Title: Improved bacterial recombineering by parallelized ... · 23 Efficient Recombineering (SEER). By performing SEER in E. coli to screen hundreds of putative 24 SSAPs, we identify

Title: Improved bacterial recombineering by parallelized protein discovery 1

Authors: Timothy M. Wannier1*, Akos Nyerges1,2, Helene M. Kuchwara1, Márton Czikkely2, 2

Dávid Balogh2, Gabriel T. Filsinger1, Nathaniel C. Borders1, Christopher J. Gregg1, Marc J. Lajoie1, 3

Xavier Rios1, Csaba Pál2, George M. Church1* 4

1 – Department of Genetics, Harvard Medical School, 77 Avenue Louis Pasteur, Boston, MA 5

02115 6

2 – Synthetic and Systems Biology Unit, Institute of Biochemistry, Biological Research Centre, 7

Szeged, HU-6726, Hungary 8

* To whom correspondence should be addressed: Timothy M. Wannier 9

[email protected], George M. Church [email protected] 10

Keywords: Recombineering, Genome Engineering, Synthetic Biology, recT, single-stranded DNA-11

annealing protein 12

13

author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the. https://doi.org/10.1101/2020.01.14.906594doi: bioRxiv preprint

https://doi.org/10.1101/2020.01.14.906594

Abstract 14

Exploiting bacteriophage-derived homologous recombination processes has enabled precise, 15

multiplex editing of microbial genomes and the construction of billions of customized genetic 16

variants in a single day. The techniques that enable this, Multiplex Automated Genome 17

Engineering (MAGE) and directed evolution with random genomic mutations (DIvERGE), are 18

however currently limited to a handful of microorganisms for which single-stranded DNA-19

annealing proteins (SSAPs) that promote efficient recombineering have been identified. Thus, to 20

enable genome-scale engineering in new hosts, highly efficient SSAPs must first be found. Here 21

we introduce a high-throughput method for SSAP discovery that we call Serial Enrichment for 22

Efficient Recombineering (SEER). By performing SEER in E. coli to screen hundreds of putative 23

SSAPs, we identify highly active variants PapRecT and CspRecT. CspRecT increases the efficiency 24

of single-locus editing to as high as 50% and improves multiplex editing by 5 to 10-fold in E. coli, 25

while PapRecT enables efficient recombineering in Pseudomonas aeruginosa, a concerning 26

human pathogen. CspRecT and PapRecT are also active in other, clinically and biotechnologically 27

relevant enterobacteria. We envision that the deployment of SEER in new species will pave the 28

way toward pooled interrogation of genotype-to-phenotype relationships in previously 29

intractable bacteria. 30

31

32


https://doi.org/10.1101/2020.01.14.906594

Introduction 33

The increasing accessibility of whole-genome sequencing to microbiologists has created a gap 34

between researchers’ ability to read vs. their ability to write/edit genetic information. As the 35

availability of sequencing data increases, so too does its hypothesis-generating capacity, which 36

motivates the development of genome-editing tools that can be employed to create user-defined 37

genetic variants in a massively parallelizable manner. Currently, most techniques for editing 38

microbial genomes that meet these criteria exploit bacteriophage-derived homologous 39

recombination processes1–3. Molecular tools derived from phages enabled recombineering 40

(recombination-mediated genetic engineering), which uses short homology arms to efficiently 41

direct double-stranded DNA (dsDNA) cassette- and single-stranded DNA (ssDNA)-integration into 42

bacterial genomes4–6. Improvements to recombineering in E. coli enabled multiplex, genome-43

scale editing and the construction of billions of genetic variants in a single experiment7–9. Two of 44

these recombineering-based methods, Multiplex Automated Genome Engineering (MAGE) and 45

directed evolution with random genomic mutations (DIvERGE) have been used for a variety of 46

high-value applications3,10–14, but at their core they offers the ability to generate populations of 47

bacteria that can contain billions of precisely targeted mutations. However, while these 48

techniques function well in E. coli and some closely related Enterobacteria, efforts to reproduce 49

these results in other bacterial species have been sporadic and stymied by low efficiencies (Table 50

S1)15–31. 51

The incorporation of genomic modifications via oligonucleotide annealing at the replication fork, 52

called oligo-mediated recombineering, is the molecular mechanism that drives MAGE and 53

DIvERGE4,32,33. This method was first described in E. coli, and is most commonly promoted by the 54


https://doi.org/10.1101/2020.01.14.906594

expression of bet (here referred to as Redβ) from the Red operon of Escherichia phage λ5,6,34. 55

Redβ is a single-stranded DNA-annealing protein (SSAP) whose role in recombineering is to 56

anneal ssDNA to complimentary genomic DNA at the replication fork. Though improvements to 57

recombineering efficiency have been made7–9, the core protein machinery has remained 58

constant, with Redβ representing the state-of-the-art in E. coli and enterobacterial 59

recombineering. Redβ additionally does not adapt well to use outside of E. coli, displaying host 60

tropism, presumably toward hosts that are targets of infection for Escherichia phage λ. To enable 61

recombineering in organisms in which Redβ does not work efficiently, most often Redβ homologs 62

from prophages of genetically similar bacteria are screened15,16,21,23,30, but high levels of 63

recombineering efficiency, as seen in E. coli, remain elusive. 64

We hypothesized that the identification of an optimal SSAP is currently the limiting factor to 65

improved recombineering efficiency and to developing multiplex genome engineering tools (i.e. 66

MAGE) in new bacterial species. To provide a solution, here we present a high-throughput 67

method for isolating SSAP homologs that efficiently promote recombineering, which we call 68

Serial Enrichment for Efficient Recombineering (SEER) (Figure 1). SEER allowed us to screen two 69

SSAP libraries in a matter of weeks, each of which contains more members than, to our 70

knowledge, the sum total from all previous SSAP-screening efforts (Table S1). We first use SEER 71

to screen a library of 122 SSAPs from seven different families: RecT, Erf, Sak4, Gp2.5, Sak, Rad52, 72

and RecA, finding that proteins of the RecT and Erf families are the most promising candidates 73

for future screening. In a follow-on screen of 107 RecT variants, we then identify CspRecT (from 74

a Collinsella stercoris phage), which doubles recombineering efficiency in E. coli over Redβ. Next, 75

by focusing on another enriched variant, PapRecT (from a Pseudomonas aeruginosa phage), we 76


https://doi.org/10.1101/2020.01.14.906594

demonstrate high efficiency genome editing in its native host Pseudomonas aeruginosa, which 77

has long lacked good genetic tools. We then broadly profile the activity of CspRecT and PapRecT 78

across diverse Gammaproteobacteria. Following on our successful demonstration in this work, 79

we believe that SEER should be an easily adaptable method for identifying optimal proteins for 80

recombineering across a wide range of bacterial species. 81

Results and Discussion 82

Identification of single-stranded DNA-annealing proteins (SSAPs) 83

Previous analyses of phylogenetic data suggest that there are seven principal families of phage-84

derived SSAPs: RecT (Pfam family: PF03837), Erf (Pfam family: PF04404), Rad52 (Pfam family: 85

PF04098), Sak4 (Pfam family: PF13479), Gp2.5 (Pfam family: PF10991), Sak (Pfam family: 86

PF06378), and RecA (Pfam family: PF00154)18,35. These have been further grouped into three 87

superfamilies, wherein RecT, Erf, and Sak have been proposed to adopt Rad52-like folds and Sak4 88

and RecA cluster together into a Rad51-like superfamily18. Furthermore, Redβ homologs are best 89

classified as a part of the larger RecT family, and UvsX homologs are best classified in the RecA 90

protein family36. To generate a library that widely sampled SSAP diversity, we used a Hidden 91

Markov Model to search metagenomic databases beginning with an ensemble of SSAPs that have 92

demonstrated activity in E. coli19,37 (see methods). A library of 131 proteins was identified and 93

codon-optimized for expression in E. coli, 121 of which were synthesized without error (Table S2) 94

and cloned into a plasmid vector with a standard arabinose-inducible expression system. This 95

assembled group of SSAP variants, which we refer to here as Broad SSAP Library, has members 96

from all seven families of SSAPs (Figure 2A) and is phylogenetically diverse (Figure 2B). To 97

facilitate tracking the library over multiple selective cycles, we added 12-nucleotide barcode 22 98


https://doi.org/10.1101/2020.01.14.906594

nucleotides downstream of the stop codon of each gene (Figure S1), which enabled us to identify 99

each SSAP variant by PCR-amplification and targeted Next Generation Sequencing (NGS). 100

Serial Enrichment for Efficient Recombineering (SEER) 101

Next, to evaluate libraries of computationally identified SSAP variants, we established a high-102

throughput assay to select for SSAPs that promote efficient recombineering. We hypothesized 103

that the iterative, serial integration of easily selectable mutations into the bacterial genome, 104

coupled to subsequent allelic selection, could allow us to select SSAP variants en masse, and thus 105

enrich best performers from diverse SSAP libraries (Figure 1). This method, that we term Serial 106

Enrichment for Efficient Recombineering (or SEER) proceeds via i.) the identification and cloning 107

of SSAP libraries into expression vectors, ii.) the transformation of such libraries into an organism-108

of-interest, followed by iii.) enrichment for SSAPs efficient at recombineering, and then iv.) 109

analysis by deep sequencing to read out library composition. Step “iii” comprises the successive 110

transformation of oligos, which upon successful integration into the host genome will confer a 111

resistant phenotype, followed by antibiotic selection for the resistant allele across the entire 112

library in liquid culture. SSAP variants that incorporate mutations most effectively will thereby be 113

enriched. Finally, if needed, after the selective handles are exhausted, the SSAP-plasmid library 114

can be extracted and re-transformed into a naïve chassis for further cycles of enrichment. 115

To perform the selections efficiently we engineered an E. coli chassis to both improve 116

recombineering efficiency and to introduce genetic handles with which we could apply selective 117

pressure against non-edited alleles. We used as our parental strain, EcNR23, a derivative of E. coli 118

K-12 MG1655 that has its methyl-directed mismatch repair (MMR) machinery disabled 119

(ΔmutS::cat) to improve recombineering efficiency (see Methods for details)7. EcNR2 was 120


https://doi.org/10.1101/2020.01.14.906594

modified to i.) improve recombineering efficiency (DnaG Q576A)8 and to ii.) introduce genetic 121

handles for antibiotic selection. To this latter aim, stop codons were introduced into both cat (in 122

the mutS locus) and tolC, making the modified strain sensitive to chloramphenicol (CHL) and 123

sodium dodecyl sulfate (SDS). We refer to this modified organism as the “SEER chassis”. Next, to 124

identify selectable markers beyond cat and tolC, and thus allow multiple SEER selection cycles, 125

we tested a group of resistant alleles that we gathered from an analysis of antibiotic resistance 126

literature (See Methods for details). We found GyrA_S83L to confer robust resistance to 127

ciprofloxacin (CIP), RpoB_S512P to rifampicin (RIF), and RpsL_K43R to streptomycin (STR) (Figure 128

S2). We then verified that these three antibiotic selections, in addition to the CHL and SDS 129

selections described above, are orthogonal and that they could be performed serially on our 130

engineered E. coli SEER chassis. This enabled us to run ten rounds of SEER with only a single 131

extraction and re-transformation of the plasmid libraries into the naïve chassis (Figure 1). 132

Following the optimization of the SEER workflow, we performed ten successive cycles of selection 133

on Broad SSAP Library, representing all major SSAP families, in the E. coli SEER chassis. Four 134

replicate populations were run: one control population underwent induction of protein 135

expression and oligo transformation, but without antibiotic selection, while three experimental 136

replicates underwent antibiotic selections in the following order: SDS ® STR ® CIP ® CHL ® RIF 137

