Enhanced production of D-pantothenic acid in Corynebacterium glutamicum using an efficient CRISPR-Cpf1 genome editing method

Abstract

Background: Corynebacterium glutamicum has industrial track records for producing a variety of valuable products such as amino acids. Although CRISPR-based genome editing technologies have undergone immense developments in recent years, the suicide-plasmid-based approaches are still predominant for C. glutamicum genome manipulation. It is crucial to develop a simple and efficient CRISPR genome editing method for C. glutamicum.

Results: In this study, we developed a RecombinAtion Prior to Induced Double-strand-break (RAPID) genome editing technology for C. glutamicum, as Cpf1 cleavage was found to disrupt RecET-mediated homologous recombination (HR) of the donor template into the genome. The RAPID toolbox enabled highly efficient gene deletion and insertion, and notably, a linear DNA template was sufficient for gene deletion. Due to the simplified procedure and iterative operation ability, this methodology could be widely applied in C. glutamicum genetic manipulations. As a proof of concept, a high-yield D-pantothenic acid (vitamin B5)-producing strain was constructed, which, to the best of our knowledge, achieved the highest reported titer of 18.62 g/L from glucose only.

Conclusions: We developed a RecET-assisted CRISPR-Cpf1 genome editing technology for C. glutamicum that harnessed CRISPR-induced DSBs as a counterselection. This method is of great importance to C. glutamicum genome editing in terms of its practical applications, which also guides the development of CRISPR genome editing tools for other microorganisms.

Keywords: CRISPR-Cpf1; Corynebacterium glutamicum; Genome editing; Metabolic engineering; Synthetic biology; Vitamin.

Fig. 1

Effects of the expression of different recombinases on HR of DNA fragment into the genome. A 528 bp DNA fragment flanked by approximately 500 bp genomic-homologous upstream and downstream arms was used. The integration efficiency was reflected by the colonies formed on kanamycin agar plates, which were counted or calculated based on dilution. Experiments were performed in triplicates. Values are presented as mean ± SD. ***P < 0.005

Fig. 2

Gene deletion and insertion using the CRISPR–Cpf1 system, and the effects of RecET expression and promoters that drive FnCpf1 expression are represented. a Colonies formed under different conditions. When FnCpf1 and crRNA were both expressed, the number of colonies was several orders of magnitude lower than that observed in the control, indicating that the Cpf1–crRNA complex is functional in C. glutamicum. The expression of RecET significantly increased the number of colonies, probably by improving the homologous recombination (HR) activity. P_tuf-driven expression of FnCpf1 decreased the number of colonies compared to that in the P_lacM-FnCpf1 expressing cassette. b Gene deletion efficiencies (deletion of 534 bp from CgDel); c Gene integration efficiencies (insertion of 528 bp into CgInt). The introduction of RecET and the P_tuf promoter significantly enhanced gene deletion and integration. Experiments were performed in triplicates. Values are presented as mean ± SD. **P < 0.01, *** P < 0.001

Fig. 3

RecET-mediated HR and IPTG-inducible Cpf1 genomic cleavage, potentially applicable for the development of a two-step CRISPR–Cpf1 system. a Linear templates with different lengths are used for integration, and kanamycin-resistant colonies represent the HR activity. b A decrease in the number of colonies after IPTG addition represents a strong inducible effect. The colonies were counted or calculated based on dilution. Experiments were performed in triplicates. Values are presented as mean ± SD. **P < 0.01, *** P < 0.001; n.s., not significant

Fig. 4

Optimization of the two-step RAPID CRISPR–Cpf1 system. a Different operating procedures for the RAPID system, scheme 1—the transformants are cultured in liquid medium without IPTG induction after 2 h recovery, prior to inoculation onto solid medium supplemented with IPTG. scheme 2—is the same as scheme 1, except that IPTG is added to the liquid culture. scheme 3—the transformants are recovered for a different time duration and directly spread onto a solid medium supplemented with IPTG. b Gene deletion efficiencies for different operating schemes. c Promoter optimization for RecET expression. d Integration efficiencies for fragments of different sizes. e Gene deletion and integration efficiencies using linear DNA template. Experiments were performed in triplicates. Values are presented as mean ± SD. *P < 0.05, **P < 0.01, *** P < 0.001; n.s., not significant

Fig. 5

Overview of the RAPID genome editing system. a Principle of the currently available one-step CRISPR technologies. b Principle of the RAPID technology, wherein CRISPR-crRNA is used as a counterselection, which significantly increases the number of edited colonies owing to the high-efficiency HR mediated by RecET. c Workflow of the RAPID technology, wherein the transformation of linear donor fragments and pXM-sacB-crRNA can be used for gene deletion. The whole process is simplified and comparable to that of the one-step CRISPR genome editing system

Fig. 6

Biosynthetic pathways of D-PA in C. glutamicum and the strategies for constructing the D-PA producing strain, and efficiencies for the corresponding genome editing using RAPID. a The red X represents the gene that was deleted, the blue boxes represent genes that were overexpressed by genomic integration using a strong promoter, and the green boxes represent genes that were overexpressed in a plasmid: ilvBN (acetohydroxy acid synthase), ilvC (ketol-acid reductoisomerase and ketopantoate reductase), ilvE (branched-chain-amino-acid aminotransferase), avtA (valine-pyruvate aminotransferase), aspB (aspartate aminotransferase), aspA (aspartate ammonia lyase from E. coli), ilvA (threonine dehydratase), panB (α-ketoisovalerate hydroxymethyltransferase from Bacillus subtilis), panC (pantothenate synthetase from B. subtilis), and panD (aspartate 1-decarboxylase from B. subtilis). b The efficiencies of the above-mentioned gene deletions and insertions. A 500‒800 bp fragment was normally deleted for gene inactivation, and a small fragment (here less than 600 bp) was replaced by the inserted gene cassette

Fig. 7

Production of D-PA using shake-flask fermentation and in a 5 L fermenter. a The biomass (optical density (OD) at 600 nm) of these strains using shake-flask fermentation. b The titers of D-PA in shake-flask fermentation. c The biomass, D-PA titer and residual glucose of strain Pan-4/pXtuf-panBCD in a 5 L fermenter. The shake-flask fermentation was performed in triplicate, and values are presented as mean ± SD