer, Table S1) S1), 200 mM HEPES pH at 8.0, 50 mM MgCl2, and 20 mM *rC.The reaction mixtures were subjected to air-drying to clear pastes under ambient air to accelerate non-templated primer extension (as previously described in Figure S1) and were subsequently quenched at 24h.S1) and 6.7 mM each of *rC, *araC and *tC.The reactions were allowed to dry spontaneously under ambient air and were subsequently quenched at 24 h.The samples were then desalted on C18 ZipTip pipette tips, followed by injection in HR LC-MS.The generated compound lists using the Agilent MassHunter Qualitative Analysis software were matched with the calculated masses of all possible +1 non-templated primer extension products and their salt adducts (Table S2).S1) and 2ʹ-5ʹ phosphodiester linkages (XJ-2, Table S1) in (A) 1:1 ratio of XJ-1 and XJ-2 and (B) 1:10 ratio of XJ-1 and XJ-2.Note that the XJ-1 and XJ-2 oligomers were synthesized using TOM-protected RNA amidites to exclude the possibility of 2ʹ to 3ʹ migration of the TBDMS group.Standards with 3ʹ-5ʹ and 2ʹ-5ʹ linkage could not be separated on LC-MS.(C) TIC of the non-templated primer extension product of a 5ʹhexynyl DNA primer (XJ-9, Table S1) showed that nucleobases did not react.The chemical structure of the 5ʹ-hexynyl group is indicated.(D) TIC of the non-templated primer extension product of a 5ʹ-OH DNA primer (XJ-11, Table S1) showed that 5ʹ-OH did not react with the activated mononucleotides.Reaction conditions of (C)-(D): 20 mM activated mononucleotides, 100 μM 6-mer primer, 200 mM HEPES at pH 8.0, and 50 mM MgCl2.The reactions were allowed to dry spontaneously under ambient air and were subsequently quenched at 24 h.The samples were then desalted on C18 ZipTip pipette tips, followed by injection in HR LC-MS.S1) and (C) 5ʹ-hexynyl RNA primer (XJ-8, Table S1).Reaction conditions for (B) and (C): 20 mM activated mononucleotides, 100 μM 6-mer primer, 200 mM HEPES at pH 8.0, and 50 mM MgCl2.The reactions were allowed to dry spontaneously under ambient air and were subsequently quenched at 24 h.The samples were then desalted on C18 ZipTip pipette tips, followed by injection in HR LC-MS.

Figure S13.
Ion counts (%) as a function of mass-to-charge (m/z) of the +1 products and their salt adducts of *rC:*araC:*tC = 1:1:1 competition experiment (Reaction1, Set 3) with a modified primer.All mass-tocharge ratios here are in the -2 charged state (M-2H) 2-.Reaction conditions: 100 μM 6-mer primer (XJ-10, Table S1) and 6.7 mM each of *rC, *araC and *tC.The reactions were allowed to dry spontaneously under ambient air and were subsequently quenched at 24 h.The samples were then desalted on C18 ZipTip pipette tips, followed by injection in HR LC-MS.The generated compound lists using the Agilent MassHunter Qualitative Analysis software were matched with the calculated masses of all possible +1 nontemplated primer extension products and their salt adducts (Table S3).For further information on the HR LC-MS, please refer to the materials and methods section.a Within a given set, reactions are independent replicates initiated simultaneously whereas reactions across different sets are independent replicates performed on separate dates.Given that the ambient flow rates and moisture levels can vary daily, reactions performed across different sets exhibited greater variability compared to those within the same set.b,c n ≥ 2 replicates.d Individual extension ratio was calculated based on the assumption that all internal sites (5 in total) and terminal sites (2 in total) contributed equally to the extension.

Figure S3 .
Figure S3.The hydrolysis rates and half-lives of (A) bridged dinucleotides (20 mM rC*rC) and (B) activated mononucleotides with 2AI (20 mM *rC with 100 mM 2AI) derived from the 31 P NMR experiments in Figure S5.Reaction conditions: the respective concentration of activated species as indicated above, 50 mM MgCl2, 200 mM HEPES at pH 8.0, 10% (v/v) D2O, incubation at RT. Error bars indicate standard deviations of the mean, n=2 replicates.

Figure S7 .
Figure S7.Variability of spontaneous air-drying experiments within the same set (performed at the same time and exposed to the identical ambient air flow rate and moisture level).(A) Schematic representation of non-templated primer extension reaction using an Alexa488-labeled RNA 6-mer with activated mononucleotides (*rC).(B) Gel electrophoresis image of non-templated primer extension reactions within the set.(C) Quantification of extended products relative to the primer (100%).Reaction conditions: 10 μM primer (XJ-Alexa-6mer, TableS1), 200 mM HEPES pH at 8.0, 50 mM MgCl2, and 20 mM *rC.The reaction mixtures were subjected to air-drying to clear pastes under ambient air to accelerate non-templated primer extension (as previously described in FigureS1) and were subsequently quenched at 24h.