(Figure 1C). To track the enrichment of each library member we amplified the 12-nucleotide 138

barcode region after each round of selection and sequenced by NGS (Table S3). Clear winners 139

emerged relatively quickly, with top variants increasing in frequency distinctly after only a few 140

rounds of selection (Figure S3). The non-selected control population, however, also displayed 141

significant enrichment effects, which indicated that the over-expression of certain SSAP variants 142


https://doi.org/10.1101/2020.01.14.906594

may impair fitness. Thus, to normalize for fitness and the cost of protein expression, we next 143

calculated an enrichment score for each library member: we divided the average frequency of 144

each variant within the selected populations by its frequency in the non-selected control 145

population (Figure 2C). Based on NGS data, we found that the RecT family was the most enriched 146

family by a large margin (85-fold over the non-selected control), followed by the ERF family (18-147

fold over the non-selected control), and no other SSAP family showed significant enrichment 148

(Figure 2D). However, it is important to note that our codon-optimized Redβ (SR085 in our library) 149

was not significantly enriched through the ten rounds of selection. To investigate this issue 150

further we compared the performance of Redβ expressed off of its wild-type codons against the 151

codon-optimized version that was included in Broad SSAP Library. This revealed significantly 152

decreased efficiency for the codon-optimized version of Redβ (Figure S4), which indicates that 153

codon choice is an important consideration for library design, and that negative results for 154

individual SSAPs from a large screen of this sort do not necessarily indicate that the protein itself 155

is not functional. 156

Finally, we chose a set of five library members that exhibited both high frequency and enrichment 157

for further analysis (Figure 2E). We tested their recombineering efficiency against Redβ 158

expressed off of its wild-type codons on the same plasmid system used for the SEER selections. 159

To ensure an accurate measurement we queried the efficiency of each SSAP by NGS after 160

performing a silent, non-coding genetic mutation at a non-essential gene, ynfF (silent mismatch 161

MAGE oligo 7: Table S4). Broad SSAP Library member: SR016, which we introduced earlier as 162

PapRecT (UniParc ID: UPI0001E9E6CB), demonstrated the highest efficiency of recombineering, 163

i.e. 31% ± 2% (Figure 2F). 164


https://doi.org/10.1101/2020.01.14.906594

Screening diverse RecT homologs identifies a highly efficient SSAP 165

By investigating the distribution of efficient recombineering-functions across the seven principal 166

families of phage-derived SSAPs, our first SEER screen suggested the RecT family (Pfam family: 167

PF03837) as the most abundant source of recombineering proteins for E. coli. Importantly, 168

previous screens by others also detected efficient SSAPs from the RecT family (Table S1). 169

Therefore, we hypothesized that by screening additional RecT variants, again exploiting the 170

increased throughput of SEER compared to previous efforts, we might discover recombineering 171

proteins further improved over Redβ and PapRecT. To this aim we constructed a second library, 172

identifying a maximally diverse group of 109 RecT variants, 106 of which were synthesized 173

successfully, which we call Broad RecT Library (see Methods for more details). Next, as previously 174

described, we performed 10 rounds of SEER selection on Broad RecT Library (Figure S5), and upon 175

plotting frequency against enrichment after the final selection, a clear winner emerged (Figure 176

3a). This protein, which we introduced earlier as CspRecT (UniParc ID: UPI0001837D7F), 177

originates from a phage of the Gram-positive bacterium Collinsella stercoris. 178

To maximize the phylogenetic reach and applicability of these new tools, we characterized 179

CspRecT, alongside Redβ and PapRecT, subcloned into the pORTMAGE plasmid system (Figure 180

S1, Addgene accession: 120418). This plasmid contains a broad-host RSF1010 origin of 181

replication, establishes tight regulation of protein expression with an m-toluic-acid inducible 182

expression system, and disables MMR by transient overexpression of a dominant-negative 183

mutant of E. coli MutL (MutL E32K)38, which makes it possible to establish high-efficiency editing 184

without modification of the host genome. Measured with a standard lacZ recombineering assay, 185

wild-type E. coli MG1655 expressing CspRecT exhibited editing efficiency of 35-51% for various 186


https://doi.org/10.1101/2020.01.14.906594

single-base mismatches, averaging 43% or more than double the efficiency of cells expressing 187

Redβ or PapRecT off of the same plasmid system (Figure 3B). This pORTMAGE plasmid expressing 188

CspRecT we refer to as pORTMAGE-Ec1 (Addgene accession: 138474). To our knowledge the 189

efficiency of CspRecT single-locus genome editing reported here is the first to significantly exceed 190

25%, the theoretical maximum for a single incorporation event39, implying that editing occurs 191

either at multiple forks or over successive rounds of genome replication. 192

We then tested CspRecT at a variety of more complex genome editing tasks. For longer strings of 193

consecutive mismatches, which are lower efficiency events, CspRecT was again about twice as 194

efficient as Redβ. Wild type E. coli MG1655 expressing CspRecT displayed 6% or 3% efficiency (vs. 195

3% or 1% for Redβ) for the insertion of oligos conferring 18-bp or 30-bp consecutive mismatches 196

into the lacZ locus respectively (Figure 3C). To further investigate the performance of CspRecT at 197

complex, highly multiplexed genome editing tasks, we designed a set of 20 oligos spaced evenly 198

around the E. coli genome, each of which incorporates a single-nucleotide synonymous mutation 199

at a non-essential gene. Next, while expressing Redβ, PapRecT, and CspRecT separately from the 200

corresponding pORTMAGE plasmid in E. coli MG1655, we performed a single cycle of genome 201

editing with equimolar pools of 1, 5, 10, 15, and 20 oligos and assayed editing efficiency at each 202

locus by PCR amplification coupled to targeted next generation sequencing (NGS). NGS analysis 203

revealed a general trend: as the number of parallel edits grew, the degree of overperformance 204

by CspRecT also grew (Figure 3D). For instance, when making 19 simultaneous edits (one oligo 205

from the pool of 20 could not be read out due to inconsistencies in allelic amplification), CspRecT 206

averaged 10.0% editing efficiency at all loci, whereas PapRecT averaged 4.0% and Redβ averaged 207

1.9%. Importantly, despite keeping total oligo concentration fixed across all pools, aggregate 208


https://doi.org/10.1101/2020.01.14.906594

editing efficiency increased as more oligos were present in each pool. For instance, when using 209

CspRecT with a 19-oligo pool, aggregate editing efficiency was more than 200%, implying that 210

across the total recovered population of E. coli there averaged more than two edits per cell. 211

Finally, based on the increased integration efficiency with CspRecT in multiplexed genome editing 212

tasks, we also tested its performance in a directed evolution with random genomic mutations 213

(DIvERGE) experiment13. DIvERGE uses large libraries of soft-randomized oligos that have a low 214

basal error rate at each nucleotide position along their entire sequence to incorporate mutational 215

diversity into a targeted genomic locus. To compare the performance of Redβ, PapRecT, and 216

CspRecT, we performed one round of DIvERGE mutagenesis by simultaneously delivering 130 217

partially overlapping DIvERGE oligos designed to randomize all four protein subunits of the drug 218

targets of ciprofloxacin (gyrA, gyrB, parC, and parE) in E. coli MG1655. Following library 219

generation, cells were subjected to 250, 500, and 1,000 ng/mL ciprofloxacin (CIP) on LB-agar 220

plates. Variant libraries that were generated by expressing CspRecT produced more than ten 221

times as many colonies at low CIP concentrations (i.e., 250 ng/mL) as Redβ and PapRecT, while 222

at 1,000 ng/mL CIP, which requires the simultaneous acquisition of at least two mutations 223

(usually at gyrA and parC) to confer a resistant phenotype, only the use of CspRecT produced 224

resistant variants (Figure 3E). Because gyrA and parC mutations are usually necessary to confer 225

high-level CIP resistance, sequence analysis of gyrA and parC from 11 randomly selected CIP-226

resistant colonies, we found many different mutations, in combinations of up to three, most of 227

which have been described in resistant clinical isolates13,40 (Table S5). In sum, in both MAGE and 228

DIvERGE experiments, which require multiplex editing, CspRecT provided more than an order of 229


https://doi.org/10.1101/2020.01.14.906594

magnitude improvement to editing efficiency over Redβ, the current state-of-the-art 230

recombineering tool. 231

Improved genome editing in diverse Gammaproteobacteria 232

SSAPs frequently show host tropism15,25,26, but there are also indications that within bacterial 233

clades certain SSAPs may function broadly21,38,41. Therefore, we next investigated the 234

functionality of PapRecT and CspRecT in selected Gammaproteobacteria and compared their 235

efficiency to that of Redβ. We chose to focus our efforts on two enterobacterial species: 236

Citrobacter freundii ATCC 8090 and Klebsiella pneumoniae ATCC 10031, along with the more 237

distantly related Pseudomonas aeruginosa PAO1. Pathogenic isolates of K. pneumoniae and P. 238

aeruginosa are among the most concerning clinical threats due to widespread multidrug 239

resistance42. In these species, oligo-recombineering based multiplexed genome editing (i.e., 240

MAGE and DIvERGE) holds the promise of enabling rapid analysis of genotype-to-phenotype 241

relationships and predicting future mechanisms of antimicrobial resistance13,43. C. freundii, by 242

contrast, is an intriguing biomanufacturing host in which the optimization of metabolic pathways 243

has remained challenging44,45 244

To test the activity of PapRecT and CspRecT in these three organisms we built on the broad-host-245

range pORTMAGE system38 described above. For experiments in E. coli we had subcloned 246

PapRecT or CspRecT in place of Redβ into pORTMAGE311B46 (Figure S1; Addgene accession: 247

120418), which transiently disrupts MMR with EcMutL_E32K, and whose RSF1010 origin of 248

replication and m-toluic-acid-based expression system allows the plasmid to be deployed over a 249

broad range of bacterial hosts47,48. We used these same pORTMAGE-based constructs for testing 250

in C. freundii and K. pneumoniae. In P. aeruginosa the plasmid architecture remained constant, 251


https://doi.org/10.1101/2020.01.14.906594

except that we replaced the origin of replication and antibiotic resistance, instead using the 252

broad-host-range pBBR1 origin, which was shown to replicate in P. aeruginosa49, and a 253

gentamicin resistance marker (Figure S1). Next we tested these constructs (See methods for 254

details), and in all three species, PapRecT and CspRecT displayed high editing efficiencies (Figure 255

4A). In C. freundii and K. pneumoniae, just as in E. coli, we found CspRecT to be the optimal choice 256

of protein, whereas in P. aeruginosa PapRecT performed the best. We further compared PapRecT 257

to two recently reported Pseudomonas putida SSAPs (Rec2 and Ssr)26,50, and found that PapRecT, 258

isolated from a large E. coli screen performed equal to or better than proteins found in smaller 259

screens run through P. putida (Figure S6). We found, however, that the efficiency of our plasmid 260

construct was lower in P. aeruginosa than in the enterobacterial species that pORTMAGE was 261

optimized for. Therefore, to increase editing efficiency in P. aeruginosa, we next i.) optimized 262

ribosomal binding sites (RBS) for PapRecT and EcMutL, ii.) replaced EcMutL_E32K with its 263

equivalent homologous mutant from P. aeruginosa (PaMutL_E36K), iii.) incorporated the native 264

P. aeruginosa coding sequence for PapRecT instead of the E. coli codon-optimized version (Figure 265

S7). Together these changes significantly improved the editing efficiency of our best plasmid 266

construct featuring PapRecT in P. aeruginosa, which we call pORTMAGE-Pa1 (Addgene accession: 267

