Undesired usage and the robust self-assembly of heterogeneous structures

Arvind Murugan, James Zou, Michael P. BrennerNature Communications 2015 · DOI

Abstract

Inspired by multiprotein complexes in biology and recent successes in synthetic DNA tile and colloidal self-assembly, we study the spontaneous assembly of structures made of many kinds of components. The major challenge with achieving high assembly yield is eliminating incomplete or incorrectly bound structures. Here, we find that such undesired structures rapidly degrade yield with increasing structural size and complexity in diverse models of assembly, if component concentrations reflect the composition (that is, stoichiometry) of the desired structure. But this yield catastrophe can be mitigated by using highly non-stoichiometric concentrations. Our results support a general principle of 'undesired usage'-concentrations of components should be chosen to account for how they are 'used' by undesired structures and not just by the desired structure. This principle could improve synthetic assembly methods, but also raises new questions about expression levels of proteins that form biological complexes such as the ribosome.

The Challenge of Heterogeneous Self-Assembly

Self-assembly is a fundamental process where components spontaneously organize into functional structures. However, as we attempt to engineer larger and more heterogeneous structures—those composed of many distinct part types—we encounter a "yield catastrophe." As complexity increases, the number of potential incorrect or incomplete "junk" structures grows exponentially, rapidly depleting the pool of available components.

Traditional assembly strategies often rely on the "Dosage Balance Hypothesis," which suggests that components should be provided in the exact ratios (stoichiometry) found in the final target structure. Murugan et al. demonstrate that this stoichiometric approach is often sub-optimal because it inadvertently provides the ideal environment for undesired intermediates to flourish.

The Principle of Undesired Usage

The authors propose the principle of undesired usage: component concentrations should be adjusted based on how they are consumed by "junk" structures, not just the final product. If an undesired structure consumes component A more frequently than component B, the concentration of A should be reduced to suppress that specific pathway, even if the final structure requires equal amounts of both.

Models of branched structures (a) and colloidal clusters (b) used to validate the undesired usage principle.

Methods and Benchmarking

The study employs a combination of statistical mechanics and kinetic simulations to explore these effects:

  • Equilibrium Modeling: Using perturbative Feynman diagram techniques to sum the partition functions of vast numbers of possible undesired structures.
  • Kinetic Simulations: Tracking component depletion over time in irreversible assembly processes, such as DNA brick or colloidal systems.

These models were tested against several structural benchmarks, including linear chains, branched "millipede" shapes, and colloidal clusters.

Results: DNA Bricks and Colloidal Validation

The most striking evidence for the undesired usage principle comes from complex synthetic structures. For a 208-tile "P-shape" modeled after DNA brick experiments, the authors found that a highly non-uniform concentration profile—starving the system of parts prone to forming junk—resulted in a 15-fold increase in assembly yield compared to standard stoichiometric ratios.

The 208-tile 'P-shape' DNA brick structure (c) and its corresponding tile map. Non-stoichiometric optimization significantly boosts yield for such complex targets.

Furthermore, the model successfully explains experimental observations in colloidal assembly, where an 8:1 ratio was required to achieve high yields of $AB_4$ clusters, despite the 4:1 stoichiometry of the target. By optimizing concentrations based on undesired usage, the authors restored yield to nearly 100% in linear structures where stoichiometric yields had dropped to near zero.

Discussion and Biological Implications

These findings challenge biological dogma regarding protein expression levels. The varying concentrations of proteins in a cell may not be mere noise, but a sophisticated control mechanism to prevent the formation of toxic or wasteful aggregates during the assembly of massive complexes like the ribosome. While applying this principle to new synthetic systems requires a detailed understanding of assembly pathways, it provides a robust framework for engineering the next generation of complex nanostructures.

Summary updated

Abstract

Inspired by multiprotein complexes in biology and recent successes in synthetic DNA tile and colloidal self-assembly, we study the spontaneous assembly of structures made of many kinds of components. The major challenge with achieving high assembly yield is eliminating incomplete or incorrectly bound structures. Here, we find that such undesired structures rapidly degrade yield with increasing structural size and complexity in diverse models of assembly, if component concentrations reflect the composition (that is, stoichiometry) of the desired structure. But this yield catastrophe can be mitigated by using highly non-stoichiometric concentrations. Our results support a general principle of 'undesired usage'-concentrations of components should be chosen to account for how they are 'used' by undesired structures and not just by the desired structure. This principle could improve synthetic assembly methods, but also raises new questions about expression levels of proteins that form biological complexes such as the ribosome.
BibTeX
@inproceedings{murugan2015undesired,
  author = {Arvind Murugan and James Zou and Michael P. Brenner},
  title = {{Undesired usage and the robust self-assembly of heterogeneous structures}},
  booktitle = {Nature Communications},
  year = {2015},
  doi = {10.1038/ncomms7203},
}

Loading related content...

Figure