Characterizing a thermostable Cas9 for bacterial genome editing and silencing

Abstract

CRISPR-Cas9-based genome engineering tools have revolutionized fundamental research and biotechnological exploitation of both eukaryotes and prokaryotes. However, the mesophilic nature of the established Cas9 systems does not allow for applications that require enhanced stability, including engineering at elevated temperatures. Here we identify and characterize ThermoCas9 from the thermophilic bacterium Geobacillus thermodenitrificans T12. We show that in vitro ThermoCas9 is active between 20 and 70 °C, has stringent PAM-preference at lower temperatures, tolerates fewer spacer-protospacer mismatches than SpCas9 and its activity at elevated temperatures depends on the sgRNA-structure. We develop ThermoCas9-based engineering tools for gene deletion and transcriptional silencing at 55 °C in Bacillus smithii and for gene deletion at 37 °C in Pseudomonas putida. Altogether, our findings provide fundamental insights into a thermophilic CRISPR-Cas family member and establish a Cas9-based bacterial genome editing and silencing tool with a broad temperature range.

Fig. 1

The Geobacillus thermodenitrificans T12 type-IIC CRISPR-Cas locus encoding ThermoCas9. a Schematic representation of the genomic locus encoding ThermoCas9. The domain architecture of ThermoCas9 based on sequence comparison, with predicted active sites residues highlighted in magenta. A homology model of ThermoCas9 generated using Phyre 2 is shown, with different colors for the domains. b Phylogenetic tree of Cas9 orthologues highly identical to ThermoCas9. Evolutionary analysis was conducted in MEGA7. c SDS-PAGE of ThermoCas9 after purification by metal-affinity chromatography and gel filtration. The migration of the obtained single band is consistent with the theoretical molecular weight of 126 kD of the apo-ThermoCas9

Fig. 2

ThermoCas9 PAM analysis. a Schematic illustrating the in vitro cleavage assay for discovering the position and identity (5′-NNNNNNN-3′) of the protospacer adjacent motif (PAM). Magenta triangles indicate the cleavage position. b Sequence logo of the consensus 7 nt long PAM of ThermoCas9, obtained by comparative analysis of the ThermoCas9-based cleavage of target libraries. Letter height at each position is measured by information content. c Extension of the PAM identity to the eighth position by in vitro cleavage assay. Four linearized plasmid targets, each containing a distinct 5′-CCCCCCAN-3′ PAM, were incubated with ThermoCas9 and sgRNA at 55 °C for 1 h, then analyzed by agarose gel electrophoresis. Supplementary Fig. 8 shows the uncropped gel image. d In vitro cleavage assays for DNA targets with different PAMs at 30 and 55 °C. Sixteen linearized plasmid targets, each containing one distinct 5′-CCCCCNNA-3′ PAM, were incubated with ThermoCas9 and sgRNA, then analyzed for cleavage efficiency by agarose gel electrophoresis. See also Supplementary Fig. 3. Supplementary Fig. 9 shows the uncropped gel images

Fig. 3

ThermoCas9 temperature range and effect of sgRNA-binding. a Schematic representation of the sgRNA and a matching target DNA. The target DNA, the PAM and the crRNA are shown in gray, blue and turquoise, respectively. The site where the crRNA is linked with the tracrRNA is shown in purple. The dark blue and light blue boxes indicate the predicted three and two loops of the tracrRNA, respectively. The 41-nt truncation of the repeat-anti-repeat region and the three loops of the sgRNA are indicated by the magenta dotted line and magenta triangles, respectively. b The importance of the predicted three stem-loops of the tracrRNA scaffold was tested by transcribing truncated variants of the sgRNA and evaluating their ability to guide ThermoCas9 to cleave target DNA at various temperatures. Average values of three replicates are shown, with error bars representing S.D. The blots of one of the replicates are shown in Supplementary Fig. 10. c The importance of the length of the spacer was tested by transcribing truncated variants of the initial spacer in the sgRNA and evaluating their ability to guide ThermoCas9 to cleave target DNA at 55 °C. Average values of three replicates are shown, with error bars representing S.D.. The blots of one of the replicates are shown in Supplementary Fig. 11. d To identify the maximum temperature, endonuclease activity of ThermoCas9:sgRNA RNP complex was assayed after incubation at 60, 65, and 70 °C for 5 or 10 min. The pre-heated DNA substrate was added and the reaction was incubated for 1 h at the corresponding temperature. e Comparison of active temperature range of ThermoCas9 and SpCas9 by activity assays conducted after 5 min of incubation at the indicated temperature. The pre-heated DNA substrate was added and the reaction was incubated for 1 h at the same temperature

Fig. 4

Protospacer targeting Specificity of ThermoCas9. a Scheme of the generated mismatch protospacers library, employed for evaluating the ThermoCas9:sgRNA targeting specificity in vitro. The mismatch spacer-protospacer positions are shown in red, the PAM in light blue with the fifth to eighth positions underlined. b Graphical representation of the ThermoCas9:sgRNA cleavage efficiency over linear or plasmid targets with different mismatches at 37 °C. The percentage of cleavage was calculated based on integrated band intensities after gel electrophoresis. Average values from three biological replicates are shown, with error bars representing S.D. The blots of one of the replicates are shown in Supplementary Fig. 12. c Graphical representation of the ThermoCas9:sgRNA cleavage efficiency over linear or plasmid targets with different mismatches at 55 °C. The percentage of cleavage was calculated based on integrated band intensities after gel electrophoresis. Average values from three biological replicates are shown, with error bars representing S.D. The blots of one of the replicates are shown in Supplementary Fig. 12

Fig. 5

ThermoCas9-based genome engineering in prokaryotes. a Schematic overview of the basic pThermoCas9_Δgene-of-interest (goi) construct. The thermocas9 gene was introduced either to the pNW33n (B. smithii) or to the pEMG (P. putida) vector. Homologous recombination flanks were introduced upstream thermocas9 and encompassed the 1 kb (B. smithii) or 0.5 kb (P. putida) upstream and 1 kb or 0.5 kb downstream region of the gene of interest (goi) in the targeted genome. A sgRNA-expressing module was introduced downstream the thermocas9 gene. As the origin of replication (ori), replication protein (rep), antibiotic resistance marker (AB) and possible accesory elements (AE) are backbone specific, they are represented with dotted outline. b Agarose gel electrophoresis showing the resulting products from genome-specific PCR on ten colonies from the ThermoCas9-based pyrF deletion process from the genome of B. smithii ET 138. All ten colonies contained the ΔpyrF genotype and one colony was a clean ΔpyrF mutant, lacking the wild-type product. Supplementary Fig. 13 shows the uncropped gel image. c Schematic overview of the basic pThermoCas9i_goi construct. Aiming for the expression of a catalytically inactive ThermoCas9 (ThermodCas9: D8A, H582A mutant), the corresponding mutations were introduced to create the thermodcas9 gene. The thermodcas9 gene was introduced to the pNW33n vector. A sgRNA-expressing module was introduced downstream the thermodcas9. d Graphical representation of the production, growth and RT-qPCR results from the ldhL silencing experiment using ThermodCas9. The graphs represent the lactate production, optical density at 600 nm and percentage of ldhL transcription in the repressed cultures compared to the control cultures. Average values from three biological replicates are shown, with error bars representing S.D