Media, chemicals and culture conditions
Routine bacterial cultures were grown in LB Miller medium (BD Biosciences). E. coli was grown at 37 °C and A. tumefaciens was grown at 28 °C shaking at 200 rpm unless otherwise noted. Cultures were supplemented with kanamycin (50 mg L−1; Sigma-Aldrich), gentamicin (30 mg L−1; Thermo Fisher Scientific), spectinomycin (100 mg L−1; Sigma-Aldrich) or rifampicin (100 mg L−1; Teknova) when indicated. All other compounds unless otherwise specified were purchased through Sigma-Aldrich.
Strains and plasmids
All bacterial strains and plasmids used in this work are listed in Supplementary Table 2. All strains and plasmids created in this work are viewable through the public instance of the Joint BioEnergy Institute (JBEI) registry (https://public-registry.jbei.org/folders/847). All strains and plasmids created in this work can be requested from the strain archivist at JBEI with a signed material transfer agreement. All plasmids generated in this paper were designed using Device Editor, which is part of the DIVA suite version 6.1.2 and Open Vector Editor software version 18.3.6, while all primers used for the construction of plasmids were designed using j5 software version 3.4.0 (refs. 37,38,39). Plasmids were assembled by Gibson assembly using standard protocols40, Golden Gate assembly using standard protocols41 or restriction digest followed by ligation with T4 ligase42. Plasmids were routinely isolated using the QIAprep spin miniprep kit (Qiagen) and all primers were purchased from Integrated DNA Technologies (IDT). Plasmid sequences were verified using whole-plasmid sequencing (Primordium Labs). Agrobacterium was routinely transformed by electroporation as described previously using a 1-mm cuvette and a 2.4-kV, 25-μF, 200-Ω pulse43.
Mutating RepA proteins using epPCR
For each origin used in this study, the RepA or RepA-like protein was randomly mutagenized using epPCR44. Briefly, the RepA ORFs were amplified with a high-fidelity Phusion polymerase (New England Biolabs (NEB), M0530S) using primers to add BsaI restriction sites to the 5′ and 3′ ends of the product. Bands of the proper size were gel-purified (Zymo Research, D4007) and the resulting product was used as a template. An epPCR master mix was made, containing MnCl2 and unequal nucleotide ratios (10 mM TrisCl pH 8.3, 50 mM KCl, 7 mM MgCl2, 1 mM deoxycytidine triphosphate, 1 mM deoxythymidine triphosphate, 0.2 mM deoxyadenosine triphosphate, 0.2 mM deoxyguanosine triphosphate, 0.5 mM MnCl2, 1 μM forward and reverse primers, 20 pg μl−1 purified RepA template and 1 μl of Taq polymerase (Thermo Fisher Scientific, EP0401) per 50-μl reaction). For each reaction, a 12-cycle amplification was run to constrain the mutation number in each product strand to around 1–3 mutations. The appropriate bands were excised and gel-extracted and the resulting product was digested with BsaI and cleaned by column purification (NEB, T1030S)
Constructing mutant libraries
To make selection vectors, a gentamicin resistance gene (gentamicin-3-acetyltransferase) driven by a salicylic-acid-inducible promoter was originally cloned into pGingerBS-NahR45 and then variants were created for each of the pVS1, pSa and RK2 origins using a Gibson-like assembly with NEBuilder HiFi DNA assembly (NEB, E2621L). The entire vector except the RepA protein was then amplified in a Phusion PCR reaction with primers containing PaqCI restriction overhangs designed to have complementary sticky ends with the epPCR products from above. These PCR products were gel-purified and digested with PaqCI before being column-purified and ligated with the digestion product from the epPCRs using T7 DNA ligase (NEB, M0318S).
