target="_blank" rel="nofollow" href="#ulink_7ca7a628-dfc7-5e88-a754-d2b4227d9aa3">Figure 4.5 Design of molecular beacon (MB)‐based DNA logic gates. (a) MB probe fluoresces upon bind complementary analyte. (b) The 4WJ‐1 design for a YES gate. Oligoethylene glycol linkers triethylene glycol (TEG) are shown as dashed lines.
Source: Based on Lake et al. [63]
. (c) The 4JW‐2 design of YES gate.
Source: Based on Cornett et al. [64].
We have designed and optimized individual 4WJ logic gates in solution with the main purpose of integrating them into computational circuits [63,64,66,67]. The designs of NOT and AND gates are presented in Figure 4.6a,b. NOT gate forms a fluorescent complex with an MB probe; addition of an oligonucleotide input disrupts the 4WJ structure, which results in the release of the MB probe from the NOT complex yielding low fluorescent signal. AND gate consists of three DNA strands folded in hairpin structures (Figure 4.6b). Two oligonucleotide inputs open their correspondent AND hairpins and trigger association of the two strands with the MB probe. Again, the complex is stabilized by the 4WJ structure. The output of the gates is a DNA fragment formed by the portions of the two AND strands (e.g. blue and green in Figure 4.6b). Therefore, the 4WJ gate design preserves the input–output homogeneity, which is important for the integration of the gates into circuits, since an output of one 4WJ gate can be recognized as an input by the downstream gate. We achieved three layers of gate integration in solution by building an XOR logic from one OR, two AND, and two NOT gates [66]. Expectedly, the signal intensity would decrease upon adding each new level of integrated gates [66].
Figure 4.6 4WJ DNA logic gates and tile‐integrated DNA circuits. (a) 4WJ NOT gate: a DNA strand (NOT) holds the opened MB probe in the absence of an input. Addition of the DNA input decomposes the complex, thus releasing the MB probe. (b) Two‐input 4WJ AND gate. The gate consists of ANDa, ANDb, ANDc, and an MB probe. The five‐stranded 4WJ association is formed only in the presence of both inputs I1 and I2. (c) Two‐input 4WJ NOR gate integrated into a DNA tile. NOT1, NOT2, ANDa, and ANDb strands are attached to a DNA crossover (X) tile at the indicated points. In the absence of inputs, the 5′‐terminal NOT1 and 3′‐terminal NOT2 output fragments are bound to the input‐recognition fragments of the ANDa and ANDb strands, respectively, which enables formation of the high‐signal (fluorescent) 4WJ NOR association. Addition of inputs i3 and/or i4, which are complementary to the input‐recognition fragments of NOT1 and NOT2, respectively, results in dissociation of the 4WJ NOR complex. The gate performs as predicted according to the observed changes in fluorescent output (lower right corner).
To facilitate inter‐gate communication and reduce the undesirable crosstalk, we attempted to confine the gates in a nano‐environment by attaching them to a DNA scaffold in proper orientation and near their direct communication partners [27]. For example, Figure 4.6c demonstrates integration of two NOT with an AND gate to produce a NOR logic: the two NOT gates feed their outputs to a 2iAND. The NOT and AND strands form a fluorescent complex with an MB probe, while addition of at least one input decomposes one of the NOT gates followed by the collapse of the entire complex into the separate strands. The obtained circuits were shown to be reusable multiple times if RNA inputs are used in the presence of RNase H as a buffer component. As the case for strand displacement‐based logic gates, a universal powering method is needed to build long chains of communicating logic gates.
4.5 Conclusion
Majority of the DNA logic gates, however, explore only two to five layers of integration, which faces significant signal reduction as the signal propagates along the chain of communicating gates. At least partially, this problem can be mitigated by localizing logic gates in a specific order and at precise positions on a DNA tile for efficient communication as it is used in electronic processors. An energy input is required to “push” the signal through the DNA association, an approach that has not been realized yet. Alternatively, parallel computation using multiple small‐scale integrated circuits can be explored. While all the technical problems can be eventually addressed given the appropriate time and effort, the future of molecular computation depends on the practical usability of DNA computers. Indeed, it becomes clear that computers based on hybridization of DNA strands cannot compete with electronic devices in terms of the processing speed due to much slower rates of DNA hybridization than electron transfer in semiconductor materials. Instead, biocompatible and biodegradable DNA‐based logic constructs can be used for manipulating biological molecules and objects (cells), which can eventually find applications in addressing biological and biomedical problems.
References
1 1 Malvino, A.P. and Brown, J.A. (1993). Digital Computer Electronics, 3e. Lake Forest: Glencoe.
2 2 de Silva, A.P. and Uchiyama, S. (2007). Nat. Nanotechnol. 2: 399–410.
3 3 de Silva, A.P., Leydet, Y., Lincheneau, C., and McClenaghan, N.D. (2006). J. Phys. Condens. Matter 18: S1847–S1872.
4 4 Ball, P. (2000). Nature 406: 118–120.
5 5 Katz, E. (2017). Anal. Bioanal. Chem. 409: 81–94.
6 6 Katz, E. (2015). Curr. Opin. Biotechnol. 34: 202–208.
7 7 Schneider, H.J. (2017). ChemPhysChem 18: 2306–2313.
8 8 Benenson, Y. (2016). Chimia (Aarau) 70: 392–394.
9 9 Erbas‐Cakmak, S., Kolemen, S., Sedgwick, A.C. et al. (2018). Chem. Soc. Rev. 47: 2228–2248.
10 10 Adleman, L.M. (1994). Science 266: 1021–1024.
11 11 Stojanovic, M.N., Stefanovic, D., and Rudchenko, S. (2014). Acc. Chem. Res. 47: 1845–1852.
12 12 Fu, T., Lyu, Y., Liu, H. et al. (2018). Trends Biochem. Sci. 43: 547–560.
13 13 Ariga, K., Nishikawa, M., Mori, T. et al. (2019). Sci. Technol. Adv. Mater. 20: 51–95.
14 14 Stojanovic, M.N., Mitchell, T.E., and Stefanovic, D. (2002). J. Am. Chem. Soc. 124: 3555–3561.
15 15 Saghatelian, A., Völcker, N.H., Guckian, K.M. et al. (2003). J. Am. Chem. Soc. 125: 346–347.
16 16 Okamoto, A., Tanaka, K., and Saito, I. (2004). J. Am. Chem. Soc. 126: 9458–9463.
17 17 Yoshida, W. and Yokobayashi, Y. (2007). Chem. Commun. 14: 195–197.
18 18 Penchovsky, R. and Breaker, R.R. (2005). Nat. Biotechnol. 23: 1424–1433.
19 19 He, H.Z., Chan, D.S., Leung, C.H., and Ma, D.L. (2013). Nucleic Acids Res. 41: 4345–4359.
20 20 Kahan‐Hanum, M., Douek, Y., Adar, R., and Shapiro, E. (2013). Sci. Rep. 3: 1535.
21 21 Li, T., Lohmann, F., and Famulok, M. (2014). Nat. Commun. 5: 4940.
22 22 Guo, Y., Zhou, L., Xu, L. et al. (2014). Sci. Rep. 4: 7315.
23 23 Fan, D., Wang, K., Zhu, J. et al. (2015). Chem. Sci. 6: 1973–1978.
24 24 Green, A.A., Kim, J., Ma, D. et al. (2017). Nature 548: 117–121.
25 25 Gao, J., Liu, Y., Lin, X. et al. (2017). Sci. Rep. 7: 14014.
26 26 He, K., Li, Y., Xiang, B. et al. (2015).