138475), to ~15%. 268

Virulent strains of P. aeruginosa are a frequent cause of acute infections in healthy individuals, 269

as well as chronic infections in high-risk patients, such as those suffering from cystic fibrosis51. 270

The rate of antibiotic resistance in this species is growing, with strains adapting quickly to all 271

clinically applied antibiotics52,53. The development of multidrug resistance in P. aeruginosa 272

requires the successive acquisition of multiple mutations, but due to the lack of efficient tools for 273


https://doi.org/10.1101/2020.01.14.906594

multiplex genome engineering in P. aeruginosa54,55, investigation of these evolutionary 274

trajectories has remained cumbersome. Therefore, and to demonstrate the utility of pORTMAGE-275

Pa1-based MAGE in P. aeruginosa, we simultaneously incorporated a panel of genomic mutations 276

that individually confer resistance to STR, RIF, and fluoroquinolones (i.e., CIP)56,57. Importantly, 277

the corresponding genes are also clinical antibiotic targets in P. aeruginosa58. Following a single 278

cycle of MAGE delivering 5 mutation-carrying oligos, a single-day experiment with pORTMAGE-279

Pa1, we were able to isolate all possible combinations of five resistant mutations, with more than 280

105 cells from a 1ml overnight recovery attaining simultaneous resistance to STR, RIF, and CIP 281

(Figure 4B). Interestingly, because rpsL and rpoB, the resistant loci for STR and RIF respectively, 282

are located only ~5kb apart from each other on the P. aeruginosa genome, these two mutations 283

co-segregated much more often than would be expected by independent inheritance, confirming 284

that co-selection functions similarly in P. aeruginosa to E. coli (Figure 4C)12. By genotyping and 285

characterizing resistant colonies, we could then easily determine the Minimum Inhibitory 286

Concentration (MIC) of CIP for various resistant genotypes (Table 1, Figure S8). The allure of this 287

method is that the entire workflow took only three days to complete, in contrast with other 288

genome engineering methods (i.e., CRISPR/Cas9 or base-editor-based strategies) that are either 289

less effective, have biased mutational spectra, and/or would require tedious plasmid cloning and 290

cell manipulation steps54,55. 291

Conclusion 292

The discovery and subsequent improvement of recombineering in E. coli enabled the 293

development of advanced genome-engineering techniques such as MAGE3, TRMR59, REXER1, 294

Retron-Recombineering10,60, and DIvERGE13, and applications ranging from genomic recoding61, 295


https://doi.org/10.1101/2020.01.14.906594

biocontainment11, and viral resistance61 to the study and prediction of antibiotic resistance13 and 296

the improved bioproduction of chemicals12. These powerful techniques all rely on efficient 297

recombineering, but despite the essential role that an optimal SSAP plays, no method had 298

hitherto been developed to rapidly and efficiently screen complex libraries of SSAPs. Here we 299

describe such a method that we term Serial Enrichment for Efficient Recombineering (SEER), a 300

technique for isolating efficient SSAPs from a large number of in silico-identified candidates, in 301

which successive rounds of recombineering are followed by allelic selection (Figure 1). In the 302

current work we have used SEER to screen two large libraries of SSAPs through E. coli and 303

demonstrated the power of this method by isolating a variant, CspRecT that doubles the already 304

high single-locus efficiency in E. coli to around 50%, a standout number and the highest reported 305

gene-editing efficiency of any system that is not nuclease-dependent that we are aware of. We 306

also demonstrate that CspRecT radically improves the multiplexability of oligo recombineering in 307

E. coli, showing 5 to 10-fold improvement over Redβ with methods that rely on multiplex editing 308

such as MAGE and DIvERGE. 309

Beyond the success of SEER in E. coli, the true promise of the technique is to increase the number 310

of bacterial species in which these powerful genome-engineering tools are available to 311

researchers. To this end we demonstrate that SEER can discover recombineering proteins that 312

work not only in the targeted organism, but in closely related species as well. We show that 313

proteins enriched in two SSAP libraries that were screened through E. coli meet or exceed the 314

efficiency of the best tools available in C. freundii, K. pneumoniae, and P. aeruginosa. We report 315

single-locus editing efficiencies of over 10% in each of these clinically and biotechnologically 316

relevant Gammaproteobacteria, which should readily enable multiplex genome engineering. 317


https://doi.org/10.1101/2020.01.14.906594

However, it is also important to highlight the current limitations to the scope of this work. To 318

expand modern genome-engineering tools to bacteria that are only distantly related to E. coli, 319

SEER will need to be performed in new host organisms in the future. To this end we detail three 320

important considerations for constructing SSAP libraries for future studies: i.) based on our 321

screens in E. coli, and in concert with previous efforts (Table S1), the recT and ERF families are 322

the most likely sources of SSAPs to drive efficient recombineering, ii.) phylogenetic diversity is 323

essential, as CspRecT, the best-performing SSAP in E. coli, originates from a phage of Collinsella 324

stercoris, a Gram-positive coriobacterium that is phylogenetically distant to E. coli, and iii.) future 325

SSAP screening efforts should give careful consideration to codon optimization, as codon choice 326

can have a large impact on apparent recombineering efficiency (as we observed with both Redβ 327

and PapRecT). In short, SEER holds the promise of being a universal screening method that allows 328

for the identification of highly active SSAPs in virtually any target bacterium, with the potential 329

to broaden the applicability of recombineering-based genome-engineering tools toward as yet 330

genetically intractable microorganisms. 331

Materials and Methods 332

Strains and Reagents 333

All strains used in this study are listed in Table S6. Unless otherwise noted, bacterial cultures were 334

grown in Lysogeny-Broth-Lennox (LBL) (10g tryptone, 5g yeast extract, 5g NaCl in 1L H2O). Super 335

optimal broth with catabolite repression (SOC) was used for recovery after electroporation. 336

MacConkey agar (17g pancreatic digest of gelatin, 3g peptone, 10g lactose, 1.5g bile salt, 5g NaCl, 337

13.5g agar, 0.03g neutral red, 0.001g crystal violet in 1L H2O) and IPTG-X-gal Mueller-Hinton II 338

agar (3g beef extract, 17.5g acid hydrolysate of casein, 1.5g starch, 13.5g agar in 1L H2O, 339


https://doi.org/10.1101/2020.01.14.906594

supplemented with 40 mg/L X-gal and 0.2 μM IPTG) were used to differentiate LacZ(+) and (-) 340

mutants. Cation-adjusted Mueller Hinton II Broth (MHBII) was used for antimicrobial 341

susceptibility tests. Antibiotics were ordered from Sigma-Aldrich. Recombineering oligos were 342

synthesized by Integrated DNA Technologies (IDT) or by the DNA Synthesis Laboratory of the 343

Biological Research Centre (Szeged, Hungary) with standard desalting as purification. 344

Oligo-mediated recombineering 345

Bacterial cultures (E. coli, K. pneumoniae, C. freundii, or P. aeruginosa) were grown in LBL at 37 °C 346

in a rotating drum. Overnight cultures were diluted 1:100, grown for 60 minutes or until OD600 347

≈ 0.3, whereupon expression of SSAPs was induced for 30 minutes with 0.2% arabinose or 1 mM 348

m-toluic acid as appropriate. Cells were then prepared for transformation. Briefly, E. coli, K. 349

pneumoniae, and C. freundii cells were put on ice for approximately ten minutes, washed three 350

times with cold water and resuspended in 1/100th culture volume of cold water. This same 351

procedure was followed for P. aeruginosa with the following differences: (1) Resuspension Buffer 352

(0.5 M sucrose + 10% glycerol) was used in place of water and (2) there was no pre-incubation 353

on ice, as competent cell prep was carried out at room temperature, which we found to be much 354

more efficient than preparation at 4°C. After competent cell prep, 9 µl of 100 µM oligo was added 355

to 81 µl of prepared cells for a final oligo concentration of 10 µM in the transformation mixture 356

(2.5 µM final oligo concentration was used for C. freundii and K. pneumoniae). This mixture was 357

transferred to an electroporation cuvette with a 0.1 cm gap and electroporated immediately on 358

a Gene Pulser (BioRad) with the following settings: 1.8 kV (2.2 kV in the case of P. aeruginosa), 359

200 Ω, 25 µF. Cultures were recovered with SOC media for one hour and then 4 ml of LB with 360


https://doi.org/10.1101/2020.01.14.906594

1.25x selective antibiotic and 1.25x antibiotic for plasmid maintenance were added for 361

outgrowth. 362

Engineering of SEER chassis 363

The E. coli strain described in this work as the SEER chassis is engineered from EcNR23. EcNR2 364

harbors a small piece of λ-phage integrated at the bioAB locus, which allows expression of λ-Red 365

genes, and a knockout of the methyl-directed mismatch repair (MMR) gene mutS, which 366

improves the efficiency of mismatch inheritance (MG1655 ΔmutS::cat Δ(ybhB-bioAB)::[λcI857 367

Δ(cro-orf206b)::tetR-bla]). Modifications made to EcNR2 to engineer the SEER chassis include: 1. 368

improvement of MAGE efficiency by mutating DNA primase (dnaG_Q576A)8, 2. introduction of a 369

handle for SDS selection (tolC_STOP), 3. introduction of a handle for CHL selection 370

(mutS::cat_STOP), and 4. removal of lambda phage with a zeocin resistance marker Δ[λcI857 371

Δ(cro- orf206b)::tetR-bla]::zeoR. The final strain which we refer to as the SEER chassis is 372

therefore: MG1655 Δ(ybhB-bioAB)::zeoR ΔmutS::cat_STOP tolC_STOP dnaG_Q576A. 373

Selective allele testing in the SEER chassis 374

To complement the SEER chassis’ two engineered selective handles, we tested the following 375

native antibiotic resistance alleles: [TMP: FolA P21→L, A26→G, and L28→R62], [KAN/GEN: 376

16SrRNA U1406→A and A1408→G], [SPT: 16SrRNA A1191→G and C1192→U63], [RIF: RpoB 377

S512→P and D516→G64], [STR: RpsL K4→R and K88→R65], and [CIP: GyrA S83→L66]. 90-bp oligos 378

conferring each mutation, with two PT bonds at their 5’ end and with complementarity to the 379

lagging strand were designed. Two oligos were designed to repair the engineered selective 380

handles: (1) elimination of a stop codon in the chloramphenicol acetyltransferase (cat) to confer 381


https://doi.org/10.1101/2020.01.14.906594

CHM resistance and (2) elimination of a stop codon in tolC to confer SDS resistance. Oligo-382

mediated recombineering was run with Redβ expressed off of the pARC8 plasmid and the 383

cultures were then plated onto a range of concentrations of the antibiotic to which the oligo was 384

expected to confer resistance. Colony counts were made and compared to a water-blank control. 385

Modifications targeted to provide TMP, KAN, and SPT resistance did not work adequately and so 386

were dropped. RpsL_K43R was chosen for STR selection and RpoB_S512P for RIF selection, 387

although in both cases there was not a significant observable difference between the two tested 388

alleles. An antibiotic concentration was chosen that provided the largest selective advantage for 389

those cultures transformed with oligo (Fig S2). The concentrations chosen for the selective 390

antibiotics were: 0.1% v/v SDS, 25 μg/ml STR, 100 µg/ml RIF, 0.1 µg/ml CIP, and 20 μg/ml CHL. 391

Identification of SSAP library members 392

To generate Broad SSAP Library we used a multiple sequence alignment of eight SSAPs that had 393

been shown to function in E. coli (Redβ, EF2132 from Enterococcus faecalis, OrfC from Legionella 394

pneumophila, S065 from Vibrio cholerae, Plu2935 from Photorhabdus luminescens, Orf48 from 395

