Sliding over the Blocks in Enzyme-Free RNA Copying – One-Pot Primer Extension in Ice

Template-directed polymerization of RNA in the absence of enzymes is the basis for an information transfer in the ‘RNA-world’ hypothesis and in novel nucleic acid based technology. Previous investigations established that only cytidine rich strands are efficient templates in bulk aqueous solutions while a few specific sequences completely block the extension of hybridized primers. We show that a eutectic water/ice system can support Pb2+/Mg2+-ion catalyzed extension of a primer across such sequences, i.e. AA, AU and AG, in a one-pot synthesis. Using mixtures of imidazole activated nucleotide 5′-monophosphates, the two first “blocking” residues could be passed during template-directed polymerization, i.e., formation of triply extended products containing a high fraction of faithful copies was demonstrated. Across the AG sequence, a mismatch sequence was formed in similar amounts to the correct product due to U·G wobble pairing. Thus, the template-directed extension occurs both across pyrimidine and purine rich sequences and insertions of pyrimidines did not inhibit the subsequent insertions. Products were mainly formed with 2′-5′-phosphodiester linkages, however, the abundance of 3′–5′-linkages was higher than previously reported for pyrimidine insertions. When enzyme-free, template-directed RNA polymerization is performed in a eutectic water ice environment, various intrinsic reaction limitations observed in bulk solution can then be overcome.


Dependence on the presence of metal ions
The non-competetive primer extensions for setup 1 (ImpG/t 1 and ImpU/t 2 , see main text Table 2) were tested in the absence of Mg 2+ and Pb 2+ ions, where about 1% of primer was elongated with G across the CC-motif and no elongation by U was detected at all across the AA-motif template. Figure S1 below shows the primer conversions after incubation at -18.4 °C with or without metal ions after 2 and 5 days (all other parameters were kept constant) Figure S1: Control reactions without metal ions. FP elongated with: red -ImpG in presence of t 1 , blue -ImpU in presence of t 2 . Control reactions performed in absence of Mg(NO 3 ) 2 and Pb(NO 3 ) 2

Monomer compositions for competitive elongation in setup 1
From earlier results of elongation reactions in the eutectic phase in ice, 1 it has been established that in case of ImpU in presence of the AA template (t 2 ) the best yield is achieved if the monomer is provided at 1.75 mM in the reaction mixture. In case of ImpG that optimal concentration is 0.6mM in presence of a CC template (t 1 ). In the competitive setup 1 using the AA-template these concentrations could either be interpreted as intrinsic to the particular primer/template system or the monomer provided. The following conditions are reported here: i) the total concentration of monomers is 1.75 mM, the optimal concentration observed using only ImpU on its respective template (t 2 ); ii) monomers are supplied at concentrations correspond to their individual optimal concentration, i.e. ImpU at 1.75 mM and ImpG at 0.6 mM; iii) a control experiment, where both monomers were provided at a significantly higher concentration in a 1:1 ratio (2.5 mM ImpU, 2.5 mM ImpG). The results shown in Figure S2, compare the three cases with the non-competitive reactions. In the first case ( Figure S2b) clearly the elution pattern of the U reaction is reproduced and at a lower yield. In the second case ( Figure S2c) peaks from both individual cases can be observed, including unspecific elongations in the FP+1 and +2 region. In the third case ( Figure S2d)), the higher concentration let additions of G completely dominates the reaction products, probably due to the stronger stacking interactions of the purine base. Most notably the total yield of the reaction is highest in case ii. (Total primer conversions: i) 23%, ii) 35%, iii) 17%). Figure S2: Monomer composition in competitive reactions. All analyses where performed after 5 day incubation in the presence of FP and t 2 , unless indicated otherwise. a) Overlay of the three conditions tested; b) 1:1 ImpU/ImpG; [ImpN] = 1.75mM, in overlay with non-competetive reactions; c) 1.75mM ImpU / 0.6mM ImpG and controls; d) 1:1 ImpU/ImpG, [ImpN] = 5mM.

