Engineering a riboswitch-based genetic platform for the self-directed evolution of acid-tolerant phenotypes

Abstract

Environmental pH is a fundamental signal continuously directing the metabolism and behavior of living cells. Programming the precise cellular response toward environmental pH is, therefore, crucial for engineering cells for increasingly sophisticated functions. Herein, we engineer a set of riboswitch-based pH-sensing genetic devices to enable the control of gene expression according to differential environmental pH. We next develop a digital pH-sensing system to utilize the analogue-sensing behavior of these devices for high-resolution recording of host cell exposure to discrete external pH levels. The application of this digital pH-sensing system is demonstrated in a genetic program that autonomously regulated the evolutionary engineering of host cells for improved tolerance to a broad spectrum of organic acids, a valuable phenotype for metabolic engineering and bioremediation applications.Cells are exposed to shifts in environmental pH, which direct their metabolism and behavior. Here the authors design pH-sensing riboswitches to create a gene expression program, digitalize the system to respond to a narrow pH range and apply it to evolve host cells with improved tolerance to a variety of organic acids.

Fig. 1

Engineering the dynamic range of a wild-type pH-riboswitch. a Mechanism of action of the wild-type pH-riboswitch. During transcription, the wild-type pH-riboswitch adopts distinct folding conformations to affect the RNAP-dependent mRNA synthesis process. At low pH (pHe 5.0), mature mRNA is produced via the RNAP non-pausing pathway to yield translationally inactive transcripts with the ribosome-binding site locked by its complementary sequence. At high pH (pHe 8.0), mRNA is produced via the RNAP pausing pathway to yield translationally active transcripts with an accessible ribosome-binding site to allow translation. b Detailed folding conformations of inactive (left) and active (right) states of the wild-type pH-riboswitch. The putative wild-type ribosome-binding sequence (AGGA) is boxed. c Schematic of the genetic construct used to characterize the dynamic range of the wild-type and synthetic pH-riboswitches. d Engineering strategy to tune the dynamic range of the wild-type pH-riboswitch. The weak ribosome-binding site and its complementary sequence within the wild-type pH-riboswitch (PRE) were replaced by those of a strong ribosome-binding site RBS34 (BBa_0034) to generate the synthetic variant PREmR34. A library of PREmR34 variants (PREmR34.x) with diverse dynamic ranges was generated by varying the bases surrounding the “core” RBS sequence of PREmR34 (GAGGAG). e Output fluorescence at varied extracellular pH (pHe) of the pH-sensing genetic devices (l-ara 0.02% m/v). f Correlation between the fluorescence output and relative strength of the RBSs embedded in the pH-riboswitch variants. The relative RBS strengths were normalized against RBS34, and the relative outputs were normalized against the output of the pH-sensing device built from RBS34. The data represent the mean of three independent experiments performed on different days

Fig. 2

Alleviation of the background leakiness of the pH-sensing device through transducer engineering. a Schematic of transducer engineering strategy involving the generation of T7 promoter mutants (pT7mt) and T7RNAP mutants (T7 RNAPmt). b Tuning the basal expression of the pH-sensing device through T7 promoter engineering. c Tuning the basal expression of the pH-sensing device through T7RNAP engineering. d Tuning the basal expression of the pH-sensing device through titration of the intracellular T7RNAP concentration. The heat map shows the graded fluorescence output of the pH-sensing device controlled by the mutant FQ under varying l-arabinose induction (bottom to top: 0, 0.001, 0.002, 0.008, 0.02, 0.1% m/v) and extracellular pH (pHe) conditions (left to right: 5.5, 6.0, 6.5, 7.0, 7.5, 8.0). e Comparing the output of the pH-sensing devices controlled by wild-type T7RNAP and the mutant FQ at l-arabinose induction 0.002% m/v. The data represent the mean of three independent experiments performed on different days