Listeria monocytogenes, Orf245 from Lactococcus lactis, and Bet from Prochlorococcus siphovirus 396

P-SS219,37) to generate a Hidden Markov Model that described the weighted positional variance 397

of these proteins. We then queried non-redundant nucleotide and environmental metagenomic 398

databases using a web-based search interface67. Candidates were filtered based on gene size and 399

annotation. Those that exhibited intra-sequence similarity of greater than 98% were removed 400

from the group. We added three eukaryotic SSAP homologs to the library68. In total, Broad SSAP 401

Library contains 120 members from the homology search, 8 members from the starting sequence 402

alignment, and 3 eukaryotic members, or a total of 131 SSAP homologs (Table S2). 403


https://doi.org/10.1101/2020.01.14.906594

Broad RecT Library was generated from the full alignment of Pfam family PF03837, containing 404

576 sequences from Pfam 31.069. Using ETE 3, a phylogenetic tree made by FastTree and accessed 405

from the Pfam31.0 database was pruned, and from it a maximum diversity subtree of 100 406

members was identified (Figure S9)70. Five members of this group were found in Library S1, and 407

so were excluded, and in their place six RecT variants from Streptomyces phages and eight other 408

RecT variants were added that had previously reported activity or were otherwise of 409

interest6,15,19,22,23, bringing the library size to 109 (Table S2). 410

Library assembly 411

Broad SSAP Library and Broad RecT Library variants with a DNA barcode 22 nucleotides 412

downstream of the stop codon were codon-optimized for E. coli and synthesized by Gen9 (S1) or 413

Twist (S2). Synthesized DNA was amplified by PCR (NEB Q5 polymerase) and cloned into 414

pDONR/Zeo (Thermo) by Gibson Assembly (NEB HiFi DNA Assembly Master Mix) and then moved 415

into pARC8-DEST for arabinose-inducible expression. pARC8-DEST was engineered from a pARC8 416

plasmid71 that shows good inducible expression in E. coli by moving Gateway sites (attR1/attR2), 417

a CHL marker, and a ccdB counter-selection marker downstream of the pBAD-araC regulatory 418

region (Figure S1). This enabled easy, one-step cloning of the entire library into pARC8-DEST by 419

Gateway cloning (Thermo). The Gateway reaction was transformed into E. cloni Supreme 420

electrocompetent cells (Lucigen), providing > 10,000x coverage of both libraries in total 421

transformants. 422

Serial Enrichment for Efficient Recombineering (SEER) 423


https://doi.org/10.1101/2020.01.14.906594

Libraries were mini-prepped (NEB Monarch Kit) and electroporated into the SEER chassis with 424

more than 1,000-fold coverage. Five cycles of oligo-mediated recombineering followed by 425

antibiotic selection were then conducted (Fig. 1B). 5 µl of the 5 ml recovery from the 426

recombineering step was immediately plated onto LBL + selective antibiotic plates to estimate 427

the total throughput of the selective step. This allowed us to ensure that the library was never 428

bottlenecked—the first round of selection was the most stringent, but we ensured that there was 429

> 500x coverage at this stage. Following five rounds of selection, the plasmid library was mini-430

prepped and transformed back into the naïve parent strain, followed by five further rounds of 431

selection (ten in total). After each selective step a 100 µl aliquot of the antibiotic-selected 432

recovery was frozen down at -80 °C in 25% glycerol for analysis by NGS. 433

Next Generation Sequencing of libraries 434

Primers were designed to amplify a 215 bp product containing the barcode region of the SSAP 435

libraries from the pARC8 plasmid and to add on Illumina adaptors. PCR amplification was done 436

with Q5 polymerase (NEB) performed on a LightCycler 96 System (Roche), with progress tracked 437

by SYBR Green dye and amplification halted during the exponential phase. Barcoding PCR for 438

Illumina library prep was performed as just described, but with NEBNext Multiplex Oligos for 439

Illumina Dual Index Primers Set 1 (NEB). Barcoded amplicons were then purified with AMPure XP 440

magnetic beads (Beckman Coulter), pooled, and the final pooled library was quantified with the 441

NEBNext Library Quant Kit for Illumina (NEB). The pooled library was diluted to 4 nM, denatured, 442

and a paired end read was run with a MiSeq Reagent Kit v3, 150 cycles (Illumina). Sequencing 443

data was downloaded from Illumina, sequences were cleaned with Sickle72 and analyzed with 444

custom scripts written in Python. 445


https://doi.org/10.1101/2020.01.14.906594

Measuring recombineering efficiency in E. coli by NGS 446

To measure single locus editing, a recombineering cycle was run with an oligo that confers a 447

single base pair non-coding mismatch in a non-essential gene (Table S4 - silent mismatch MAGE 448

oligo 7). The allele was then amplified by PCR and editing efficiency was measured by NGS as 449

described above. To test multiplex editing, the concentration of oligo was held fixed (10 µM in 450

the final electroporation mixture), but the total number of oligos in the mixture was varied. Pools 451

of oligos to test editing at 5, 10, 15, or 20 alleles simultaneously were designed so as to space the 452

edits relatively evenly around the genome. The 5-oligo pool contained oligo #’s 3,7,11,15,17, the 453

10-oligo pool added oligo #’s 1,5,9,13,19, the 15-oligo pool added oligo #’s 4,8,12,16,18, and the 454

final 20-oligo pool contained all of the silent mismatch MAGE oligos listed in Table S4. One locus 455

(locus 8) showed major irregularities when sequenced, and so we eliminated it from our analyses. 456

DIvERGE-based simultaneous mutagenesis of gyrA, gyrB, parE, and parC 457

A single round of DIvERGE mutagenesis was carried out to simultaneously mutagenize gyrA, gyrB, 458

parE, and parC in E. coli MG1655 by the transformation of an equimolar mixture of 130 soft-459

randomized DIvERGE oligos, tiling the four target genes. The sequences and composition of these 460

oligos were published previously (Nyerges, A., et. al, PNAS, 2018)13. To perform DIvERGE, 4 µl of 461

this 100 µM oligo mixture was electroporated into E. coli K-12 MG1655 cells expressing Redb 462

from pORTMAGE311B, PapRecT from pORTMAGE312B, or CspRecT from pORTMAGE-Ec1, in 5 463

parallel replicates according to our previously described protocol46. Following electroporation, 464

the replicates were combined into 10 ml fresh TB media. Following recovery for 2 hours, cells 465

were diluted by the addition of 10 ml LB and allowed to reach stationary phase at 37 °C, 250 rpm. 466


https://doi.org/10.1101/2020.01.14.906594

Library generation experiments were performed in triplicates. Following library generation, 1 mL 467

of outgrowth from each library was subjected to 250, 500, and 1,000 ng/mL ciprofloxacin (CIP) 468

stresses on 145 mm-diameter LB-agar plates. Colony counts were determined after 72 hours of 469

incubation at 37 °C, and individual colonies were subjected to further genotypic (i.e., capillary 470

DNA sequencing) analysis and phenotypic (i.e., Minimum Inhibitory Concentration) 471

measurements. 472

pORTMAGE plasmid construction and optimization 473

All plasmids used in the study are listed in Table S7. Cloning reactions were performed with Q5 474

High-Fidelity Master Mix and HiFi DNA Assembly Master Mix (New England Biolabs). 475

pORTMAGE312B (Addgene accession: 128969) and pORTMAGE-Ec1 (Addgene accession: 476

138474) were constructed by replacing the Redβ open reading frame (ORF) of pORTMAGE311B 477

plasmid (Addgene accession: 120418)43 with PapRecT and CspRecT respectively. pORTMAGE-Pa1 478

was constructed in many steps: i.) the Kanamycin resistance cassette and the RSF1010 origin-of-479

replication on pORTMAGE312B with Gentamicin resistance marker and pBBR1 origin-of-480

replication, amplified from pSEVA63173 (gift from Prof. Victor de Lorenzo - CSIC, Spain), ii.) 481

optimization of RBSs in pORTMAGE-Pa1 was done by designing a 30-nt optimal RBS in front of 482

the SSAP ORF and in between the SSAP and MutL ORFs with an automated design program, De 483

Novo DNA74, iii.) PaMutL was amplified from Pseudomonas aeruginosa genomic DNA and cloned 484

in place of EcMutL_E32K, and finally iv.) PaMutL was mutated by site-directed mutagenesis to 485

encode E36K. Ssr and Rec2 were ordered as gblocks from IDT and cloned in place of PapRecT into 486

earlier versions of pORTMAGE-Pa1 for the comparisons in Figure S6. 487

Measuring recombineering efficiency in Gammaproteobacteria by selective plating 488


https://doi.org/10.1101/2020.01.14.906594

Oligos were designed to introduce I) premature STOP codons into lacZ for E. coli, K. pneumoniae, 489

and C. freundii, or II) RpsL K43→R; GyrA T83→I; ParC S83→L; RpoB D521→V, or a premature 490

STOP codon into nfxB for P. aeruginosa. Oligo-mediated recombineering was performed as 491

described above on all bacterial strains. After recovery overnight, cells were plated at empirically-492

determined dilutions to a density of 200-500 colonies per plate. In the case of LacZ screening, 493

plating was assayed on MacConkey agar plates or on X-Gal/IPTG LBL agar plates in the case of K. 494

pneumoniae. In the case of selective antibiotic screening, cultures were plated onto both 495

selective and non-selective plates. Selective antibiotic concentrations used were the same as 496

those described for the selective testing above, except that in P. aeruginosa 100 µg/ml STR and 497

1.5 µg/ml CIP were used unless otherwise noted. Variants that were resistant to multiple 498

antibiotics were selected on LBL agar plates that contained the combination of all corresponding 499

antibiotics. Non-selective plates were antibiotic-free LBL agar plates. In all cases, allelic-500

replacement frequencies were calculated by dividing the number of recombinant CFUs by the 501

number of total CFUs. Plasmid maintenance was ensured by supplementing all media and agar 502

plates with either KAN (50 µg/ml) or GEN (20 µg/ml). 503

Minimum Inhibitory Concentration (MIC) measurement in P. aeruginosa 504

MICs were determined using a standard serial broth microdilution technique according to the 505

CLSI guidelines (ISO 20776-1:2006, Part 1: Reference method for testing the in vitro activity of 506

antimicrobial agents against rapidly growing aerobic bacteria involved in infectious diseases). 507

Briefly, bacterial strains were inoculated from frozen cultures onto MHBII agar plates and were 508

grown overnight at 37 °C. Next, independent colonies from each strain were inoculated into 1 ml 509

MHBII medium and were propagated at 37 °C, 250 rpm overnight. To perform MIC tests, 12-step 510


https://doi.org/10.1101/2020.01.14.906594

serial dilutions using 2-fold dilution-steps of the given antibiotic were generated in 96-well 511

microtiter plates (Sarstedt 96-well microtest plate). Antibiotics were diluted in 100 μl of MHBII 512

medium. Following dilutions, each well was seeded with an inoculum of 5×104 bacterial cells. 513

Each measurement was performed in 3 parallel replicates. Plates were incubated at 37 °C under 514

continuous shaking at 150 rpm for 18 hours in an INFORS HT shaker. After incubation, the OD600 515

of each well was measured using a Biotek Synergy 2 microplate reader. MIC was defined as the 516

antibiotic concentration which inhibited the growth of the bacterial culture, i.e., the drug 517

concentration where the average OD600 increment of the three replicates was below 0.05. 518

519

Acknowledgements 520

We thank Erkin Kuru, Max Schubert, Aditya Kunjapur, and Devon Stork for helpful discussions, 521

and Prof. Victor de Lorenzo (Centro Nacional de Biotecnologia, CSIC, Spain) for providing plasmids 522

for the study. Claire O’Callaghan, and Verena Volf provided experimental help in various 523

capacities in closely related work. John Aach was helpful for thinking through conceptual ideas. 524

