Overview

Gene editing plays a key role in understanding gene functions, which form the basis of heredity and determine the function, growth, development, and reproduction of all living organisms. One of the important gene editing technologies, recombineering, is homologous recombination-mediated genetic engineering catalyzed by bacteriophage-encoded homologous recombination systems with single- or double-stranded DNA (ss- or dsDNA) containing the desired changes, flanked by homologous sequences.^1,2,3 It can make precise and targeted modifications in chromosome and episomal replicons, such as plasmids and bacterial artificial chromosomes. This in vivo technology enables the modification of DNA sequences without the use of restriction sites, a crucial component of classical genetic engineering involving restriction enzymes and ligases to combine DNA sequences in a specific order. Recombineering and its derived methods have rapidly evolved from one genomic alteration at a time in a few species to multiple changes across many loci in numerous species, such as multiplex automated genome engineering (MAGE). Furthermore, combined with clustered, regularly interspaced short palindromic repeats and their associated proteins (CRISPR-Cas), recombineering becomes a powerful tool for genome engineering with high efficiency in a wide variety of organisms, including Escherichia coli.²

The main steps for recombineering are 1) designing the donor DNA, i.e., single-stranded oligo used in this lab module, 2) preparing recombineering competent cells, 3) electrotransforming the donor DNA into the cells, 4) plating cells and screening for edited/mutant phenotype, i.e., streptomycin-resistant (strepR) in this lab module, and 5) validating proper gene editing in cells/colonies obtained from step 4. In bacteria, one obstacle to overcome for successful recombineering is that most commercially available bacterial strains derived from E. coli are not readily transformed by linear DNA.⁴ This is largely due to rapid DNA degradation by the endogenous RecBCD exonuclease. Thus, mutant RecBCD bacterial strains were made for gene editing. However, earlier studies have shown that they grow poorly, are defective for recombination, and do not support efficient replication of many plasmids used for recombinant DNA.^4,5,6 To overcome these deficiencies, scientists have combined the λRed homologous recombination system with ss- or dsDNA to make insertion, deletion, and point mutations in microbial genomes, including E. coli.

Three proteins involved in the λRed recombination system are λGam inhibiting RecBCD activities, λExo resecting the 5′-ended strand of a dsDNA end in a 5′ → 3′ direction to leave  3′ ssDNA overhangs, and λBeta binding to ssDNA to promote its annealing to a complementary target and thus, increasing the homologous recombination between the ssDNA and the target site to be edited (Figure 3-1).³ Murphy demonstrated successful bacterial recombineering by transforming linear DNA with long homologies into a bacterial RecBCD mutant in the presence of the bacteriophage λ proteins Exo and Beta.⁷ Later, Yu et al. constructed a recombinogenic E. coli strain containing the three λRed prophage recombination genes, exo, beta, and gam, under the control of a temperature-sensitive lambda cI-repressor to prohibit their expression at 32°C until it was turned on at 42°C.⁸ While recombineering with dsDNA requires both the λExo and λBeta proteins, ssDNA requires only the λBeta and promotes the recombination efficiency to 6% of treated cells, a breakthrough in the ease of making point mutations and increasing efficiency.⁹ Later study by Mosberg et al. showed that the λRed recombineering in E. coli occurs through a fully single-stranded intermediate.¹⁰

The E. coli methyl-directed mismatch repair (MMR) affects the ssDNA recombination events. Restoring the mutant allele to its original sequence reduces the recombination frequency by > 100-fold.¹¹ Without MMR, λRed-mediated oligo recombination point can cause mutations in up to 25% of electroporated surviving cells, and λBeta was the only bacteriophage function required for this level of recombination, suggesting Beta directs the ssDNA to the replication fork as it passes the target sequence during DNA replication.¹² For this lab module, the MMR-deficient bacterial strain HME63 will be used to decrease or eliminate the correction of edited nucleotides by MMR, thereby enhancing recombineering efficiency.

