Accurate characterization of dynamic microbial gene expression and growth rate profiles

Abstract

Genetic circuits are subject to variability due to cellular and compositional contexts. Cells face changing internal states and environments, the cellular context, to which they sense and respond by changing their gene expression and growth rates. Furthermore, each gene in a genetic circuit operates in a compositional context of genes which may interact with each other and the host cell in complex ways. The context of genetic circuits can therefore change gene expression and growth rates, and measuring their dynamics is essential to understanding natural and synthetic regulatory networks that give rise to functional phenotypes. However, reconstruction of microbial gene expression and growth rate profiles from typical noisy measurements of cell populations is difficult due to the effects of noise at low cell densities among other factors. We present here a method for estimation of dynamic microbial gene expression rates and growth rates from noisy measurement data. Compared to the current state-of-the-art, our method significantly reduced the mean squared error of reconstructions from simulated data of growth and gene expression rates, improving the estimation of timing and magnitude of relevant shapes of profiles. We applied our method to characterize a triple-reporter plasmid library combining multiple transcription units in different compositional and cellular contexts in Escherichia coli. Our analysis reveals cellular and compositional context effects on microbial growth and gene expression rate dynamics, and suggests a method for dynamic ratiometric characterization of constitutive promoters relative to an in vivo reference.

Towards automation of the Design-Build-Test-Learn (DBTL) bioengineering cycle: Application to the testing and characterization of standard bioparts

“…There is work being done towards a fully automated DBTL cycle tackling one of the steps at the time; mainly for the automation of the design step (Buecherl and Myers, 2022;Radivojević et al, 2020;Vidal et al, 2022b;Boada et al, 2021Boada et al, , 2022a, build step (Ko et al, 2022;Kang et al, 2022;Bryant Jr et al, 2023), for the test step there are advances in the calibration of the measurements (Beal et al, 2022;González-Cebrián et al, 2023), and in the automation of wetlab protocols in general (Vidal-Peña, 2023). Finally for the learn step (Vidal et al, 2022a;Yanez Feliu et al, 2020;Boada et al, 2019b). However, there are not many examples of automation of the test and learn steps combined.…”