Funding for this research was graciously provided by the DOE under grant DE-FG02-02ER63445 525

(to G.M.C), the European Research Council H2020-ERC-2014-CoG 648364 ‘Resistance Evolution’ 526

(to C.P.), GINOP (MolMedEx TUMORDNS) GINOP-2.3.2-15-2016-00020, GINOP (EVOMER) GINOP-527

2.3.2-15-2016-00014 (to C.P.); the ‘Lendület’ Program of the Hungarian Academy of Sciences (to 528

C.P.), and a PhD fellowship from the Boehringer Ingelheim Fonds and an EMBO Long-Term 529

Fellowship (to Á.N.). M.C. was supported by the Szeged Scientists Academy under the 530

sponsorship of the Hungarian Ministry of Human Capacities (EMMI: 13725-2/2018/INTFIN), by 531

the UNKP-18-2 New National Excellence Program of the Hungarian Ministry of Human Capacities 532


https://doi.org/10.1101/2020.01.14.906594

and UNKP 10-2 New National Excellence Program of the Hungarian Ministry for Innovation and 533

Technology. 534

Competing financial interests 535

GMC has related financial interests in EnEvolv, GRO Biosciences, and 64-x. For a complete list of 536

GMC’s financial interests, please visit http://arep.med.harvard.edu/gmc/tech.html. GMC, CJG, 537

MJL, and XR have submitted a patent application relating to pieces of this work 538

(WO2017184227A2). T.W., G.F., and G.C. have submitted a patent application related to the 539

improved SSAP variants referenced here. A.N. and C.P. have submitted a patent application 540

related to DIvERGE (PCT/EP2017/082574 (WO2018108987) Mutagenizing Intracellular Nucleic 541

Acids). 542

543


https://doi.org/10.1101/2020.01.14.906594

References 544

1. Wang, K. et al. Defining synonymous codon compression schemes by genome recoding. 545

Nature 539, 59–64 (2016). 546

2. Bird, A. W. et al. High-efficiency counterselection recombineering for site-directed 547

mutagenesis in bacterial artificial chromosomes. Nat. Methods 9, 103–109 (2011). 548

3. Wang, H. H. et al. Programming cells by multiplex genome engineering and accelerated 549

evolution. Nature 460, 894–898 (2009). 550

4. Ellis, H. M., Yu, D., DiTizio, T. & Court, D. L. High efficiency mutagenesis, repair, and 551

engineering of chromosomal DNA using single-stranded oligonucleotides. Proc. Natl. Acad. 552

Sci. U. S. A. 98, 6742–6746 (2001). 553

5. Yu, D. et al. An efficient recombination system for chromosome engineering in Escherichia 554

coli. Proc. Natl. Acad. Sci. 97, 5978–5983 (2000). 555

6. Zhang, Y., Buchholz, F., Muyrers, J. P. & Stewart, A. F. A new logic for DNA engineering using 556

recombination in Escherichia coli. Nat. Genet. 20, 123–128 (1998). 557

7. Costantino, N. & Court, D. L. Enhanced levels of lambda Red-mediated recombinants in 558

mismatch repair mutants. Proc. Natl. Acad. Sci. U. S. A. 100, 15748–15753 (2003). 559

8. Lajoie, M. J., Gregg, C. J., Mosberg, J. A., Washington, G. C. & Church, G. M. Manipulating 560

replisome dynamics to enhance lambda Red-mediated multiplex genome engineering. 561

Nucleic Acids Res. 40, e170 (2012). 562

9. Grogan, D. W. & Stengel, K. R. Recombination of synthetic oligonucleotides with prokaryotic 563

chromosomes: substrate requirements of the Escherichia coli/lambdaRed and Sulfolobus 564

acidocaldarius recombination systems. Mol. Microbiol. 69, 1255–1265 (2008). 565


https://doi.org/10.1101/2020.01.14.906594

10. Farzadfard, F. & Lu, T. K. Genomically encoded analog memory with precise in vivo DNA 566

writing in living cell populations. Science 346, 1256272 (2014). 567

11. Mandell, D. J. et al. Biocontainment of genetically modified organisms by synthetic protein 568

design. Nature 518, 55–60 (2015). 569

12. Wang, H. H. et al. Genome-scale promoter engineering by coselection MAGE. Nat. Methods 570

9, 591–593 (2012). 571

13. Nyerges, Á. et al. Directed evolution of multiple genomic loci allows the prediction of 572

antibiotic resistance. Proc. Natl. Acad. Sci. U. S. A. 115, E5726–E5735 (2018). 573

14. Wannier, T. M. et al. Adaptive evolution of genomically recoded Escherichia coli. Proc. Natl. 574

Acad. Sci. U. S. A. 115, 3090–3095 (2018). 575

15. Sun, Z. et al. A high-efficiency recombineering system with PCR-based ssDNA in Bacillus 576

subtilis mediated by the native phage recombinase GP35. Appl. Microbiol. Biotechnol. 99, 577

5151–5162 (2015). 578

16. Binder, S., Siedler, S., Marienhagen, J., Bott, M. & Eggeling, L. Recombineering in 579

Corynebacterium glutamicum combined with optical nanosensors: a general strategy for 580

fast producer strain generation. Nucleic Acids Res. 41, 6360–6369 (2013). 581

17. Chang, Y., Wang, Q., Su, T. & Qi, Q. The efficiency for recombineering is dependent on the 582

source of the phage recombinase function unit. bioRxiv doi:10.1101/745448. 583

18. Lopes, A., Amarir-Bouhram, J., Faure, G., Petit, M.-A. & Guerois, R. Detection of novel 584

recombinases in bacteriophage genomes unveils Rad52, Rad51 and Gp2.5 remote 585

homologs. Nucleic Acids Res. 38, 3952–3962 (2010). 586


https://doi.org/10.1101/2020.01.14.906594

19. Datta, S., Costantino, N., Zhou, X. & Court, D. L. Identification and analysis of 587

recombineering functions from Gram-negative and Gram-positive bacteria and their 588

phages. Proc. Natl. Acad. Sci. U. S. A. 105, 1626–1631 (2008). 589

20. Guo, T., Xin, Y., Zhang, Y., Gu, X. & Kong, J. A rapid and versatile tool for genomic 590

engineering in Lactococcus lactis. Microb. Cell Factories 18, (2019). 591

21. van Pijkeren, J.-P. & Britton, R. A. High efficiency recombineering in lactic acid bacteria. 592

Nucleic Acids Res. 40, e76 (2012). 593

22. van Pijkeren, J.-P., Neoh, K. M., Sirias, D., Findley, A. S. & Britton, R. A. Exploring 594

optimization parameters to increase ssDNA recombineering in Lactococcus lactis and 595

Lactobacillus reuteri. Bioengineered 3, 209–217 (2012). 596

23. van Kessel, J. C. & Hatfull, G. F. Recombineering in Mycobacterium tuberculosis. Nat. 597

Methods 4, 147–152 (2007). 598

24. van Kessel, J. C. & Hatfull, G. F. Efficient point mutagenesis in mycobacteria using single-599

stranded DNA recombineering: characterization of antimycobacterial drug targets. Mol. 600

Microbiol. 67, 1094–1107 (2008). 601

25. Yin, J. et al. Single-Stranded DNA-Binding Protein and Exogenous RecBCD Inhibitors Enhance 602

Phage-Derived Homologous Recombination in Pseudomonas. iScience 14, 1–14 (2019). 603

26. Ricaurte, D. E. et al. A standardized workflow for surveying recombinases expands bacterial 604

genome-editing capabilities. Microb. Biotechnol. 11, 176–188 (2018). 605

27. Aparicio, T., de Lorenzo, V. & Martínez-García, E. A Broad Host Range Plasmid-Based 606

Roadmap for ssDNA-Based Recombineering in Gram-Negative Bacteria. Methods Mol. Biol. 607

Clifton NJ 2075, 383–398 (2020). 608


https://doi.org/10.1101/2020.01.14.906594

28. Aparicio, T. et al. Mismatch repair hierarchy of Pseudomonas putida revealed by mutagenic 609

ssDNA recombineering of the pyrF gene. Environ. Microbiol. (2019) doi:10.1111/1462-610

2920.14814. 611

29. Corts, A. D., Thomason, L. C., Gill, R. T. & Gralnick, J. A. A new recombineering system for 612

precise genome-editing in Shewanella oneidensis strain MR-1 using single-stranded 613

oligonucleotides. Sci. Rep. 9, 1–10 (2019). 614

30. Penewit, K. et al. Efficient and Scalable Precision Genome Editing in Staphylococcus aureus 615

through Conditional Recombineering and CRISPR/Cas9-Mediated Counterselection. mBio 9, 616

e00067 (2018). 617

31. Aparicio, T., Nyerges, A., Martínez-García, E. & Lorenzo, V. de. High-efficiency multi-site 618

genomic editing (HEMSE) of Pseudomonas putida through thermoinducible ssDNA 619

recombineering. bioRxiv 851576 (2019) doi:10.1101/851576. 620

32. Maresca, M. et al. Single-stranded heteroduplex intermediates in lambda Red homologous 621

recombination. BMC Mol. Biol. 11, 54 (2010). 622

33. Mosberg, J. A., Lajoie, M. J. & Church, G. M. Lambda red recombineering in Escherichia coli 623

occurs through a fully single-stranded intermediate. Genetics 186, 791–799 (2010). 624

34. Murphy, K. C. Use of bacteriophage lambda recombination functions to promote gene 625

replacement in Escherichia coli. J. Bacteriol. 180, 2063–2071 (1998). 626

35. Iyer, L. M., Koonin, E. V. & Aravind, L. Classification and evolutionary history of the single-627

strand annealing proteins, RecT, Redbeta, ERF and RAD52. BMC Genomics 3, 8 (2002). 628


https://doi.org/10.1101/2020.01.14.906594

36. Story, R. M., Bishop, D. K., Kleckner, N. & Steitz, T. A. Structural relationship of bacterial 629

RecA proteins to recombination proteins from bacteriophage T4 and yeast. Science 259, 630

1892–1896 (1993). 631

37. Sullivan, M. B. et al. The genome and structural proteome of an ocean siphovirus: a new 632

window into the cyanobacterial ‘mobilome’. Environ. Microbiol. 11, 2935–2951 (2009). 633

38. Nyerges, Á. et al. A highly precise and portable genome engineering method allows 634

comparison of mutational effects across bacterial species. Proc. Natl. Acad. Sci. U. S. A. 113, 635

2502–2507 (2016). 636

39. Pines, G., Freed, E. F., Winkler, J. D. & Gill, R. T. Bacterial Recombineering: Genome 637

Engineering via Phage-Based Homologous Recombination. ACS Synth. Biol. 4, 1176–1185 638

(2015). 639

40. Piddock, L. J. Mechanisms of fluoroquinolone resistance: an update 1994-1998. Drugs 58 640

Suppl 2, 11–18 (1999). 641

41. van Kessel, J. C., Marinelli, L. J. & Hatfull, G. F. Recombineering mycobacteria and their 642

phages. Nat. Rev. Microbiol. 6, 851–857 (2008). 643

42. Tommasi, R., Brown, D. G., Walkup, G. K., Manchester, J. I. & Miller, A. A. ESKAPEing the 644

labyrinth of antibacterial discovery. Nat. Rev. Drug Discov. 14, 529–542 (2015). 645