Fig. 3

Characterization of the digitalized pH-sensing system. a Schematic of the digitalized pH-sensing system. The system regulates the pH-dependent expression of an integrase (int2) that catalyzes the unidirectional reorientation of a strong constitutive promoter (BBa_J23119) flanked by its recognition sequences. At acidic pH (pHe 5.0), integrase expression is suppressed, and cells are maintained in the “OFF” state (no fluorescence). At neutral pH (pHe 7.5), integrase expression leads to the reorientation of the promoter J23119 to drive RFP expression, and cells are switched to the “ON” state (high fluorescence). b Outputs of the digital pH-sensing systems constructed from the T7 promoter variants. c Outputs of the digitalized pH-sensing systems constructed from the T7RNAP variants. d Time-course characterization of the digitalized pH-sensing system. Top: The histograms show the fluorescence distribution of the cell population induced at pHe 5.0 or 7.5 over time. The shaded regions indicate the “ON” state (red fluorescence > 10^3.8 au). Bottom: Gel electrophoresis of PCR products corresponding to the “ON” (0.85 kb) and “OFF” (1.90 kb) states of the cell population induced at pHe 5.0 or 7.5 over time. e Tuning of the transition band of the digital pH-sensing system through titration of the intracellular T7RNAP concentration. Low-threshold (filled circles) and high-threshold (open circles) transition bands were obtained at near-saturated (0.1% m/v) and lower induction concentrations (0.002% m/v) of l-arabinose, respectively. The shaded regions indicate the transition regions. The dashed lines indicate 10% and 85% population switching. f Stability assessment of the digitalized pH-sensing system at pHe 5.0 (blue) and pHe 7.5 (red) (with 0.002% m/v l-arabinose) over long periods. The switching behavior of the system was monitored daily over a 21-day period. The data represent the mean of three independent experiments performed on different days

Fig. 4

Programming autonomous evolution of acid-tolerant phenotypes. a Schematic of a genetic platform that enables the self-directed evolution of an acid-tolerant phenotype (RiDE). b Conceptual design of RiDE. A core pH-sensing module was employed to determine the single-cell pHi. In the default configuration, RiDE sets host cells to a rapidly evolving state and generates genetic heterogeneity among the progeny population. Mutants that evolve to stably maintain the pHi near neutrality activate integrase expression, which then inverts the J23119 promoter to turn OFF in vivo mutagenesis and turn ON RFP expression. c Flow analysis of RiDE culture after 36 h of growth. A small population of a highly fluorescent evolved phenotype is boxed in red. d High-throughput experimental workflow to isolate acid-tolerant phenotypes from RiDE-based evolution experiment. For each of 16 combinations of pHe (4.0, 4.5, 4.8, 5.0) and l-arabinose (0.0001%, 0.001%, 0.002%, 0.02% m/v), three biological replicates were prepared to form a total of 48 parallel cultures. As negative controls, cells transformed with digitalized pH-sensing system (without ednaQ expression) were used to prepare another 48 parallel cultures with similar pHe and l-arabinose conditions. RFP expressions in the parallel cultures were monitored by a robotic flow analyzer every 12 h. The highly fluorescent population was sorted, propagated, and subjected to a high-throughput growth assay at acidic pH (pHe 4.5). e Growth kinetics and specific growth rates of acid-tolerant strains at pHe 4.5. The data represent the mean of three biological replicates

Fig. 5

Phenotype and genotype comparison of the evolved strains AT01, AT02, and wild-type T10. a Growth rates of AT01, AT02, T10 at varied pHe. b Growth profiles of AT01, AT02, T10 in M9 medium at pHe 4.20. c Culture turbidity of AT01, AT02, T10 grown in M9 medium at pHe 4.20 for 18 h. d pHi of AT01, AT02, T10 at different pHe. Student t-test: asterix = P < 0.05. e Mutations in the AT01 genome compared to the DH10B reference. The middle blue ring represents coding sequences (CDSs) in the reference DH10B genome. The arcs in the inner ring represent single-nucleotide polymorphisms (SNPs) in AT01 compared to the reference. The arcs in the outer ring show coding sequences (ykfH, ykfF, ypjJ) with insertion or deletion mutations compared to the reference. f List of modified CDSs in AT01. The data represent the mean of three biological replicates

Fig. 6

Tolerance of AT01, AT02, and T10 against industrial organic acids. a Survival percentage of AT01, AT02, and T10 cultures growth at varied organic acid concentrations at (37 °C, 12 h) are reported. For each organic acid, the stationery-phase cell density at each concentration were divided against maximal cell density grown in neutral media (pHe 7.0). b Comparison of intracellular pH (pHi) of AT01 and T10 under organic acids challenge (25 mM). c Comparison of intracellular pH (pHi) of AT02 and T10 under organic acids (25 mM). The pHi values are measured 6 h upon organic acid addition. Student t-test: asterix = P < 0.05, double asterix = P < 0.01. The data represent the mean of three biological replicates