Experimental Tests of the Virtual Circular Genome Model for Nonenzymatic RNA Replication

Abstract

The virtual circular genome (VCG) model was proposed as a means of going beyond template copying to indefinite cycles of nonenzymatic RNA replication during the origin of life. In the VCG model, the protocellular genome is a collection of short oligonucleotides that map to both strands of a virtual circular sequence. Replication is driven by templated nonenzymatic primer extensions on a subset of kinetically trapped partially base-paired configurations, followed by the shuffling of these configurations to enable continued oligonucleotide elongation. Here, we describe initial experimental studies of the feasibility of the VCG model for replication. We designed a small 12-nucleotide model VCG and synthesized all 247 oligonucleotides of lengths 2 to 12 corresponding to this genome. We experimentally monitored the fate of individual labeled primers in the pool of VCG oligonucleotides following the addition of activated nucleotides and investigated the effect of factors such as oligonucleotide length, concentration, composition, and temperature on the extent of primer extension. We observe a surprisingly prolonged equilibration process in the VCG system that enables a considerable extent of reaction. We find that environmental fluctuations would be essential for continuous templated extension of the entire VCG system since the shortest oligonucleotides can only bind to templates at low temperatures, while the longest oligonucleotides require high-temperature spikes to escape from inactive configurations. Finally, we demonstrate that primer extension is significantly enhanced when the mix of VCG oligonucleotides is preactivated. We discuss the necessity of ongoing in situ activation chemistry for continuous and accurate VCG replication.

Figure 1

(A) Schematic illustration of the virtual circular genome model. The green circle represents the virtual genome that does not correspond to any actual oligonucleotide. A subset of the real oligonucleotides in the VCG system is illustrated as the blue and red arrows. Dotted lines, along with the bold arrows, showed how the oligomers map onto the virtual circular genome. The two complementary sequences selected for this study are shown inside the green circle. The direction from 5′ to 3′ is clockwise for the blue sequence and counterclockwise for the red sequence, which is the same direction as the arrows. Adapted from Figure 3 of ref (23) with permission under a Creative Commons Attribution 4.0 International License. Copyright 2021 Zhou et al.; Cold Spring Harbor Laboratory Press for the RNA Society. (B) Examples of productive and nonproductive configurations of annealed oligonucleotides.

Figure 2

Demonstration of extension inside the virtual circular genome system. (A) Comparison between the VCG system and a single-template system, with schematic representations of the two experiments shown flanking the PAGE gel image. The VCG system contains 1 μM of all of the VCG oligos listed in Table S1. The single-template system contains 1 μM of the template and 1 μM of the primer. Extensions were monitored using trace 5′-³²P-GUGAUG added to the reactions. The small VCG diagram is adapted from Figure 3 of ref (23) with permission under a Creative Commons Attribution 4.0 International License. Copyright 2021 Zhou et al.; Cold Spring Harbor Laboratory Press for the RNA Society. (B) Continuous VCG extension for 3 days, with and without periodic replenishment of 20 mM activated N*N and 90 °C heat pulses. The scheme on the right demonstrates different treatments for the three reactions. All reactions were conducted at room temperature, with 50 mM MgCl₂, 200 mM Tris–HCl (pH 8.0), and 20 mM pre-equilibrated N*N.

Figure 3

Virtual circular genome extension with different oligonucleotide compositions. (A) VCG extension when concentrated vs diluted. (i) Concentration of each oligonucleotide at the indicated length. (ii) VCG extension measured by % unextended 5′-³²P-GUGAUG. (iii) Melting temperature of p-GUGAUG measured as a function of concentration in the primer extension buffer. (B) VCG extension at different concentration gradients. The concentration gradient is expressed as [(pN)_i]/[(pN)_i+1], starting at 1 μM of each 12-mer. (i) Oligonucleotide concentrations in each gradient. The concentrations of 2–5-mer in 1.41× exceed the y-axis limit. See Table S2 for all concentrations. (ii) VCG extension under different concentration gradients. (C) Extension in partial VCG mixtures containing only the longer or shorter oligomers. (i) Oligonucleotide composition in the partial VCG system. (ii) VCG extension in the partial system. All reactions were measured by the extension of 5′-³²P-GUGAUG (<0.05 μM) conducted at room temperature, with 50 mM MgCl₂, 200 mM Tris–HCl (pH 8.0), and 20 mM pre-equilibrated N*N. See Table S2 for detailed oligomer concentrations in different VCG mixtures.

Figure 4

Length dependence and temperature effect on the primer extension in the 1× VCG system. (A(i)) Schematic representation of possible base-paired configurations between radiolabeled oligonucleotides of varying lengths and complementary 12-mers. (ii) Extension of oligonucleotides with different lengths in the 1× VCG oligo mix, represented by the percentage of unextended 5′-³²P-labeled oligonucleotide over time. (iii) Sequences of the labeled oligomers and their melting temperatures, measured in the primer extension buffer. (B) Heat pulses facilitate the continuous VCG extension of a 10-mer (5′-³²P-UGUGGUGAUG). (i) Experimental scheme. (ii) Measured extension with or without the heat pulses. The heat pulses were performed by 10 s of 90 °C heating, followed by immediate 1 min cooling on ice. The replenishments were performed by adding 10 mM of equilibrated and lyophilized N*N powder. (C) Lower temperature facilitates the VCG extension of a 4-mer (5′-³²P-GAUG). (i) A scheme showing that a tetramer in the VCG has a higher chance to anneal to a complementary strand at lower temperatures. (ii) 4-mer extension in VCG at different temperatures. All reactions were conducted in 1× VCG with 200 mM Tris–HCl (pH 8.0), 50 mM MgCl₂, and an initial addition of 20 mM N*N.

Figure 5

Detection of 5′-pyrophosphate-capped oligonucleotides. (A) Schematic representation of the phosphatase deradiolabeling of the extension products. The 5′-³²P labels were shown as stars. The 5′-³²P-oligonucleotides would be dephosphorylated while the 5′-Np³²P-oligonucleotides would be protected. (B) PAGE gel analysis of the extension products with or without phosphatase digestion. The VCG extension was performed with the 1× VCG mixture and 20 mM N*N, while the untemplated reactions were performed with 1 μM 6-mer and 20 mM of the indicated imidazolium-bridged homodimers. See Figure S7B for more details. All reactions were run at room temperature for 24 h. Phosphatase-digested products were loaded at the same concentration as the untreated sample. Authentic samples were run alongside the PAGE gel and are indicated in the figure.

Figure 6

Demonstration of possible strategies to improve VCG extension (A–B) Significantly enhanced VCG extension after preactivation with either a 1.41× or U-shaped gradient. (i) Oligo concentration of each length. (ii) Comparison between the extensions of 5′-³²P-GUGAUG inside the VCG system with or without preactivation. (C) Faster extension in a U-shaped VCG mix with the more reactive 3′-NH₂-2AIpddN modification as a model system.