43. Szili, P. et al. Antibiotic usage promotes the evolution of resistance against gepotidacin, a 646

novel multi-targeting drug. bioRxiv 495630 (2018) doi:10.1101/495630. 647

44. Yang, C. et al. Fed-batch fermentation of recombinant Citrobacter freundii with expression 648

of a violacein-synthesizing gene cluster for efficient violacein production from glycerol. 649

Biochem. Eng. J. 57, 55–62 (2011). 650


https://doi.org/10.1101/2020.01.14.906594

45. Jiang, P. et al. Pathway redesign for deoxyviolacein biosynthesis in Citrobacter freundii and 651

characterization of this pigment. Appl. Microbiol. Biotechnol. 94, 1521–1532 (2012). 652

46. Szili, P. et al. Rapid evolution of reduced susceptibility against a balanced dual-targeting 653

antibiotic through stepping-stone mutations. Antimicrob. Agents Chemother. AAC.00207-19 654

(2019) doi:10.1128/AAC.00207-19. 655

47. Honda, Y. et al. Functional division and reconstruction of a plasmid replication origin: 656

molecular dissection of the oriV of the broad-host-range plasmid RSF1010. Proc. Natl. Acad. 657

Sci. U. S. A. 88, 179–183 (1991). 658

48. Gawin, A., Valla, S. & Brautaset, T. The XylS/Pm regulator/promoter system and its use in 659

fundamental studies of bacterial gene expression, recombinant protein production and 660

metabolic engineering. Microb. Biotechnol. 10, 702 (2017). 661

49. Szpirer, C. Y., Faelen, M. & Couturier, M. Mobilization function of the pBHR1 plasmid, a 662

derivative of the broad-host-range plasmid pBBR1. J. Bacteriol. 183, 2101–2110 (2001). 663

50. Aparicio, T., Jensen, S. I., Nielsen, A. T., de Lorenzo, V. & Martínez-García, E. The Ssr protein 664

(T1E_1405) from Pseudomonas putida DOT-T1E enables oligonucleotide-based 665

recombineering in platform strain P. putida EM42. Biotechnol. J. 11, 1309–1319 (2016). 666

51. Marvig, R. L., Sommer, L. M., Molin, S. & Johansen, H. K. Convergent evolution and 667

adaptation of Pseudomonas aeruginosa within patients with cystic fibrosis. Nat. Genet. 47, 668

57–64 (2015). 669

52. AbdulWahab, A. et al. The emergence of multidrug-resistant Pseudomonas aeruginosa in 670

cystic fibrosis patients on inhaled antibiotics. Lung India Off. Organ Indian Chest Soc. 34, 671

527–531 (2017). 672


https://doi.org/10.1101/2020.01.14.906594

53. Tacconelli, E. et al. Discovery, research, and development of new antibiotics: the WHO 673

priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect. Dis. 18, 318–327 674

(2018). 675

54. Agnello, M. & Wong-Beringer, A. The use of oligonucleotide recombination to generate 676

isogenic mutants of clinical isolates of Pseudomonas aeruginosa. J. Microbiol. Methods 98, 677

23–25 (2014). 678

55. Chen, W. et al. CRISPR/Cas9-based Genome Editing in Pseudomonas aeruginosa and 679

Cytidine Deaminase-Mediated Base Editing in Pseudomonas Species. iScience 6, 222–231 680

(2018). 681

56. Cabot, G. et al. Evolution of Pseudomonas aeruginosa Antimicrobial Resistance and Fitness 682

under Low and High Mutation Rates. Antimicrob. Agents Chemother. 60, 1767–1778 (2016). 683

57. Jatsenko, T., Tover, A., Tegova, R. & Kivisaar, M. Molecular characterization of Rif(r) 684

mutations in Pseudomonas aeruginosa and Pseudomonas putida. Mutat. Res. 683, 106–114 685

(2010). 686

58. PEW Charitable Trust. Antibiotics Currently in Global Clinical Development. 687

http://pew.org/1YkUFkT. 688

59. Sandoval, N. R. et al. Strategy for directing combinatorial genome engineering in Escherichia 689

coli. Proc. Natl. Acad. Sci. 109, 10540–10545 (2012). 690

60. Simon, A. J., Morrow, B. R. & Ellington, A. D. Retroelement-Based Genome Editing and 691

Evolution. ACS Synth. Biol. 7, 2600–2611 (2018). 692

61. Lajoie, M. J. et al. Genomically Recoded Organisms Expand Biological Functions. Science 693

342, 357–360 (2013). 694


https://doi.org/10.1101/2020.01.14.906594

62. Novais, Â. et al. Evolutionary Trajectories of Beta-Lactamase CTX-M-1 Cluster Enzymes: 695

Predicting Antibiotic Resistance. PLoS Pathog. 6, (2010). 696

63. Criswell, D., Tobiason, V. L., Lodmell, J. S. & Samuels, D. S. Mutations conferring 697

aminoglycoside and spectinomycin resistance in Borrelia burgdorferi. Antimicrob. Agents 698

Chemother. 50, 445–452 (2006). 699

64. Campbell, E. A. et al. Structural mechanism for rifampicin inhibition of bacterial rna 700

polymerase. Cell 104, 901–912 (2001). 701

65. Okamoto-Hosoya, Y., Hosaka, T. & Ochi, K. An aberrant protein synthesis activity is linked 702

with antibiotic overproduction in rpsL mutants of Streptomyces coelicolor A3(2). Microbiol. 703

Read. Engl. 149, 3299–3309 (2003). 704

66. Yoshida, H., Bogaki, M., Nakamura, M. & Nakamura, S. Quinolone resistance-determining 705

region in the DNA gyrase gyrA gene of Escherichia coli. Antimicrob. Agents Chemother. 34, 706

1271–1272 (1990). 707

67. Finn, R. D., Clements, J. & Eddy, S. R. HMMER web server: interactive sequence similarity 708

searching. Nucleic Acids Res. 39, W29–W37 (2011). 709

68. Eisen, A. & Camerini-Otero, R. D. A recombinase from Drosophila melanogaster embryos. 710

Proc. Natl. Acad. Sci. U. S. A. 85, 7481–7485 (1988). 711

69. El-Gebali, S. et al. The Pfam protein families database in 2019. Nucleic Acids Res. 47, D427–712

D432 (2019). 713

70. Huerta-Cepas, J., Serra, F. & Bork, P. ETE 3: Reconstruction, Analysis, and Visualization of 714

Phylogenomic Data. Mol. Biol. Evol. 33, 1635–1638 (2016). 715


https://doi.org/10.1101/2020.01.14.906594

71. Choe, W., Chandrasegaran, S. & Ostermeier, M. Protein fragment complementation in 716

M.HhaI DNA methyltransferase. Biochem. Biophys. Res. Commun. 334, 1233–1240 (2005). 717

72. Joshi, N. & Fass, J. Sickle: A sliding-window, adaptive, quality-based trimming tool for FastQ 718

files. (2019). 719

73. Martínez-García, E., Aparicio, T., Goñi-Moreno, A., Fraile, S. & de Lorenzo, V. SEVA 2.0: an 720

update of the Standard European Vector Architecture for de-/re-construction of bacterial 721

functionalities. Nucleic Acids Res. 43, D1183–D1189 (2015). 722

74. Salis, H. M., Mirsky, E. A. & Voigt, C. A. Automated Design of Synthetic Ribosome Binding 723

Sites to Precisely Control Protein Expression. Nat. Biotechnol. 27, 946–950 (2009). 724

75. Marcusson, L. L., Frimodt-Møller, N. & Hughes, D. Interplay in the selection of 725

fluoroquinolone resistance and bacterial fitness. PLoS Pathog. 5, e1000541 (2009). 726

727

728


https://doi.org/10.1101/2020.01.14.906594

Table 1 – Minimum inhibitory concentration (MIC) was measured for various combinations of 729

ciprofloxacin resistance-conferring alleles. GyrA_T83I displays strong positive epistasis with 730

ParC_S87L, and so clonal populations with mutations to parC but not gyrA were not pulled out 731

of our antibiotic selection75. 732

Genotype Ciprofloxacin MIC (µg/ml)

PAO1 wild-type 0.25 nfxB knockout 4

GyrA_T83I 16 GyrA_T83I + ParC_S87L 32

GyrA_T83I + ParC_S87L + NfxB

knockout >128

733

734


https://doi.org/10.1101/2020.01.14.906594

Figures 735

736

Figure 1. Serial Evolutionary Enhancement of Recombineering (SEER) workflow. The SEER 737

workflow is depicted across the top from left to right, with libraries being first assembled, then 738

moved into a chassis organism, enriched over 3-10 cycles for efficient recombineering proteins, 739

and finally analyzed by deep sequencing. This process can then be iterated on by learning from 740

results and making improvements to the library design. The specific selective enrichment strategy 741

that we designed for E. coli is shown in a gray callout bubble. Five successive antibiotic selections 742

were applied to a library of E. coli cells expressing SSAP variants, and after the selective handles 743

were exhausted the plasmid library was extracted and re-transformed into the naïve SEER chassis 744

for five further cycles of selection. 745

746

Vector Backbone

SSAPs + barcode

+

Plasmid Library

Transform

SEERChassis

Naïve Library Enriched Library

Library Prep PCRand Indexing

An

alysis and

Lib

rary Desig

n

1. tolCmut�ĺ�WRO&WT / SDS selection2. rpsLWT�ĺ�USV/K43R / STM selection3. gyrAWT�ĺ�J\U$S83L / CIP selection4. catmut�ĺ�FDWWT / CHL selection

5. rpoBWT�ĺ�USR%S512P / RIF selection SEER

Chassis

1x Rep

eat

Library Mini-PrepsTransform

Assemble3 -10 cycles of

selective enrichment

Deep Sequencing

E. coli selective enrichment strategy


https://doi.org/10.1101/2020.01.14.906594

747

Figure 2. Broad SSAP Library (A) Circles of various sizes represent the number of variants present 748

in Broad SSAP Library from different protein families, as categorized by the Pfam database. The 749

seven principle families of SSAPS are grouped into three clusters based on structural and 750

0 2 4 6 8 1010-3

10-2

10-1

100

101

102

Round of SEER

Fold

enr

ichm

ent o

ver

non-

sele

ctiv

e co

ntro

l

SR016 (PapRecT)SR120SR011SR095SR055

0 1 2 3 4 5 6 7 8 9 100

200

400

600

Round of SEER

Tota

l SS

AP

fam

ily e

nric

hmen

t ERFGp2.5Rad52

RecARecTSak

Sak4Unclassified

no SSAP

Red�

SR055

SR095

SR011

SR120

SR016 (

PapRec

T)0.0

0.1

0.2

0.3

0.4

Edi

ting

Eff

icie

ncy

*****

ns

**** ****

0.00001 0.0001 0.001 0.01 0.1 10.01

0.1

1

10

100

Broad SSAP Library Frequency

Bro

ad S

SA

P Li

brar

y E

nric

hmen

t

Actinobacteria

Firmicutes

Proteobacteria

Bacilli

BetaproteobacteriaClostridiaCorynebacteriales

Gammaproteobacteria

Bacillales

BurkholderialesClostridialesEnterobacterales

Lactobacillales

Pseudomonadales

EnterobacteriaceaeLactobacillaceae

Paenibacillaceae

StaphylococcaceaeStreptococcaceae

Phyla Class

Order Family

a

c

e

Gp2.5

Candidate Proteins

Rad51 Cluster

Unclassified

Rad52 Cluster

Protein Family Representation in Broad SSAP Library Phylogenetic Distribution of Broad SSAP Libraryb

d

f

SR011

SR016


https://doi.org/10.1101/2020.01.14.906594

phylogenetic information as proposed by Lopes et al. (2010). (B) The phylogenetic distribution of 751