After 30 min of room temperature incubation, the ligation reaction was column-purified and 1 μl was electroporated into E. coli per reaction (NEB, C3020K). Following a 1-h recovery in the supplied medium, the electroporated samples were placed into 50 ml of LB + spectinomycin. A 1:100 dilution was made for each flask and was plated onto solidified LB + spectinomycin to allow for an estimate of library size. Around five flasks were prepared for each origin to create a library size of at least 100,000 independent transformants. After an overnight growth at 37 °C shaking at 200 rpm, 5-ml aliquots from each flask were miniprepped (Qiagen, 27104) and all samples from each origin were combined into single tubes, comprising the mutant vector library of ~100,000 mutants.
Next, 1 μl of these libraries per reaction were then electroporated into A. tumefaciens C58C1 electrocompetent cells. Using the same method as above with 50-ml recovery flasks, cells were recovered at 30 °C overnight shaking at 200 rpm and enough flasks were prepared to obtain a library size of around 150,000 mutants. Then, 5 ml of cells from each flask were combined and made into glycerol stocks for use in the higher-copy-number selection screen.
Selecting higher-copy-number mutants from constructed libraries
A checkerboard assay was used to select for higher-copy-number variants. The mutant libraries described above for each ORI contain A. tumefaciens C58C1 cells that each harbor a selection vector comprised of a constitutively expressed spectinomycin resistance gene, a salicylic-acid-inducible gentamicin resistance gene and the ORI of interest with a repA gene that was subjected to epPCR to induce random mutations. A 1-ml glycerol stock of each of the mutant libraries was grown in 50 ml of LB + spectinomycin (50 mg L−1) overnight at 30 °C shaking at 200 rpm along with a culture of the WT strain for each ORI. Spectinomycin was added to maintain the plasmid within the population during this overnight growth before the actual selection began with gentamicin.
For each ORI, 100 ml of LB + spectinomycin was inoculated with saturated culture from the mutant library or WT control in a 1:200 ratio and 500 μl of this solution was added to each well of a 96-well block. Gentamicin was then added to all wells in the block following a serial dilution row-wise from rows A to H with a dilution factor of 2, starting from 3,000 mg L−1 and ending with 23.4 mg L−1. Salicylic acid was then added to all wells in the plate following a serial dilution column-wise from columns 1 to 12 with a dilution factor of 2, starting from 5 μM in column 12 and ending with 2.44 nM in column 1. This created 96 unique selection conditions within the block corresponding to different gentamicin and salicylic acid concentrations, enabling selection of the population over a large dynamic range of selection conditions.
The resulting 96-well blocks were then covered in vent film and incubated overnight at 30 °C shaking at 200 rpm. Growth was determined by calculating the OD600 of each well the following day using a Genesys 50 (Thermo Fisher Scientific) and the mutant population was compared to its WT counterpart to determine gentamicin–salicylic acid combinations that were lethal to the WT ORI but that permitted growth of the mutant population. Two selection conditions that were WT lethal per ORI were chosen and grown in triplicate in 10-ml cultures of LB + spectinomycin + gentamicin + salicylic acid inoculated with a fresh glycerol stock of the mutant population. A culture containing just WT plasmid was also prepared for each selection condition as a negative control. Additionally, a triplicate of cultures containing the mutant library without gentamicin + salicylic acid selection were grown as an unselected control.
The following selection conditions were used: pVS1, 750 mg L−1 gentamicin + 5 μM salicylic acid and 375 mg L−1 gentamicin + 156 nM salicylic acid; pSa, 1,500 mg L−1 gentamicin + 5 μM salicylic acid and 750 mg L−1 gentamicin + 625 nM salicylic acid; BBR1, 2,250 mg L−1 gentamicin + 5 μM salicylic acid and 1,000 mg L−1 gentamicin + 78 nM salicylic acid; RK2, 1,500 mg L−1 gentamicin + 5 μM salicylic acid. The overnight cultures were grown at 30 °C shaking at 200 rpm and plasmid-prepped according to a modified Qiagen protocol (https://www.qiagen.com/us/resources/resourcedetail?id=95083ccb-9583-489e-b215-99bd91c0604e&lang=en).
