Thermal Habitat for RNA Amplification and Accumulation

The RNA world scenario posits replication by RNA polymerases. On early Earth, a geophysical setting is required to separate hybridized strands after their replication and to localize them against diffusion. We present a pointed heat source that drives exponential, RNA-catalyzed amplification of short RNA with high efficiency in a confined chamber. While shorter strands were periodically melted by laminar convection, the temperature gradient caused aggregated polymerase molecules to accumulate, protecting them from degradation in hot regions of the chamber. These findings demonstrate a size-selective pathway for autonomous RNA-based replication in natural nonequilibrium conditions.

Figure 1

Heat flux across water-filled pores drives RNA-catalyzed RNA replication. (a) A heat flow is used to create temperature differences across a water-filled pore. The temperature differences induce both thermophoresis of molecules (dashed arrows), moving them along thermal gradients, and convection of water (solid arrows). Simulations predict that the interplay of the two physical nonequilibrium effects locally concentrates the RNA polymerase toward cold regions of the cylindrical chamber, where it is protected against thermal degradation. Because of the strongly length-dependent thermophoretic properties of RNA [22], shorter RNA molecules are not accumulated, but are subjected to temperature oscillations to achieve the necessary strand separation in the warm spot after their template-directed replication in the cold areas. (b) A natural setting for such a heat flow could be the dissipation of heat across volcanic or hydrothermal rocks. This leads to temperature differences over porous structures of various shapes and lengths.

Figure 2

Convective RNA-catalyzed replication of RNA. (a) The convection system and a thermal cycler showed similar yields of primer extension on the 35-nucleotide RNA template. (b) The polymerase ribozyme exponentially amplified starting template concentrations as low as 100 fM in 24 h. The amplification can be described theoretically by a two-parameter growth equation (Supplemental Material [24]). A maximum amplification of $2\times {10}^5$-fold was observed in the convection chamber. Error bars indicate the deviation from duplicate experiments. (c) Thin lines show simulated stochastic trajectories for a 35mer inside the convection chamber of $500\mu \mathrm{m}$ height and 2.25 mm radius. (d) In the convection chamber, RNA mostly resided at low-temperature regions, where polymerization could occur, and passed quickly through high-temperature regions that enabled strand separation. Stochastic simulations found a mean temperature cycling time of 26 min. Thermal cycler experiments were performed with cycles of $17°\mathrm{C}$ for 20 min, then $68°\mathrm{C}$ for 2 s. (e) Degradation of RNA is almost 5 orders of magnitude faster at higher temperatures [26].

Figure 3

Thermophoretic accumulation of nucleic acids. (a) Accumulation of nucleic acids inside the convection chamber with a central heat source implemented via an IR laser is shown for 35mer single-stranded DNA, 210mer double-stranded DNA, and folded single-stranded DNA and RNA corresponding to the polymerase sequences (from left to right). The scale bar corresponds to 1 mm. (b) Imaging the different strands inside the reaction buffer revealed the formation of conglomerates in the case of the polymerase sequences. The scale bar corresponds to $250\mu \mathrm{m}$ (c) The height average $⟨c/c_0{⟩}_z$ of the simulated relative concentrations (lines) reproduced the experimental fluorescence signal (symbols) for all species. The data points represent the radial averaging of the background corrected fluorescence images with respect to the central point of the chamber. (d) By including diffusive, convective, thermophoretic, and diffusiophoretic transport in the finite element simulation, the model could capture the ring-shaped concentration enhancement of the polymerases observed in the experiments, with a maximum of 66-fold concentration increase after 60 min. The simulation was executed in a compartment of $500\mu \mathrm{m}$ and 2.25 mm height and radius, respectively.