Figure S8 .
Figure S8.Effect of primer concentrations on spontaneous air-drying experiments.(A) Schematic representation of non-templated primer extension reaction using an Alexa488 labeled RNA 6-mer (TableS1) with activated mononucleotides (*rC).(B) Gel electrophoresis image of non-templated primer extension reactions across different primer concentrations.(C) Quantification of +1 product relative to primer (100%).(D) Derived total amount of +1 product at different primer concentrations for a 60 μL reaction.100 μM primer concentration yielded the highest amount of +1 product for characterization and was therefore chosen as the most optimal condition for subsequent spontaneous air-drying and competition experiments.Reaction conditions: primer (10 μM, 20 μM, 50 μM or 100 μM), 200 mM HEPES pH 8.0, 50 mM MgCl2 and 20 mM *rC.The reaction mixtures were dried down to clear pastes under ambient air to speed up non-templated primer extension (FigureS1) and were subsequently quenched at 24h.

Figure S9 .
Figure S9.Plot of counts (%) as a function of mass-to-charge (m/z) of the +1 products and their salt adducts of *rC:*araC:*tC = 1:1:1 competition experiment using an 5ʹ-OH RNA 6-mer.Species shown are in the -2 charged state (M-2H) 2-.Reaction conditions: 100 μM 6-mer primer (XJ-5, TableS1) and 6.7 mM each of *rC, *araC and *tC.The reactions were allowed to dry spontaneously under ambient air and were subsequently quenched at 24 h.The samples were then desalted on C18 ZipTip pipette tips, followed by injection in HR LC-MS.The generated compound lists using the Agilent MassHunter Qualitative Analysis software were matched with the calculated masses of all possible +1 non-templated primer extension products and their salt adducts (TableS2).

Figure S11 .
Figure S11.Origin of the double peak pattern in the chromatograms of the competition experiments in Figure 4. (A-B) Total ion chromatograms (TIC) from the co-injection of oligomers ending with 3ʹ-5ʹ (XJ-1, TableS1) and 2ʹ-5ʹ phosphodiester linkages (XJ-2, TableS1) in (A) 1:1 ratio of XJ-1 and XJ-2 and (B) 1:10 ratio of XJ-1 and XJ-2.Note that the XJ-1 and XJ-2 oligomers were synthesized using TOM-protected RNA amidites to exclude the possibility of 2ʹ to 3ʹ migration of the TBDMS group.Standards with 3ʹ-5ʹ and 2ʹ-5ʹ linkage could not be separated on LC-MS.(C) TIC of the non-templated primer extension product of a 5ʹhexynyl DNA primer (XJ-9, TableS1) showed that nucleobases did not react.The chemical structure of the 5ʹ-hexynyl group is indicated.(D) TIC of the non-templated primer extension product of a 5ʹ-OH DNA primer (XJ-11, TableS1) showed that 5ʹ-OH did not react with the activated mononucleotides.Reaction conditions of (C)-(D): 20 mM activated mononucleotides, 100 μM 6-mer primer, 200 mM HEPES at pH 8.0, and 50 mM MgCl2.The reactions were allowed to dry spontaneously under ambient air and were subsequently quenched at 24 h.The samples were then desalted on C18 ZipTip pipette tips, followed by injection in HR LC-MS.

Figure S12 .
Figure S12.Both internal and terminal hydroxyls can react in non-templated primer extension.(A) Schematic representation of the internal 2ʹ-OHs and terminal OHs in an RNA oligomer (5ʹ-hexynyl-rU rA rA rC rU rC).Overlay of TCC (purple) and ECC (blue: primer; pink: internal 2ʹ-OH reaction; yellow: terminal OH extension) of the non-templated primer extension product with (B) 5ʹ-hexynyl RNA primer ending with dideoxy cytosine (XJ-7, TableS1) and (C) 5ʹ-hexynyl RNA primer (XJ-8, TableS1).Reaction conditions for (B) and (C): 20 mM activated mononucleotides, 100 μM 6-mer primer, 200 mM HEPES at pH 8.0, and 50 mM MgCl2.The reactions were allowed to dry spontaneously under ambient air and were subsequently quenched at 24 h.The samples were then desalted on C18 ZipTip pipette tips, followed by injection in HR LC-MS.

Table S6 . Observed and calculated masses of the hetero-bridged dinucleotides containing tC in a one-pot mixture of 1:1:1 *rC:*araC:*tC. a
Activated mononucleotides (*rC, *araC, and *tC) were mixed in the ratio of 1:1:1.The mixture got spontaneously air-dried.It was then resuspended in LC-MS grade water and immediately submitted for LC-MS analysis. a