These samples were then used for the following Tagmentation procedure to barcode the RepA ORFs to allow for identification of enriched mutations within the selected population relative to the unselected control.
Sequencing mutant library survivors to identify enriched RepA mutations
The extracted plasmids from the selected populations were sequenced using an Illumina MiSeq to identify enriched mutations that may have contributed to their survival in WT-lethal conditions. The entire RepA ORF + 150 bp upstream and downstream was PCR-amplified and gel-purified. Purified samples were then fragmented into random 600-bp pieces and barcoded using the Illumina bead-linked transposon Tagmentation kit (Illumina, 20060059) according to the protocol provided by the manufacturer. The concentration of the barcoded fragments was then measured using a Qubit fluorometer with the Qubit dsDNA HS assay kit (Invitrogen). Library size analysis was conducted using a Bioanalyzer (Agilent Technologies). To enhance the quality of Illumina reads for regions with low variations in the conserved PCR amplicons, 20% PhiX Control v3 DNA (Illumina, FC-110-3001) was added to the library. The libraries were evenly pooled and the library, along with the PhiX mixture at a concentration of 18 pM, was loaded onto the MiSeq platform using the MiSeq Reagent Kit v3 (paired-end, 2 × 300-bp reads; Illumina).
Mapping of Illumina reads was performed as described previously46. Briefly, quality control for the reads from this run was performed using FastQC version 0.11.8 and MultiQC version 1.10.1. Read trimming was performed using trimmomatic version 0.39 (ref. 47). Trimmed reads were aligned to their corresponding reference file using Bowtie version 2.4.5 (ref. 48). Mutations in sequencing reads were identified using pysamstats version 1.1.2 and pysam version 0.18.0. All code from this work is publicly available on GitHub (https://github.com/shih-lab/origin_story) and all sequencing reads are available on the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) under BioProject accession PRJNA1031697.
Building plant GFP expression vectors with enriched RepA mutations from the MiSeq data
To construct GFP expression vectors, promoters of variable strengths derived from the PCONS suite of constitutive plant promoters were used34. Expression vectors containing a GFP construct driven by the constitutive plant promoters pCM2 (for RK2 and pSa) or pCL2 (pVS1 and BBR1) along with a 35S::KanR component for stable-line selection were made using Gibson assembly. RK2 was run as a pilot for this project using the stronger pCM2 promoter to drive GFP. Because of the high GFP output of the WT forms of pVS1 and BBR1, a very large fold-change increase in GFP expression could potentially saturate the GFP channel of the plate reader; as such, the weaker pCL2 was used for these origins. As pSa had a weaker starting WT expression, pCM2 was used for this origin. To demonstrate the performance of the highest-performing mutants expressed from the same promoter for all origins, an additional set of vectors were cloned with pCM2::GFP for pVS1 and BBR1 (Supplementary Fig. 6). For the Ruby vector described in Supplementary Fig. 12, the strongly expressing pCH5 promoter was used to drive the Ruby reporter49.
Single SNPs were introduced into the RepA coding sequence using PCR with the SNP of interest designed into a primer overhang. Proper PCR band lengths were verified in a high-throughput manner using an Agilent ZAG (zero-gel electrophoresis). PCR products were column-purified and assembled using NEBuilder HiFi DNA assembly (NEB, E2621L). The resulting assemblies were transformed by heat shock into XL1-Blue E. coli and grown overnight at 37 °C on solidified LB + kanamycin plates. Single colonies were selected, grown in a 4 ml of LB + kanamycin liquid culture and miniprepped; then, the integrity of the plasmid sequence including the proper SNP was verified using whole-plasmid sequencing from Primordium Labs (https://www.primordiumlabs.com/). Verified plasmids were then transformed into the EHA105 strain of A. tumefaciens by electroporation and selected on plates of solidified LB + kanamycin (50 mg L−1) and rifampicin (100 mg L−1) at 28 °C shaking at 200 rpm.
