Extended Data Fig. 2: Turnover of the reusable catalyst.
From: Heat-rechargeable computation in DNA logic circuits and neural networks
a-b, Simulations and fluorescence kinetics experiments of a reusable DNA catalyst (a) compared with a use-once DNA catalyst (b). Four distinct combinations of strand purities were investigated. Gate strands and input strands were ordered from IDT DNA either unpurified (standard desalting) or PAGE purified. In both cases, we annealed and then PAGE purified the hairpin gate in the reusable catalyst and the double-stranded gate in the use-once catalyst in our lab. For double-stranded gates, in-house purification is necessary to reduce stoichiometry errors. For hairpin gates, we expect in-house purification to help reduce malformed structures. In modeling the reusable catalyst, kd = 5 and 2 per second were applied to IDT unpurified and purified input strand, respectively, accounting for synthesis errors within the invading toehold on the input strand; kd1 = 0.02 and 0.05 per second were applied to IDT unpurified and purified gate strand, respectively, accounting for synthesis errors within the dissociation toehold on the gate strand. A 58-fold and 82-fold catalytic turnover were observed at 24 hours with 0.01 × input when both gate and input strands were IDT unpurified and purified, respectively. In modeling the use-once catalyst, effective strand displacement rate k = 1.1 to 3.2 × 105 per molar per second were estimated based on the experimental data. A 95-fold and 90-fold catalytic turnover were observed at 24 hours with 0.01 × input when both gate and input strands were IDT unpurified and purified, respectively. In general, purified input strand resulted in slightly faster kinetics whereas purified gate strand resulted in slower kinetics. This is counterintuitive but consistent with our prior observation shown in Fig. S18 of the logic circuit construction using the seesaw gate motif14. The benefit of IDT purified gate strand is evident for the reusable catalyst but not for the use-once catalyst, presumably due to the strand length difference: the two strands in the use-once gate are 29 and 35 nucleotides, whereas the hairpin strand in the reusable gate is 74 nucleotides. With a 99.4% nucleotide coupling efficiency by IDT, 81 to 84% full-length products are expected for the two strands but only 64% full-length product is expected for the hairpin strand.