Broad SSAP Library is represented as vertical bar charts. For each phylogenetic level, any group 752

that represents more than 5% of the total library is called out to the right of the bar. (C) The 753

enrichment of each Broad SSAP Library member is plotted over ten successive rounds of 754

selection. Enrichment of each library member was calculated by dividing the average frequency 755

across selective replicates by the frequency of the non-selective control. Frequency data is 756

measured by amplification of the barcoded region of the plasmid library and next generation 757

sequencing (NGS). (D) Total enrichment is plotted for each protein family over the ten rounds of 758

SEER. (E) Frequency is plotted against enrichment for each Broad SSAP Library member after the 759

tenth round of selection. Five candidate proteins, which are shaded in a yellow box, were 760

selected for further characterization. (F) Five top candidates were tested for their efficiency at 761

incorporating a single-base-pair silent mutation at a non-essential gene, ynfF. Efficiency was read 762

out by NGS. Significance values are indicated for a parametric t-test between two groups, where 763

ns, *, **, ***, and ***** indicate p > 0.05, p < 0.05, p < 0.01, p < 0.001, and p < 0.0001 764

respectively. 765

766


https://doi.org/10.1101/2020.01.14.906594

767

Figure 3. Broad RecT Library and CspRecT (A) Frequency is plotted against enrichment for each 768

Broad RecT Library member after the tenth round of selection. One candidate protein, CspRecT 769

(shaded box), was the standout winner. In all subsequent panels, Redβ, PapRecT, and CspRecT 770

are compared when expressed from a pORTMAGE-based construct (Figure S1) in wild-type 771

MG1655 E. coli. Significance values are indicated for a grouped parametric t-test, where ns and 772

***** indicate p > 0.05 and p < 0.0001 respectively. Editing efficiency was measured by 773

blue/white screening at the LacZ locus for (B) eight different single-base mismatches (n=3) and 774

(C) 18-base and 30-base mismatches (n=3). (D) MAGE editing targeting 1, 5, 10, 15, or 20 genomic 775

loci at once in triplicate, was read out by NGS. The solid lines represent the mean editing 776

efficiency across all targeted loci, while the dashed lines represent the sum of all single-locus 777

efficiencies, which we refer to as aggregate efficiency. (E) A 130-oligo DIvERGE experiment using 778

oligos that were designed to tile four different genomic loci that encode the drug targets of 779

0.0001 0.001 0.01 0.1 10.01

0.1

1

10

100

Broad RecT Library Frequency

Bro

ad R

ecT

Libr

ary

Enr

ichm

ent

250 500 10000

20

40

60

80

Cip (ng/ml)

Res

ista

nt C

FUs

Red�PapRecTCspRecT

nd nd

0 5 10 15 200.01

0.1

1

Number of genetic loci targeted

Ave

rage

Edi

ting

Eff

icie

ncy CspRecT

Red�

PapRecTAggregate

editing efficiency

Single locus editing

efficiency

CspRecT

Red�

PapRecT

A:AG:G T:TA:GC:A C:TG:A G:T A:AG:G T:TA:GC:A C:T

G:A G:T A:AG:G T:TA:GC:A C:TG:A G:T

0.0

0.2

0.4

0.6

Edi

ting

Eff

icie

ncy

ns

****

Red� PapRecT CspRecT

18nt mismatch 30nt mismatch0.00

0.02

0.04

0.06

0.08

Edi

ting

Eff

icie

ncy

Red�PapRecTCspRecT

a b

c d

Candidate Protein

e


https://doi.org/10.1101/2020.01.14.906594

fluoroquinolone antibiotics and are known hotspots for CIP resistance. The oligos contained 1.5% 780

degeneracy at each nucleotide position along their entire length. All 130 oligos were mixed and 781

transformed together into cells (n=3). Colony forming units were measured at three different CIP 782

concentrations after plating 1/100th of the final recovery volume. 783

784


https://doi.org/10.1101/2020.01.14.906594

785

Figure 4. Recombineering in Gammaproteobacteria. (A) Recombineering experiments were run 786

with Redβ, PapRecT, and CspRecT expressed off of the pORTMAGE311B backbone, or with a 787

pBBR1 origin in the case of P. aeruginosa. Editing efficiency was measured by colony counts on 788

selective vs. non-selective plates (n=3; see methods). Vector optimization resulted in improved 789

efficiency of PapRecT in P. aeruginosa (see Figure S7) (B) Diagram of a simple multi-drug 790

resistance experiment in P. aeruginosa harboring an optimized PapRecT plasmid expression 791

system, pORTMAGE-Pa1. In a single round of MAGE, a pool of five oligos was used to incorporate 792

genetic modifications that would provide resistance to STR, RIF, and CIP (n=3). These populations 793

were then selected by plating on all combinations of 1-, 2-, or 3-antibiotic agarose plates and 794

0.000

001

0.000

01

0.000

10.0

01 0.01 0.1 1

E. coli

C. freundii

K. pneumoniae

P. aeruginosa

Editing Efficiency

BetaPapRecTCspRecT

Str RifCip

Rif + Str

Cip + Str

Cip + Rif

Cip + Rif +

Str0.00001

0.0001

0.001

0.01

0.1

P. a

erug

inos

a Ed

iting

Effi

cien

cyExpected Multilocus Efficiency

Observed

a

cb

Plasmid Optimization


https://doi.org/10.1101/2020.01.14.906594

compared with a non-selective control. (C) Observed efficiencies were calculated by comparing 795

colony counts on selective vs. non-selective plates. Expected efficiencies for multi-locus events 796

were calculated as the product of all relevant single-locus efficiencies. 797

798


https://doi.org/10.1101/2020.01.14.906594

Supplementary Figures 799

800

Figure S1 – Vector maps. (A) pARC8-DEST was created to have a pBAD regulatory region, beta 801

lactamase, a p15a origin, and a lethal ccdB gene flanked by attR sites for Gateway cloning. 802

Introduction by the LR Gateway reaction of for instance SR001, would create the vector on the 803

right, with an arabinose-inducible SR001 followed by a barcode. (B) Two pORTMAGE vectors are 804

provided for broad-spectrum recombineering. pORTMAGE-Ec1 was demonstrated effective in E. 805

coli, C. freundii, and K. pneumoniae, while pORTMAGE-Pa1 was demonstrated effective in P. 806

aeruginosa. 807

Chm

Resistance

ccdB

attR2

attR1pBAD-araC

regulatory region

araC

Car

Resistancep15a origin

pARC8-DEST

Pink region removed when introducing genesby Gateway Cloning

LR Gateway

reaction

(e.g. + SR001)

attB2

attB1pBAD-araC

regulatory region

araC

Car

Resistance p15a origin

pARC8 +

sample SEER

library member

SR001

SR001

Barcode

a

b


https://doi.org/10.1101/2020.01.14.906594

808

809

Figure S2 – Antibiotic testing in E. coli. Candidate oligos used to confer resistance to the five 810

antibiotic selections used in the SEER chassis were tested for their efficacy at varied 811

concentrations of antibiotic. The number of CFUs is plotted against antibiotic concentration for 812

each antibiotic. Oligos that were used in SEER selection are colored light blue. The antibiotic 813

concentration chosen for the SEER selection is indicated by a vertical dashed line. (A) SDS 814

resistance was restored by correcting a stop codon insertion in tolC. (B) Ciprofloxacin, (C) 815

Rifampicin, and (D) Streptomycin resistance were imparted with a mutation to native E. coli 816

genes. (E) Chloramphenicol resistance was restored by correcting a stop codon insertion in cat. 817

A B

C D

E

0 50 100 150

1

10

100

1000

10000

Rif (�g/ml)

# C

olon

ies

(pla

ted

at 1

0-1)

Rifampicin resistance

Water Blank

Oligo 398

Oligo 399

Oligo 403

Oligo 404

0.001 0.01 0.1 1

1

10

100

1000

10000

SDS (% v/v)

# C

olon

ies

(pla

ted

at 1

0-1)

SDS resistance

Water Blank

Oligo 307

0.0 0.2 0.4 0.6

1

10

100

1000

10000

Cip (�g/ml)

# C

olon

ies

(pla

ted

at 1

0-1)

Ciprofloxacin resistance

Water Blank

Oligo 402

0 20 40 60

1

10

100

1000

10000

Stm (�g/ml)#

Col

onie

s (p

late

d at

10-1

)

Streptomycin resistance

Water Blank

Oligo 400

0 10 20 30 40 50

1

10

100

1000

10000

Chm (�g/ml)

# C

olon

ies

(pla

ted

at 1

0-1)

Chloramphenicol resistance

Water Blank

Oligo 460


https://doi.org/10.1101/2020.01.14.906594

818


https://doi.org/10.1101/2020.01.14.906594

819

Figure S3 – (Top) the frequency of each library member from the Broad SSAP Library is plotted 820

across ten rounds of SEER. Data are measured by amplification of the barcoded region of the 821

plasmid library and next generation sequencing (NGS). (Bottom) The frequency of SSAP families 822

is shown across the same selective cycles on the Broad SSAP Library. 823

0 2 4 6 8 1010-6

10-5

10-4

10-3

10-2

10-1

100

Round of SEER

Ave

rage

freq

uenc

y

SR016 (PapRecT)

SR120SR011

SR095SR055

0 1 2 3 4 5 6 7 8 9 100.0

0.5

1.0

Round of SEER

Ave

rage

SS

AP

fam

ily fr

eque

ncy

ERFGp2.5Rad52RecARecTSakSak4Unclassified


https://doi.org/10.1101/2020.01.14.906594

824

Figure S4 – The effect of codon-usage on Redβ editing efficiency in E. coli. The efficiency of 825

Redβ from the Broad SSAP Library was compared with Redβ expressed off of its wild-type 826

codons. Efficiency of making a single base pair mutation in a non-coding gene (silent mismatch 827

MAGE oligo 7) was measured by NGS. 828

829

Red�

Red� [

WT co

dons]0.00

0.05

0.10

0.15

0.20

0.25

Edi

ting

Eff

icie

ncy


https://doi.org/10.1101/2020.01.14.906594

830

Figure S5 – Frequency and Enrichment of members of Broad RecT Library over ten rounds of 831

SEER enrichment. 832

833

b

0 2 4 6 8 100.000001

0.00001

0.0001

0.001

0.01

0.1

1

Round of SEER

Ave

rage

freq

uenc

yov

er th

ree

sele

ctiv

e re

plic

ates

Library S2 SSAP Frequency

PF020 (CspRecT)

PF077

PF047

PF076

0 2 4 6 8 100.001

0.01

0.1

1

10

100

1000

Round of SEER

Fold

enr

ichm

ent o

ver

non-

sele

ctiv

e co

ntro

l

Library S2 SSAP Enrichment

PF020 (CspRecT)

PF077PF047PF076

a


https://doi.org/10.1101/2020.01.14.906594

834

Figure S6 – Recombineering efficiency in P. aeruginosa was measured for PapRecT with E. coli 835

codons, PapRecT with its wild-type codons, and two SSAPs that have been reported to work in 836

Pseudomonas putida. This was measured both with the original pORTMAGE311B RBS and an 837

RBS optimized for P. aeruginosa. Significance values are indicated for a parametric t-test 838

between two groups, where ns, *, **, ***, and ***** indicate p > 0.05, p < 0.05, p < 0.01, p < 839

0.001, and p < 0.0001 respectively. 840

PapRec

T

PapRec

T_WT

ssr

Rec2

PapRec

T_WT (R

BSopt)

ssr (

RBSopt)