Mass Spectrometry of selected elongation products
Products collected from anion-exchange using a fresh DNAPac PA 200 column mM NaClO 4 , 0-2 min, 0% B; 2-42 min, convex gradient to 65% B, 42 nm). The peaks selected for MS analysis are indicated on the HPLC chromatogram was desalted using pipette tips prepared followed by evaporation to up-concentrate the samples (> 5 pmol/µl). Ali to 1.5µl matrix droplets (Matrix: 300 mM trihydroxyacetophenone (THAP) in ethanol, containing 30% v/v 0.1M diammonium citrate in water). Spectra obtained in linear Figures S3, S5 and S7 show the matching the indicated fraction counters 2, 3 and 4, respectively. Figure S3: Fractionation of setup 2 products. collected for MALDI MS spectrometry (indicated by black bars). The numbers of the fractions correspond to the numbering of MS spectra in Figure S4, where 1 is the primer  Overview of isolated product species for setup 2. For each HPLC fraction of the reaction mixture the observed species ote that for each molecule at least two regio-isomers exist, that elute differently from HPLC. Regiochemistry indicated if known, otherwise labels (1) and (2) Figure S5. The detection range of each MS spectrum overlaps with the m/z range of the product ± 1 nt (except for the primer, FP). For each peak a concomitant degradation peak at about 151 m/z units below the main peak corresponding to the loss of a guanine base (this peak was annotated in spectrum 2 but omitted in later spectra). Figure S5: Fractionation of setup 3 products. were later collected for MALDI MS spectrometry (indicated by black bars). The numbers of the fractions correspond to the numbering of MS spectra in Figure S6, where elongation products.   Figure S6: MALDI spectrum of the HPLC purified fractions 1-11 of setup 3. The numbers correspond to the indicated fractions in Figure S5. For each peak a concomitant degradation peak at about 151 m/z units below the main peak corresponding to the loss of a guanine base (this peak was annotated in 2 but omitted in later spectra). Figure S7: Fractionation of setup 4 products. were later collected for MALDI MS spectrometry (indicated by black bars). numbering of MS spectra in Figure S8, where elongation reaction mixture the observed species are Table S3: Overview of isolated product species of setup 4 and found masses are listed. Note that for each molecule at least two regio products are labeled (main) while trace products are labeled (trace). nucleotides are written in square brackets with variables number of inserted monomers. The combinations pertaining to, these cases, the molecular masses are given as the average ± Fraction Species

Setup 4 mass spectrometry data
Setup 4 mass spectrometry data etup 4 products. Analytical HPLC Chromatogram obtained from setup were later collected for MALDI MS spectrometry (indicated by black bars). The numbers of the fractions correspond to the , where 1 is the primer FP and 2-8 are the subsequent elongation products or mixtures of reaction mixture the observed species are shown. of isolated product species of setup 4. For each HPLC fraction of the reaction mixture the observed are listed. Note that for each molecule at least two regio-isomers exist, that elute differently from HPLC. Major eled (main) while trace products are labeled (trace). When different sequences nucleotides are written in square brackets with variables i and j as counters : FP-[X i Y j ], where inserted monomers. The combinations pertaining to, e.g., FP-[U i C j ], n = 2 are FP-UU, these cases, the molecular masses are given as the average ± ∆m/z (U-C) = 0.98.  Figure S8: MALDI spectrum of the HPLC purified fractions 1-5, 7 and 8 of setup 4. The numbers correspond to the indicated fractions in Figure S7. For each peak a concomitant degradation peak at about 151 m/z units below the main peak corresponding to the loss of a guanine base (this peak was annotated in 2 but omitted in later spectra).

Control reactions for setup 3 and 4
As described in the main manuscript, various control reactions (Table S4) were needed to help identifying the FP+3 products fromed during primer extensions in setups 3 ( Figure S9) and 4 ( Figure S10). For each setup an overlay is shown to illustrate which products can be formed if not all nucleotides are present. This allows the identification of FP+1 incorporations of the various nucleotides and to compare the unfaithfully formed FP+2 and FP+3 products to the ones that are observed when the full set of required monomers is supplied. Thus, by removing *pU from the reaction pool, a template mismatch is inevitable if FP+1 is formed , i.e., FP-Â or FP-Ĉ for setup 3 or 4, respectively. If *pA (setup 3) or *pC (setup 4) is removed from the reaction mixture the second insertion will create a mismatch with the template, i.e., on forming FP+UÛ across AU or AG, respectively. In all these controls, the amount of formed, bears information about the fidelity of the reaction in the full system, i.e. with all required monomers present. A further control reaction was carried out using FP-3′ U (commercially synthesized oligomer) as a primer, to gauge the efficiency of the reaction after a first correct elongation step. The template motif for the first two insertion in these controls no longer qualifies as "blocking sequence" and accordingly it could be shown that, if the first insertion (U across A) is efficient and correct, the completion to FP+3 is also efficient. C, dG n/a S10, b U, dG n/a S10, c 5 (FP-3′ U) t3 (GC) C, dG n/a S10, d Figure S9: Setup 3 control experiments. a) Overlay of the full setup 3 reaction with controls a) with *pU or *pU/*pA only, b) control without *pU, c) control without *pA. d) overlay with the elongation of FP-3′ U by A and dG.

Figure S10: Setup 4 control experiments.
Overlay of the full setup 4 reaction with controls a) *pU and *pU/*pC only b) control without *pU. c) control without *pC. d) overlay with the elongation of FP-3′ U by C and dG.