The rpsL encoding E. coli 30S ribosomal protein S12 is a good candidate for gene editing because of its well-characterized mutants and abundance in cells. Protein S12 affects ribosome assembly and translational function, drug sensitivity, and ribosome biogenesis.¹³ When streptomycin binds the protein, it raises error rates of bacterial translation by increasing the ribosome’s affinity for non-cognate tRNAs and interfering with its proofreading. Thus, streptomycin can be used as an antibiotic because the production and accumulation of defective proteins increase, resulting in bacterial cell death. In contrast, mutations in the S12 protein can prevent it from binding to the antibiotic, allowing cells to survive. One of the common streptomycin-resistant mutations in the rpsL, K42T (lysine to threonine conversion at amino acid #42), was generated by changing a single nucleotide at position 128 (A→C) in the E. coli K-12 strain.¹⁴ The resulting amino acid mutant, K42T, prevents streptomycin from binding, resulting in the rapid emergence of streptomycin-resistant bacteria without significantly affecting their translational accuracy.

DNA replication and transcription use the same template and occur at high frequencies in cells. Transcription complexes can act as a roadblock for replication and vice versa, causing transcription-replication conflicts. Thus, transcriptions are often co-directional with the direction of DNA replication fork movement to avoid the harmful consequences of head-on replication-transcription conflict, which can lead to replication fork arrest and genomic instability, as well as negative selection against deleterious effects of the conflict.^15,16 Thus, most bacterial genes, particularly those that are essential and highly transcribed, such as ribosomal genes, are located on the leading strand, a phenomenon known as strand-biased gene distribution (SGD).¹⁷ In contrast, genes of some functional categories, such as transcription factors, have higher preferences for the lagging strands. There is also a balancing force that tends to prevent genes from all moving to the leading and more efficient strand.¹⁵

SGD can affect gene editing efficiency. Court et al. proposed that the single-strand regions of the incoming linear DNA bound by the λBeta protein are annealed to complementary single-strand gaps arising at the replication fork during DNA replication.¹⁸ Consistent with this model, Ellis et al. showed that when comparing recombination efficiency between the two complementary single-stranded oligos as donor DNA, the ‘lagging strand’ ssODN corresponding to sequences of Okazaki fragments should have a higher frequency due to the increased availability of ssDNA regions during lagging strand synthesis (Figure 3-2 left).⁹ Like the short RNA primers on DNA replication, the Beta-bound oligo has more transient ssDNA regions available on its template strand for annealing during recombineering (Figure 3-2 right) and, thus, results in a higher editing frequency.

Although the λRed recombination system combined with ssODN is made for successful target-specific recombineering, the editing efficiency can still be low in some cases due to the lack of selection markers for screening desired mutants. To overcome the obstacle, Jiang et al. combined CRISPR/Cas9 of Streptococcus pyogenes with recombineering and showed that 65% of recovered cells are recombinants due to the ‘passive’ counter-selection via CRISPR/Cas9-mediated cleavage resulted in the death of non-edited cells.^2,19 For this lab module, we will start with E. coli recombineering with DNA oligos containing the desired editing sequence, flanked by homologous sequences at both ends, to target the gene of S12 protein in E. coli genome.

In the first two weeks, we will compare E. coli recombineering efficiencies with two complementary ssONDs (sense and antisense) as DNA templates. The edited cells should contain threonine (T) in place of lysine (K) at amino acid #42 (K42T) of the S12 protein (AA→CC at nucleotides #128 and #129) and prevent the protein from binding to streptomycin, surviving in the streptomycin selection medium. The second part of the module is using CRISPR-Cas9 to eliminate non-edited cells through the target’s recognition and cleavage by Cas9, which also induces the recombination of the DNA oligo at the target site. It aims to determine whether the editing efficiency is enhanced with the help of CRISPR-Cas9, as Cas9 cleavage kills the unedited cells and induces the recombination of ssODN at the target site. An outline of weekly lab exercises for this module is shown in Figure 3-3.