The primordial life of DNA dynamic networks

DNA-based dynamic networks show adaptation to external stimuli toward the generation of the fittest constituent. This selection principle has now been implemented to control the catalytic efficiency of an enzymatic reaction.

Physics strives to reveal the laws of the universe. Biology aims at grasping the secrets of life. However, while the former conveys the general rules that govern both inanimate and animate things, overlooking their fundamental difference, the latter navigates its own way in the complexity of living matter without being able to explain its primordial origin. How can we fill this gap? How can we weave the logic thread that goes from inanimate to animate matter? From non-living to living beings?

As envisioned by J. M. Lehn more than 10 years ago1, a possible answer to these challenging questions comes from chemistry, particularly ‘systems chemistry’, the field that deals with the special ability of some molecules to self-organize into systems with emergent properties. Such systems are constituted by a network of chemical species connected by reversible reactions that set them into antagonistic or agonistic relationships depending on whether or not they share a component. The simplest case is that of four components (A, B, A′ and B′) that organize themselves into four distinct species (or constituents), namely AA′, AB′, BA′ and BB′ (Fig. 1). At thermodynamic equilibrium, the system is described by a defined partition of all species (Fig. 1b). As these are interconnected to each other, perturbation of one of the components of the network by a physical stimulus or a chemical effector will lead to a redistribution of the species, until a new equilibrium state is reached. Thus, for example, an increase (or decrease) in the population of AA′ will drive the down- or up-regulation of the directly connected species — that is, the antagonists AB′ and BA′ — that are competing for the same resource (Fig. 1c and a, respectively). At the same time, the agonistic species BB′ that is not directly connected to the target of change will follow the same destiny as AA′, that is, it will be free to grow (or committed to vanish). Such constitutional dynamic networks (CDNs) can thus show us that initial logic thread in the form of self-organizing systems capable of sensing a signal from the environment, responding to it and adapting to a new condition. By promoting the survival of the fittest at the expense of the least fit, CDNs obey the universal Darwinian principle that regulates life on Earth and embody the feature that makes living matter so special, that is, the ability to evolve.

Fig. 1: Orthogonal operation of biocatalytic cascades guided by coupled DNA networks.

ac, The GOx/HRP CDN is composed of four constituents: GOx/HRP-AA′, GOx-AB′, BA′-HRP and BB′. The spatial proximity of the two enzymes in the constituent AA′ allows the operation of the cascade, leading to the oxidation of substrate (S) into product (P). Treatment of the CDN with the T1 trigger enhances the activity of the GOx/HRP cascade (c), whereas subsequent addition of the counter-trigger T1′ brings the system back to a moderate catalytic activity (b). In a similar fashion, subjecting the network to the T2/T2′ strands induces the reversible switching of the network between a moderate (b) and a low activity state (a). df, The same strategy is applied for the ADH/NAD+ cascade, whose catalytic activity state is regulated by the addition of T3/T3′ and T4/T4′ triggers. The two enzyme networks communicate through hairpin motifs (H1 and H2) that specifically establish connections between two otherwise unrelated species of the two networks. Whereas H1 connects CC′ of the second network to AA′ of the first network (path 1), H2 connects BB′ of the first network to DD′ of the second network (path 2), enabling orthogonal operation of the two cascades in a fully programmable fashion.

Despite its undoubtful appeal, the implementation of this concept has been limited to structural transformations that, besides being coupled to some sort of read-out signal, are not truly operating on the external environment, and in this sense, are devoid of a proper function. Now, writing in Nature Catalysis, Itamar Willner and co-workers2 overcome this limit. Using nucleic acid components functionalized with glucose oxidase (GOx) and horseradish peroxidase (HRP), the researchers generate a DNA-based dynamic network, whose reversible triggered reconfiguration results in the switchable up- and down-regulation of the GOx/HRP catalytic cascade (Fig. 1). Similar to the four-component-model initially developed by Lehn, the system described here consists of four nucleic acid components: two of them are conjugated to one single enzyme of the cascade (GOx-A or HRP-A′) while the other two are enzyme-free (B and B′). At equilibrium, the reaction mixture contains four species: GOx/HRP-AA′, GOx-AB′, BA′-HRP and BB′ (Fig. 1b). The spatial proximity of the two enzymes in the constituent AA′ allows operation of the cascade, leading to oxidation of ABTS2– into ABTS•– or dopamine into aminochrome, in the presence of glucose and oxygen. In addition, the DNA network displays two important features. First, each constituent includes a conjugated Mg2+-ion-dependent DNAzyme unit that acts as a reporter to quantify the content of the corresponding species in solution. Second, each constituent displays a distinct loop domain whose sequences are designed such that, in the presence of the auxiliary trigger T1 (or T2), the species AA′ (or BA′), is stabilized into a triplex structure. Thus, whereas the magnesium-dependent domain enables to associate the efficiency of the biocatalytic cascade to the content of each species, the loop is the essential property that makes the switching behaviour possible.

