Dual genetic level modification engineering accelerate genome evolution of Corynebacterium glutamicum

Abstract

High spontaneous mutation rate is crucial for obtaining ideal phenotype and exploring the relationship between genes and phenotype. How to break the genetic stability of organisms and increase the mutation frequency has become a research hotspot. Here, we present a practical and controllable evolutionary tool (oMut-Cgts) based on dual genetic level modification engineering for Corynebacterium glutamicum. Firstly, the modification engineering of transcription and replication levels based on RNA polymerase α subunit and DNA helicase Cgl0854 as the 'dock' of cytidine deaminase (pmCDA1) significantly increased the mutation rate, proving that the localization of pmCDA1 around transient ssDNA is necessary for genome mutation. Then, the combined modification and optimization of engineering at dual genetic level achieved 1.02 × 104-fold increased mutation rate. The genome sequencing revealed that the oMut-Cgts perform uniform and efficient C:G→T:A transitions on a genome-wide scale. Furthermore, oMut-Cgts-mediated rapid evolution of C. glutamicum with stress (acid, oxidative and ethanol) tolerance proved that the tool has powerful functions in multi-dimensional biological engineering (rapid phenotype evolution, gene function mining and protein evolution). The strategies for rapid genome evolution provided in this study are expected to be applicable to a variety of applications in all prokaryotic cells.

Graphical Abstract

Figure 1.

Modification engineering at the transcription level increased spontaneous mutation rates in C. glutamicum. (A) Schematic illustration of modification engineering at the transcription level mediated random genome editing. pmCDA1 was fused to the RNAP subunit via a flexible linker. The RNAP locates pmCDA1 to ssDNA (purple line) which was transiently generated during transcription. Subsequently, C:G→T:A base conversion (red star) was introduced by pmCDA1. linker, a 16-residue flexible linker, amino acid sequence was provided in the supplementary data. (B) Rifampicin resistance assay for screening for the most effective RNAP subunits and fused approaches. WT, wild-type strain. (C) Mechanism of C:G→T:A base conversion catalyzed by pmCDA1. AP, apurinic/apyrimidinic; BER, DNA base excision repair. (D) Schematic illustration of the complex forms of RNAP, pmCDA1, and UGI examined in this study. Mut-α0, empty vector control; Mut-α1, independent pmCDA1; Mut-α2, pmCDA1-α fusion; Mut-α3, independent UGI; Mut-α4, independent pmCDA1 and UGI; Mut-α5, pmCDA1-UGI fusion; Mut-α6, pmCDA1-α fusion and independent UGI; Mut-α7, UGI-pmCDA1-α fusion. (E) Mutation rate of rifampicin resistance assay of C. glutamicum cells expressing each molecular device. (F) Rifampicin resistance phenotype of C. glutamicum cells expressing each molecular device. Cells with mutagenic plasmids were induced with 0.5 mM IPTG at 30°C, 200 rpm. After 24 h, 100ul cultures were plated on BHI + 8 μg/ml rifampin plate, and then cultured in 30°C for the same time. ‘ + rif ’, rifampicin were added in the medium. Data are presented as dot plots with mean ± standard deviation (SD) (n = 5). * P < 0.05, ** P < 0.01, *** P < 0.001 (Student's two-tailed t-test).

Figure 2.

Modification engineering at the replication level accelerate genome mutation. (A) Schematic representation of helicase-based modification engineering at the replication level mediated random genome mutation. pmCDA1 was fused to the DNA helicases via a 100-residue flexible linker. During replication, DNA helicase breaks hydrogen bonds to produce temporary ssDNA, and at the same time, pmCDA1 located to ssDNA catalyzes cytosine deamination to cause C:G→T:A base conversion (red star) in genome. The amino acid sequence of 100-residue flexible linker was provided in the supplementary data. (B, C) Rifampicin resistance assay for screening for the most effective DNA helicases and the optimal fused approaches. The pmCDA1 was fused to the N-terminus (B) or C-terminus (C) of each DNA helicase with flexible linker. The wild type C. glutamicum and strains expressing only pmCDA1 were used as control groups. Data are presented as dot plots with mean ± standard deviation (SD) (n = 5). * P < 0.05, ** P < 0.01, *** P < 0.001 (Student's two-tailed t-test). (D) The change trends of rifampicin/streptomycin resistance mutations during three passages of continuous evolution. Every 24 h, the proportion of resistant cells was measured, and the cells were transferred to fresh media. The wild type and strains expressing only pmCDA1 were used as control groups. Values and error bars represent means ± standard deviations (SD) (n = 3).

Figure 3.

Developing dual genetic path modification engineering with DNA replication and transcription. (A) Schematic representation of two-plasmid system and all-in-one plasmid system (Mut-Cg). Mut-α7 and Mut-Cg plasmids used pXMJ19 plasmid as backbone, Chl^r. Cgl0854-pmCDA1 plasmids used pEC-XK99E plasmid as backbone, Kan^r. (B) Rifampicin resistance assay of two-plasmid system and all-in-one plasmid system (Mut-Cg). The arrows (black) and corresponding numbers represent net mutation rate. (C) The net mutation rate assay of Mut-Cg with different amounts of lacO. Net mutation rate be represented by the ratio of mutation rates under induced and non-induced conditions. Schematic representation (D) and net mutation rate assay (E) of SD sequence was truncated step by step. (F) Rifampicin resistance assay of optimized Mut-Cg (oMut-Cg) and temperature-sensitive oMut-Cg (oMut-Cg^ts). The oMut-Cg^ts which replication origin of oMut-Cg^ts is pBL^ts, could be eliminated at 34°C. The arrows (black) and corresponding numbers represent net mutation rate. (G) Rifampicin resistance phenotype difference of C. glutamicum cells before and after plasmid oMut-Cg^ts elimination. The cells with plasmids oMut-Cg^ts and the cells after plasmid curing were induced with 0.5 mM IPTG at 30°C, 200 rpm. After 24 h, 100 ul cultures were plated on BHI + 8 μg/ml rifampin plate, another 100 ul cultures was 10-fold serial dilution by sterile physiological saline and plated on BHI plate. All plates were cultured in 30°C for the same time. Data are presented as dot plots with mean ± standard deviation (SD) (n = 5). ns, not significant, * P < 0.05, ** P < 0.01, *** P < 0.001 (Student's two-tailed t-test).