Enzymatic digestions
The regioselectivity of the products manuscript) was analyzed by enzymatic cleavage selective for 3 schematic of this process and illustrates how the cleaved and remaining products were quantified. The amount of 3′-5′ linkages formed in the first elongation was measured by integral of the digestion product (phosphate on the primer, FP-p). Aliquots of the reaction mixtures were incubated for 30 min at 37° C with RNAse ONE™ (10u/µl) 1 . Figures S12 and S13 show HPLC chromatogram the enzyme (dig.) or without (ref.)). At least three repetitions of each experiment were done. The ratio of the 2 vs. 3′-5′ regio-isomers for the first nucleotide incorporation was calculat chromatogram integrals after incubation at 37 The incubated control was also compared to the original sample to determine degradation due to temperature. The digestion allowed assessing the maximal length of products formed (to our limits of sensitivity), as the enzyme does not cleave products with all 2 with U (longest observable product, G (FP-(G) 6 ). Indeed, the yield dropped abruptly after the first non setup 2, whereas the yields of FP-(U) Figure S11: Schematic illustration of the RNAse digestion regioselectivity ratio from the chromatographic traces before (blue, "ref.") and after (red, "dig.") treatment with RNAse. consumed is shown by the dark-grey shaded areas. The digestion product ( the 3′-5′ to 2′-5′ ratio of the first elongation. If this area corresponds to the total of digested product ( products exclusively are elongated with the same regiochemistry.
The regioselectivity of the products formed in the non-competetive studies in setups manuscript) was analyzed by enzymatic cleavage selective for 3′-5′-phosphodiester linkages. Figure S1 schematic of this process and illustrates how the cleaved and remaining products were quantified. The amount of es formed in the first elongation was measured by integral of the digestion product (phosphate on the p). Aliquots of the reaction mixtures were incubated for 30 min at 37° C with RNAse ONE™ (10u/µl) show HPLC chromatograms for ImpG/t 1 and ImpU/t 2 , respectively ( . At least three repetitions of each experiment were done. The ratio of the 2 isomers for the first nucleotide incorporation was calculated from the difference of the chromatogram integrals after incubation at 37 ˚C of digested vs. control (horizontally shaded area in The incubated control was also compared to the original sample to determine degradation due to temperature. The digestion allowed assessing the maximal length of products formed (to our limits of sensitivity), as the enzyme does not cleave products with all 2′-5′ connectivity. Consequently, it was observed that the elongation with U (longest observable product, FP-(U) 9 ) was less influenced by the presence of a template strand than with Indeed, the yield dropped abruptly after the first non-cognate incorporation at (U) n (setup 1) decayed in an exponential-like fashion from Schematic illustration of the RNAse digestion. and. The example illustrates the analytical procedure to calculate regioselectivity ratio from the chromatographic traces before (blue, "ref.") and after (red, "dig.") treatment with RNAse.
grey shaded areas. The digestion product (FP-p, light-grey shaded area) was quantified to obtain ratio of the first elongation. If this area corresponds to the total of digested product ( products exclusively are elongated with the same regiochemistry. that cleaves between any two ribonucleotides leading to a 2′-3′-cyclic nucleotide monophosphate These 15 setups 1 (see Table 1 in the phosphodiester linkages. Figure S11 shows a schematic of this process and illustrates how the cleaved and remaining products were quantified. The amount of es formed in the first elongation was measured by integral of the digestion product (phosphate on the p). Aliquots of the reaction mixtures were incubated for 30 min at 37° C with RNAse ONE™ (10u/µl) , respectively (after incubation with . At least three repetitions of each experiment were done. The ratio of the 2′-5′ ed from the difference of the ˚C of digested vs. control (horizontally shaded area in Figure S11). The incubated control was also compared to the original sample to determine degradation due to temperature.
The digestion allowed assessing the maximal length of products formed (to our limits of sensitivity), as the connectivity. Consequently, it was observed that the elongation was less influenced by the presence of a template strand than with cognate incorporation at FP-(G) 4 in case of like fashion from FP-(U) 3 to FP-(U) 6 .
illustrates the analytical procedure to calculate regioselectivity ratio from the chromatographic traces before (blue, "ref.") and after (red, "dig.") treatment with RNAse. Material grey shaded area) was quantified to obtain ratio of the first elongation. If this area corresponds to the total of digested product (dark-grey shaded area), 2′-5′ cyclic nucleotide monophosphate These Figure S12: Enzymatic digestion of elongation products with ImpU in setup 1.
dTdGG. FP-p denotes the digestion product, see Figure S1 Figure S13: Enzymatic digestion of elongation products with ImpG in setup 2 dTdGG. FP-p denotes the digestion product, see Figure S Enzymatic digestion of elongation products with ImpU in setup 1. FP = fluorescent primer, 6 denotes the digestion product, see Figure S11. Digestions performed with 20 µl aliquots of 14 day sample atic digestion of elongation products with ImpG in setup 2. FP = fluorescent primer, 6 denotes the digestion product, see Figure

Enzymatic digestions after competitive elongation.
A typical HPLC analysis for setups 2 the overlay of crude product and degraded with RNAse ONE (as described above).  Figure S14, as the overlay of crude product and degraded with RNAse ONE (as described above).

Digestions of triply elongated primers
By isolating FP+3 prodcts and subjecting them to RNAse treatment, it became possible to analyse the digestion products of the triply elongated primers. These products would otherwise co-elute with, e.g. FP+1 and FP+2 strands. For the reported data, please note the following degradation schemes ( 2′ = 2′-5′-, 3′ = 3′-5′-linkage, x = both regioisomers are possible educts):   Table 4 in the main manuscript, with addition of breaking down the FP-2′ N x N x N group of regiomers into specific regiomers.