Morphogenesis in molecular systems

We take inspiration from early embryo development to engineer synthetic biochemical systems that are capable of spatio-temporal self-organization. 

The objective of our biomimetic approach is two-fold. On the one hand, by studying simple, physics-like and controllable molecular systems that retain the essential features of their biological analogues we hope to provide new insights on the emergence of complex biological behaviors -such as gene regulation, and morphogenesis. On the other hand, these dynamic molecular systems can be regarded as a new kind of "life-like" materials capable of adapting and responding autonomously to their environment. To this end we essentially use systems based on nucleic acid hybridization reactions because their reactivity can be easily predicted by, roughly, Watson-Crick pairing rules. These systems self-organize in space and time through two archetypal mechanisms: reaction-diffusion and active matter.

A DNA concentration gradient inside a microchannel (in blue) is interpreted by a molecular program and creates a 3-band self-organized chemical pattern. First in solution, this pattern is later materialized with the help of DNA-decorated colloids (bottom).


More precisely we work with three experimental systems that self-organize in space and time through different mechanisms and at different scales. 1) The PEN DNA toolbox, a molecular programming language based on short DNA strands and three enzymes that can be maintained out of equilibrium for tens of hours and self-organize into various concentration patterns through a reaction-diffusion mechanism that we can finely control. 2) Active matter composed of cyostkeletal motors and filaments. 3) Cell cultures that interact with DNA strands.


  • Demonstration of reaction-diffusion traveling waves and spirals in a programmable DNA-based chemical system (Padirac et al, JACS 2013). Cover of JACS.
  • Demonstration of chemical fronts that counter-propagate with minimal interaction (Zadorin et al, PRL, 2015). Selected as a PRL editor highlight in Physics. Highlighted in Nature nanotech and in Chemistry world.
  • Synthetic realization of a "French flag" concentration pattern, an archetypal model in Developmental Biology (Zadorin et al, Nature chem. 2017).

Future directions

Project 1. Metabolic soft matter with life-like properties: We take a synthetic, chemical approach and combine self-organized patterns with materials to create "life-like" materials capable of adapting and responding autonomously to their environment. These'metabolic', or 'animated' materials are constructed by combining a stimuli-responsive material with an active biochemical solution, i.e. a chemical reaction network maintained out of equilibrium. We will take advantage of DNA/enzyme active solutions to lay the foundation of materials able to emulate some of the fascinating dynamic abilities of living matter: processing complex chemical information, generating forces under the control of chemical stimuli and self-organizing into a reproducible final shape. This project has been funded by an ERC consolidator grant.

Project 2. Synthetic chemical patterns as a minimal model of pattern formation in development: The objective here is to use a model in vitro system to try to shed light into the molecular patterning process taking place during early embryo development. To do so we will use an artificial experimental model of gradient-based pattern formation that we have recently developed, based on DNA reaction networks (Zadorin et al, 2017). We will ask the following question: can a purely reaction-diffusion mechanism help us understand the formation of the sharp concentration patterns observed in the Drosophila blastoderm? In particular, we will probe experimentally for the influence of stochasticity in molecular pattern-formation.