Figure 4.

Whole genome sequencing indicate the mutagenesis characteristics of oMut-Cg^ts. (A) Variant frequency of all possible types of single nucleotide variations (SNVs) shown in the wild-type cells and the cells containing oMut-Cg^ts. All cells underwent 24h induction experiments by 0.5 mM IPTG at 30°C. Only SNV with 0.2% or higher variant frequency are shown. Data are presented as dot plots. **** P < 0.0001 by one-way ANOVA assay. (B) Position map and average variant frequency of C > T and G > A transitions in each region of genome. The position of C > T and G > A transitions with 0.2% or higher variant frequency in the genome (X-axis) were marked by yellow spots and purple spots, respectively. The histograms show the average variant frequency of C > T and G > A transitions in each ragion of genome, and the corresponding values are indicated on the column. (C, D) The context of ATCG distribution on both sides of mutant C and G bases. The five bases on both sides of mutant C and G were collected and used to generate the sequence logos by WebLogo3 (https://weblogo.threeplusone.com/manual.html). Only C and G bases with 0.2% or higher variant frequency were analysed.

Figure 5.

oMut-Cg^ts accelerated genome evolution of C. glutamicum for improving acid and oxidative stress tolerance. (A) Flow chart of evolving acid tolerance. The 4-morpholineethanesulfonic acid (MES) was added to medium until the desired pH level was reached. (B) The growth tendency of control strain (CK, yellow line) and strain (oMut-Cg^ts, blue line) with oMut-Cg^ts in continuous evolution of acid stress tolerance. The purple line represents the pH of the culture medium. The cultivation was conducted using 24-well plates. (C) The survival difference between the wild-type strain (WT) and the evolved strain NS-6 after being subjected to lethal acid stress (pH = 4.0). The 1 ul culture washed and diluted by sterile saline solution were spotted onto the plate with pH = 7.0. (D) The growth tendency of control strain (CK, yellow line) and strain (oMut-Cg^ts, blue line) with oMut-Cg^ts in continuous evolution of oxidative stress tolerance. The red line represents the concentration of CHP. (E) The intracellular ROS level of wild-type strain (WT) and the evolved strain NO-3, NO-8 and NO-12 by the DCFH-DA fluorescent probe. The ROS level was represented by the fluorescence intensity (excitation = 502 nm, emission = 530 nm) normalized with OD₆₀₀. Data are presented with mean ± standard deviation (SD) (n = 3).

Figure 6.

oMut-Cg^ts accelerated genome evolution of C. glutamicum for improving ethanol tolerance. (A) Flow chart of evolving ethanol tolerance, include three steps: the strains were serially cultured in medium with increased ethanol concentration, single colonies were isolated on plates containing ethanol, tolerance tests were performed in 24-well plates. (B) The growth tendency of control strain (CK, yellow line) and strain (oMut-Cg^ts, blue line) with oMut-Cg^ts in continuous evolution of ethanol tolerance. The green line represents the ethanol concentration. (C) Growth assays of the wild-type strain (WT, yellow line) and evolved strain (NE-8, blue line) at different ethanol concentration. The cultivation was conducted using 24-well plates. (D) The growth difference of wild-type strain (WT) and evolved strain NE-8 on plate with 12% ethanol. The OD₆₀₀ of culture was determined by a BioTek Epoch full wavelength absorbent light label. Data are presented with mean ± standard deviation (SD) (n = 3).

Figure 7.

Analysis and verification of mutations in strain NE-8. (A) The distribution position of oMut-Cg^ts mediated SNVs on the NE-8 genome. (B) Statistics for all types of mutations in strain NE-8. (C) Growth of the wild-type strain (WT) and nine recombinant strains without ethanol. (D) Growth of the wild-type strain (WT) and nine recombinant strains with 8% ethanol. (E) Growth of the wild-type strain (WT) and nine recombinant strains with 10% ethanol. All growth experiments were performed in 24-well plates containing CGXII medium. The OD₆₀₀ of culture was determined by a BioTek Epoch full wavelength absorbent light label. Data are presented with mean ± standard deviation (SD) (n = 3).

Figure 8.

Enzymatic properties and structural analysis of Cgl2467_WT and Cgl2467_L213W. (A) Optimum catalytic temperature. (B) Temperature stability. Time courses of inactivation at 30°C. The maximum activity was set to 100%. Data are presented with mean ± standard deviation (SD) (n = 3). (C) The difference in RMSF between Cgl2467_WT and Cgl2467_L213W. (D, E) The autodocking of substrate and ligands with Cgl2467_WT and Cgl2467_L213W. The NAD⁺, K⁺ and acetaldehyde were light yellow, light purple, and light cyan, respectively. (F, G) The interaction between the substrate and the adjacent residue in Cgl2467_WT and Cgl2467_L213W.