Self-organization controls expression more than abundance of molecular components of transcription and translation in confined cell-free gene expression

Abstract

Cell-free gene expression using purified components or cell extracts has become an important platform for synthetic biology that is finding a growing number of practical applications. Unfortunately, at cell-relevant reactor volumes, cell-free expression suffers from excessive variability (noise) such that protein concentrations may vary by more than an order of magnitude across a population of identically constructed reaction chambers. Consensus opinion holds that variability in expression is due to the stochastic distribution of expression resources (DNA, RNAP, ribosomes, etc.) across the population of reaction chambers. In contrast, here we find that chamber-to-chamber variation in the expression efficiency generates the large variability in protein production. Through analysis and modeling, we show that chambers self-organize into expression centers that control expression efficiency. Chambers that organize into many centers, each having relatively few expression resources, exhibit high expression efficiency. Conversely, chambers that organize into just a few centers where each center has an abundance of resources, exhibit low expression efficiency. A particularly surprising finding is that diluting expression resources reduces the chamber-to-chamber variation in protein production. Chambers with dilute pools of expression resources exhibit higher expression efficiency and lower expression noise than those with more concentrated expression resources. In addition to demonstrating the means to tune expression noise, these results demonstrate that in cell-free systems, self-organization may exert even more influence over expression than the abundance of the molecular components of transcription and translation. These observations in cell-free platform may elucidate how self-organized, membrane-less structures emerge and function in cells.

Synergistic interactions between confinement and macromolecular crowding spatially order transcription and translation in cell-free expression

“…Transcriptional and total expression transients were extracted from individual vesicles using custom MATLAB code for image processing. The expression noise was extracted from these transients using a protocol described in previous work 6,9,12,29 ( In contrast to either confinement 6 or crowding 9 alone, the shape and timing of the transcriptional transient response varied significantly as crowding was increased in a confined environment ( Fig. 3A).…”

“…2 burst size as the intensity of expression within the centers 12 . There is evidence of expression patterns indicative of these distinct expression centers even without crowding 12 , but with crowding these centers (at least at the transcriptional level) are visible using optical microscopy 8,9 . The transcriptional and total expression burst sizes ( = (also known as the Fano factor); Fig.…”

“…Numerous studies in confined and crowded cellular environment demonstrate how the gene expression process self-organizes into spatial subregions [12][13][14] .…”

“…This self-organized, membrane-less structure in cells creates heterogeneous environments of crowding and confinement that control the sharing of gene expression resources and tune the patterns of expression bursting 13,14,[17][18][19] . Cell-free systems can mimic some physical features of cells 6,9,12,20,21 , and crowding studies lacking cellrelevant confinement show some spatial organization of transcription 8,9 . Yet, more fully mimicking cell-like self-organization has been elusive.…”

“…Two experiments were performed for each Ficoll-70 crowder concentration of 0, 10, 40, 60, and 90 mg/mL. For an individual vesicle in each experiment, a 6-hour mRNA and protein expression trace was extracted for noise analysis, as described in previous work 6,9,12,30 .…”