Rec2 (

RBSopt)0.00

0.05

0.10

0.15

0.20

P. a

erug

inos

a E

ditin

g E

ffic

ienc

y

**

ns**

ns**

*

Original RBS Optimized RBS


https://doi.org/10.1101/2020.01.14.906594

841

Figure S7 – Editing efficiency in making a single-base mutation at the rpsL locus in P. aeruginosa 842

with various plasmid variants expressing PapRecT. An unoptimized plasmid (far left) was 843

constructed by replacing, in pORTMAGE312B (Addgene accession: 128969), the RSF1010 origin 844

of replication and the kanamycin resistance gene with a pBBR1 origin of replication and a 845

gentamicin resistance gene. The best-performing plasmid variant (third from right) was 846

renamed pORTMAGE-Pa1 (Addgene accession: 138475). Constructs examining the role of MutL 847

in single-base recombineering efficiency were made by first restoring wild-type PaMutL and 848

then by removing it entirely (second from right, and far right respectively). 849

0.001

0.01

0.1

1

P. a

erug

inos

a Ed

iting

Effi

cien

cy

Optimized PapRecT RBS

PapRecT WTcodons

PapRecT ECcodons

EcMutL_E32K

PaMutL_E36K

PaMutL

Optimized MutL RBS

- + + + + + +

- - - - + + +

- - + + + + +

+ + - - - - -

+ + + - - - -

- - - + + - -

- - - - - + -

RBS optimization

PapRecT codons

MutL variant


https://doi.org/10.1101/2020.01.14.906594

850

Figure S8 – One round of MAGE with a pool of three oligos that confer Ciprofloxacin resistance 851

was conducted in P. aeruginosa with pORTMAGE-Pa1. Editing efficiency is shown after plating on 852

three different concentrations of antibiotic. 853

854

0.001 0.0

1 0.1 1

Cip0.5

Cip1.5

Cip4.5

P. aeruginosa Editing Efficiency

Oligo Pool: PAgyrA_T83I (CIP)PAparC_S87L (CIP)PAnfxB_KO (CIP)


https://doi.org/10.1101/2020.01.14.906594

855

Figure S9 – Phylogenetic tree of the Pfam RecT protein family. The Broad RecT Library was 856

generated from the full alignment of Pfam family PF03837, containing 576 sequences from 857

Pfam 31.069, as generated by FastTree. Using ETE 3, the phylogenetic tree was pruned and a 858

maximum diversity subtree of 100 members was identified70. Members of this 100-member 859

subtree are indicated in blue. 860


https://doi.org/10.1101/2020.01.14.906594

861

Figure S10 – The growth rates of bacteria expressing each of the top five candidates from Broad 862

SSAP Library or Redβ were measured by plate-reader growth assay and plotted against the 863

maximum attained OD600 of the culture. 864

1.2 1.4 1.6 1.8 2.0 2.2 2.41.0

1.1

1.2

1.3

1.4

1.5

1.6

Growth Rate (�)

Max

OD

600

SR011

SR016 (PapRecT)

SR055

Red�

SR095

SR120


https://doi.org/10.1101/2020.01.14.906594

Vector Backbone

SSAPs + barcode

+

Plasmid Library

Transform

SEERChassis

Naïve Library Enriched Library

Library Prep PCRand Indexing

Analysis and Library D

esign

1. tolCmut → tolCWT / SDS selection2. rpsLWT → rpsLK43R / STM selection3. gyrAWT → gyrAS83L / CIP selection4. catmut → catWT / CHL selection

5. rpoBWT → rpoBS512P / RIF selection SEER

Chassis

1x Repeat

Library Mini-PrepsTransform

Assemble3 -10 cycles of

selective enrichment

Deep Sequencing

E. coli selective enrichment strategy


https://doi.org/10.1101/2020.01.14.906594

0 2 4 6 8 1010-3

10-2

10-1

100

101

102

Round of SEER

Fold

enr

ichm

ent o

ver

non-

sele

ctiv

e co

ntro

l

SR016 (PapRecT)SR120SR011SR095SR055

0 1 2 3 4 5 6 7 8 9 100

200

400

600

Round of SEER

Tota

l SS

AP

fam

ily e

nric

hmen

t ERFGp2.5Rad52

RecARecTSak

Sak4Unclassified

no SSAP

Red!

SR055

SR095

SR011

SR120

SR016 (

PapRec

T)0.0

0.1

0.2

0.3

0.4

Edi

ting

Eff

icie

ncy

*****

ns

**** ****

0.00001 0.0001 0.001 0.01 0.1 10.01

0.1

1

10

100

Broad SSAP Library Frequency

Bro

ad S

SA

P Li

brar

y E

nric

hmen

t

Actinobacteria

Firmicutes

Proteobacteria

Bacilli

BetaproteobacteriaClostridiaCorynebacteriales

Gammaproteobacteria

Bacillales

BurkholderialesClostridialesEnterobacterales

Lactobacillales

Pseudomonadales

EnterobacteriaceaeLactobacillaceae

Paenibacillaceae

StaphylococcaceaeStreptococcaceae

Phyla Class

Order Family

a

c

e

Gp2.5

Candidate Proteins

Rad51 Cluster

Unclassified

Rad52 Cluster

Protein Family Representation in Broad SSAP Library Phylogenetic Distribution of Broad SSAP Libraryb

d

f

SR011

SR016


https://doi.org/10.1101/2020.01.14.906594

0.0001 0.001 0.01 0.1 10.01

0.1

1

10

100

Broad RecT Library Frequency

Bro

ad R

ecT

Libr

ary

Enr

ichm

ent

250 500 10000

20

40

60

80

Cip (ng/ml)

Res

ista

nt C

FUs

Red!PapRecTCspRecT

nd nd

0 5 10 15 200.01

0.1

1

Number of genetic loci targeted

Ave

rage

Edi

ting

Eff

icie

ncy CspRecT

Red!

PapRecTAggregate

editing efficiency

Single locus editing

efficiency

CspRecT

Red!

PapRecT

A:AG:G T:TA:GC:A C:TG:A G:T A:AG:G T:TA:GC:A C:T

G:A G:T A:AG:G T:TA:GC:A C:TG:A G:T

0.0

0.2

0.4

0.6

Edi

ting

Eff

icie

ncy

ns

****

Red! PapRecT CspRecT

18nt mismatch 30nt mismatch0.00

0.02

0.04

0.06

0.08

Edi

ting

Eff

icie

ncy

Red!PapRecTCspRecT

a b

c d

Candidate Protein

e


https://doi.org/10.1101/2020.01.14.906594

0.000

001

0.000

01

0.000

10.0

01 0.01 0.1 1

E. coli

C. freundii

K. pneumoniae

P. aeruginosa

Editing Efficiency

BetaPapRecTCspRecT

Str RifCip

Rif + Str

Cip + Str

Cip + Rif

Cip + Rif +

Str0.00001

0.0001

0.001

0.01

0.1

P. a

erug

inos

a Ed

iting

Effi

cien

cyExpected Multilocus Efficiency

Observed

a

cb

Plasmid Optimization


https://doi.org/10.1101/2020.01.14.906594

ChmResistance

ccdB

attR2

attR1pBAD-araCregulatory region

araC

CarResistance

p15a origin

pARC8-DEST

Pink region removed when introducing genesby Gateway Cloning

LR Gatewayreaction(e.g. + SR001)

attB2

attB1pBAD-araCregulatory region

araC

CarResistance p15a origin

pARC8 +sample SEERlibrary member

SR001

SR001Barcode

a

b


https://doi.org/10.1101/2020.01.14.906594

0 50 100 150

1

10

100

1000

10000

Rif (!g/ml)

# C

olon

ies

(pla

ted

at 1

0-1)

Rifampicin resistance

Water Blank

Oligo 398

Oligo 399

Oligo 403

Oligo 404

0.001 0.01 0.1 1

1

10

100

1000

10000

SDS (% v/v)

# C

olon

ies

(pla

ted

at 1

0-1)

SDS resistance

Water Blank

Oligo 307

0.0 0.2 0.4 0.6

1

10

100

1000

10000

Cip (!g/ml)

# C

olon

ies

(pla

ted

at 1

0-1)

Ciprofloxacin resistance

Water Blank

Oligo 402

0 20 40 60

1

10

100

1000

10000

Stm (!g/ml)#

Col

onie

s (p

late

d at

10-1

)

Streptomycin resistance

Water Blank

Oligo 400

0 10 20 30 40 50

1

10

100

1000

10000

Chm (!g/ml)

# C

olon

ies

(pla

ted

at 1

0-1)

Chloramphenicol resistance

Water Blank

Oligo 460

a b

c d

e


https://doi.org/10.1101/2020.01.14.906594

0 2 4 6 8 1010-6

10-5

10-4

10-3

10-2

10-1

100

Round of SEER

Ave

rage

freq

uenc

y

SR016 (PapRecT)

SR120SR011

SR095SR055

0 1 2 3 4 5 6 7 8 9 100.0

0.5

1.0

Round of SEER

Ave

rage

SS

AP

fam

ily fr

eque

ncy

ERFGp2.5Rad52RecARecTSakSak4Unclassified


https://doi.org/10.1101/2020.01.14.906594

Red!

Red! [

WT co

dons]0.00

0.05

0.10

0.15

0.20

0.25E

ditin

g E

ffic

ienc

y


https://doi.org/10.1101/2020.01.14.906594

b

0 2 4 6 8 100.000001

0.00001

0.0001

0.001

0.01

0.1

1

Round of SEER

Ave

rage

freq

uenc

yov

er th

ree

sele

ctiv

e re

plic

ates

Library S2 SSAP Frequency

PF020 (CspRecT)

PF077

PF047

PF076

0 2 4 6 8 100.001

0.01

0.1

1

10

100

1000

Round of SEER

Fold

enr

ichm

ent o

ver

non-

sele

ctiv

e co

ntro

l

Library S2 SSAP Enrichment

PF020 (CspRecT)

PF077PF047PF076

a


https://doi.org/10.1101/2020.01.14.906594

PapRec

T

PapRec

T_WT

ssr

Rec2

PapRec

T_WT (R

BSopt)

ssr (

RBSopt)

Rec2 (

RBSopt)0.00

0.05

0.10

0.15

0.20

P. a

erug

inos

a E

ditin

g E

ffic

ienc

y

**

ns**

ns**

*

Original RBS Optimized RBS


https://doi.org/10.1101/2020.01.14.906594

0.001

0.01

0.1

1

P. a

erug

inos

a Ed

iting

Effi

cien

cy

Optimized PapRecT RBS

PapRecT WTcodons

PapRecT ECcodons

EcMutL_E32K

PaMutL_E36K

PaMutL

Optimized MutL RBS

- + + + + + +

- - - - + + -

- - + + + + +

+ + - - - - -

+ + + - - - -

- - - + + - -

- - - - - + -

RBS optimization

PapRecT codons

MutL variant


https://doi.org/10.1101/2020.01.14.906594

0.00

0.05

0.10

0.15

0.20

Cip0.5

Cip1.5

Cip4.5

P. aeruginosa Editing Efficiency

Oligo Pool: PAgyrA_T83I (CIP)PAparC_S87L (CIP)PAnfxB_KO (CIP)


https://doi.org/10.1101/2020.01.14.906594


https://doi.org/10.1101/2020.01.14.906594

1.2 1.4 1.6 1.8 2.0 2.2 2.41.0

1.1

1.2

1.3

1.4

1.5

1.6

Growth Rate (!)

Max

OD

600

SR011

SR016 (PapRecT)

SR055

Red!

SR095

SR120


https://doi.org/10.1101/2020.01.14.906594


https://doi.org/10.1101/2020.01.14.906594

Documents

Title: Improved bacterial recombineering by parallelized ... · 23 Efficient Recombineering (SEER). By performing SEER in E. coli to screen hundreds of putative 24 SSAPs, we identify