Treatment of the CDN with the T1 trigger enhances the activity of the GOx/HRP cascade, whereas subsequent addition of the counter-trigger T1′ brings the system back to a moderate catalytic activity (Fig. 1c). In a similar fashion, subjecting the network to the T2/T2′ strands induces the reversible switching of the network between a moderate and a low activity state (Fig. 1a). The final scenario is that of a GOx/HRP cascade that can explore three distinct states of catalytic performance in a step-wise manner, with the lowest and highest activity states characterized by opposite distributions of the two paired species (AA′/BB′ and AB′/BA′). The general applicability of this strategy is exemplified by the generation of a second network that switches the activity of a different enzyme cascade, composed of the alcohol dehydrogenase (ADH) and the nicotinamide adenine dinucleotide cofactor (NAD+). Again, three distinct states can be accessed in a fully reversible and programmable fashion (Fig. 1d–f). Willner and co-workers finally couple the two enzyme networks through the smart design of hairpin motifs that enable the connection of two otherwise independent systems. The final result is the orthogonal operation of two biocatalytic cascades driven by the induced intercommunication of two CDNs in an elegant synchronization of eight different species.

Previous studies from the Willner group have proven the suitability of the DNA molecule for the implementation of CDNs3,4,5,6,7. The self-recognition properties of the DNA, together with the non-covalent nature of its self-assembly, make this molecule ideal to construct the constituents of dynamic networks, allowing them to program both the structure and the dynamics of each species. However, the present study adds a fundamental feature to previous DNA-based dynamic networks, namely, the coupling of the switch to an enzymatic reaction. The information encoded in the structure of the constituents and their connectivity is not only transferred from one state to the other but is also associated to an action, a work that the system performs on the external environment, altering it in a permanent fashion.

Now, let´s consider whether these systems are really suitable to address our initial questions. Clearly, DNA-driven CDNs, until now, operate at equilibrium conditions and essentially rely on the thermodynamically controlled distribution of constituents in a given set of conditions. Life instead is a far-from-equilibrium process, a ceaseless transformation of energy and matter that drives living beings to higher forms of self-organization at the expense of a greater entropic gap between the order inside the system and the disorder outside8. The search for the mechanism that enabled the passage from self-organization to emergent complexity, emulating a true evolutionary process through increasingly fittest species, necessarily requires a change of paradigm, from thermodynamic to kinetic control of the network, from equilibrium to out-of-equilibrium conditions, from reversible to irreversible changes. Only in this way may we be able to emulate a primordial life-like process and try to answer what is probably one of the most intriguing and fascinating questions of humankind.

Considerable efforts have already been made along this direction. Recent studies demonstrate how the DNA molecule may help us in this endeavour9, enabling scientists to construct, reshape and link structural modules into systems of increasing irreversibility grade and complexity, thus showing us — once again — the enormous possibilities held in the simplicity of the Watson–Crick base-pairing rule. The challenge is certainly ambitious but we might be nearer than ever.


  1. 1.

    Lehn, J. M. Chem. Soc. Rev. 36, 151–60 (2007).

    CAS  Article  Google Scholar 

  2. 2.

    Wang, C., Yue, L. & Willner, I. Nat. Catal. (2020).

  3. 3.

    Wang, S. et al. J. Am. Chem. Soc. 139, 9662–9671 (2017).

    CAS  Article  Google Scholar 

  4. 4.

    Wang, S. et al. Angew. Chem. Int. Ed. 57, 8105–8109 (2018).

    CAS  Article  Google Scholar 

  5. 5.

    Yue, L. et al. J. Am. Chem. Soc. 140, 8721–8731 (2018).

    CAS  Article  Google Scholar 

  6. 6.

    Zhou, Z., Yue, L., Wang, S., Lehn, J. M. & Willner, I. J. Am. Chem. Soc. 140, 12077–12089 (2018).

    CAS  Article  Google Scholar 

  7. 7.

    Yue, L., Wang, S., Wulf, V. & Willner, I. Nat. Commun. 10, 4774 (2019).

    Article  Google Scholar 

  8. 8.

    Schwille, P. Angew. Chem. Int. Ed. 56, 10998–11002 (2017).

    CAS  Article  Google Scholar 

  9. 9.

    Green, L. N. et al. Nat. Chem. 11, 510–520 (2019).

    CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Barbara Saccà.

Ethics declarations

Competing interests

The authors declare no competing interests.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Winat, L.J., Saccà, B. The primordial life of DNA dynamic networks. Nat Catal 3, 865–866 (2020).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing