Efficient and Scalable Precision Genome Editing in Staphylococcus aureus through Conditional Recombineering and CRISPR/Cas9-Mediated Counterselection

Abstract

Staphylococcus aureus is an important human pathogen, but studies of the organism have suffered from the lack of a robust tool set for its genetic and genomic manipulation. Here we report the development of a system for the facile and high-throughput genomic engineering of S. aureus using single-stranded DNA (ssDNA) oligonucleotide recombineering coupled with clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9-mediated counterselection. We identify recombinase EF2132, derived from Enterococcus faecalis, as being capable of integrating single-stranded DNA oligonucleotides into the S. aureus genome. We found that EF2132 can readily mediate recombineering across multiple characterized strains (3 of 3 tested) and primary clinical isolates (6 of 6 tested), typically yielding thousands of recombinants per transformation. Surprisingly, we also found that some S. aureus strains are naturally recombinogenic at measurable frequencies when oligonucleotides are introduced by electroporation, even without exogenous recombinase expression. We construct a temperature-sensitive, two-vector system which enables conditional recombineering and CRISPR/Cas9-mediated counterselection in S. aureus without permanently introducing exogenous genetic material or unintended genetic lesions. We demonstrate the ability of this system to efficiently and precisely engineer point mutations and large single-gene deletions in the S. aureus genome and to yield highly enriched populations of engineered recombinants even in the absence of an externally selectable phenotype. By virtue of utilizing inexpensive, commercially synthesized synthetic DNA oligonucleotides as substrates for recombineering and counterselection, this system provides a scalable, versatile, precise, inexpensive, and generally useful tool for producing isogenic strains in S. aureus which will enable the high-throughput functional assessment of genome variation and gene function across multiple strain backgrounds.IMPORTANCE Engineering genetic changes in bacteria is critical to understanding the function of particular genes or mutations but is currently a laborious and technically challenging process to perform for the important human pathogen Staphylococcus aureus In an effort to develop methods which are rapid, easy, scalable, versatile, and inexpensive, here we describe a system for incorporating synthetic, mutagenic DNA molecules into the S. aureus genome and for eliminating cells that lack the engineered mutation. This method allows efficient, precise, and high-throughput genetic engineering of S. aureus strains and will facilitate studies seeking to address a variety of issues about the function of particular genes and specific mutations.

Keywords: CRISPR; Cas9; Staphylococcus aureus; genetic engineering; genome editing; mutS; recombineering.

FIG 1

Recombineering in S. aureus. (A) Activity of different recombinases when expressed exogenously in S. aureus compared to paired, mock-transformation controls lacking mutagenic oligonucleotide. Recombineering utilized mutagenic oligonucleotides encoding rifampin resistance. Recombination frequencies significantly higher than those observed in paired controls (≤P = 0.002 [two-tailed t test]) are indicated by an asterisk. TE, Tris-EDTA. (B) Sequence confirmation of recombineered isolates. Residues 477 through 485 of S. aureus rpoB are shown, with the H481Y missense mutation indicated as a white letter in a purple field. Mutations introduced by the recombineering oligonucleotide are highlighted between wild-type and engineered strains. (C) Recombineering efficiencies across laboratory and clinical strains. Results are shown for wild-type strains and strains exogenously expressing recombinase EF2132 compared to paired mock-transformation controls. Recombination frequencies significantly higher than those observed in controls (≤P = 0.008 [two-tailed t test]) are indicated by a red asterisk for wild-type strains and a purple asterisk for strains expressing recombinase. Sequence type (ST) and clonal complex (CC) data are indicated in brackets next to each strain name. (D) Effect of various oligonucleotide lengths, mismatch repair-evading silent substitutions, and phosphorothioate modifications on recombination efficiency. Recombination frequencies significantly higher than those observed in the mock transformation control (≤P = 0.002 [two-tailed t test]) are indicated by an asterisk. Error bars in all panels indicate standard errors of the means.

FIG 2

Overview of recombineering and Cas9-mediated counterselection for S. aureus genome engineering. (A) Temperature-sensitive recombineering vector pCN-EF2132tet, encoding recombinase (green circles), is transformed into the strain to be edited. Key elements of the vector are diagrammed and include recombinase EF2132 followed by a transcriptional terminator; high-copy-number temperature-sensitive S. aureus replicon T181cop-634ts; a chloramphenicol resistance gene for selection in S. aureus; and the E. coli ColE1 origin of replication with an ampicillin resistance selectable marker for maintenance in E. coli. (B) Temperature-sensitive counterselection plasmid pCAS9counter is introduced at the time that recombineering is performed. Key elements of the vector are diagrammed and include a synthetic guide RNA (sgRNA) targeted to the genomic site being modified; Cas9 followed by a transcriptional terminator; low-copy-number temperature-sensitive S. aureus replicon E194ts (which is compatible with T181cop-634ts); an erythromycin resistance gene for selection in S. aureus; and the E. coli ColE1 origin of replication with an ampicillin resistance selectable marker for maintenance in E. coli. (C and D) Recombineering is performed using a mutagenic oligonucleotide (blue curved lines) encoding the desired change (yellow diamond). Transformation with these elements leads to two possible outcomes: integration of the mutagenic oligonucleotide is successful and Cas9 (yellow Pac-Man symbol) is unable to cleave the genome (C), or integration of the mutagenic oligonucleotide does not occur and Cas9 introduces a double-stranded break into the host genome, killing unedited cells (D). (E) Brief growth of bacteria at elevated temperatures which are nonpermissive for the plasmid system results in its loss, leaving isogenic, edited cells.

FIG 3

Recombineering with concurrent Cas9-mediated counterselection. (A) Frequency of colonies recovered using different combinations of counterselection vectors, rifampin resistance recombineering oligonucleotide, and selective media when recombineering oligonucleotides and counterselection vectors are introduced concurrently. (B) Fraction of colonies carrying counterselection vectors which display resistance to rifampin after recombineering with cotransformation of counterselection vectors. Error bars in both panels indicate standard errors of the means.

FIG 4

Recombineering of a 2.4-kb mutS deletion strain and effects of mutS deletion on subsequent recombineering. (A) PCR amplification of the mutS locus from representative colonies surviving counterselection. The wild-type allele corresponds to an amplicon 3.5 kb in size, whereas the engineered deletion results in a product at 1.1 kb. A 1-kb ladder is shown. (B) Recombineering efficiencies of the mutS deletion strain and mutS-intact parental strain compared to paired, mock-transformation controls lacking mutagenic oligonucleotide. Recombineering oligonucleotides encode rifampin resistance and either incorporate or lack silent mutations promoting bypass of MMR pathway repair. Error bars indicate standard errors of the means.