Screening origin mutants in N.
benthamiana
A previously established method for transiently expressing genes in N. benthamiana was used to screen the impact of the mutant ORIs in planta50. Single colonies of transformed EHA105 were selected and used to inoculate 5 ml of LB + kanamycin and rifampicin for overnight growth at 28 °C shaking at 200 rpm. The following morning, 10 ml of fresh LB + kanamycin + rifampicin was added to each vial and the cultures were grown for 2 h. An aliquot from each culture was taken and used to measure the OD600. Then, 10 ml of each culture was centrifuged at 3,200g for 15 min. The supernatant was removed and cells were diluted to an OD600 of 1.0 using an appropriate volume of tobacco infiltration buffer (10 mM MgCl2, 10 mM MES and 200 μM acetosyringone (added fresh), pH 5.7). Resuspended cultures were allowed to induce for 2 h at room temperature while gently rocking.
N. benthamiana was grown at 25 °C under long-day conditions (16 h of light, 8 h of darkness) of 150 μmol m−2 s−1 photosynthetically active radiation (PAR; wavelength: 400–700 nm). Sunshine no. 4 growing mixture supplemented with Osmocote was used as the planting medium. For each origin mutant, six 4-week old N. benthamiana plants were syringe-infiltrated on the fourth and fifth leaves from the top of the plant. Infiltrated plants were watered and returned to the growth room for 3 days. After this 3-day period, four leaf discs of infiltrated tissue per leaf were taken using a standard 6-mm hole puncher and these discs were placed abaxial side up into a 96-well culture plate filled with 320 μl of water such that the discs were floating on the surface. These 96-well plates were then assayed for GFP fluorescence using a BioTek Synergy H1 microplate reader set for conditions of 483-nm excitation and 512-nm emission. GFP fluorescence readings for each disc were analyzed in R to compare mutant performances compared to the WT origins.
L.
sativa transient expression assays
Buttercrunch lettuce plants were grown in 18-cell flats at 25 °C under long-day conditions of 150 μmol m−2 s−1 PAR using the same Sunshine no. 4 + Osmocote planting medium. Then, 5-week-old plants were infiltrated with a blend of two EHA105 strains consisting of the same GFP binary vector strains as the tobacco experiments and a strain harboring a binary vector for the P19 gene. P19 is a viral gene that suppresses plant gene silencing and coinfiltration of this construct aids in transient expression, particularly in more recalcitrant backgrounds51. These two strains were grown overnight at 28 °C shaking at 200 rpm and were brought to an OD600 of 1.0 and mixed in a 1:1 ratio before inducing in the tobacco infiltration medium mentioned above for 2 h. The fifth leaf of each plant was infiltrated and a 4-day incubation period at 25 °C was used before isolating leaf discs for GFP quantification.
Copy number quantification
Plasmid copy number was determined using nanoplate dPCR. A QIAcuity EvaGreen PCR kit (Qiagen, 250111) was used for all dPCR reactions. Samples were transferred into an 8,500-partition 96-well QIAcuity nanoplate and loaded into a QIAcuity One system. dPCR reactions were conducted with a standard protocol (2 min at 95 °C followed by 40 cycles of 15 s at 95 °C, 15 s at 56 °C and 15 s at 72 °C). The nanoplate was imaged with an exposure duration of 150 ms and gain of 2. Images were analyzed with the QIAcuity Software Suite. The concentration of plasmids were calculated using Poisson statistical methods by the QIAcuity Software Suite. For each biological replicate, two dPCR reactions were run, one targeting the single-copy rpoB gene found within A. tumefaciens C58 and one targeting the kanamycin resistance (KanR) gene located on the binary vector. The final copy number was determined as the ratio of KanR to rpoB copies, which served as a measurement of the number of plasmid counts to genomic counts respectively. The following primers were ordered from IDT and used for the dPCR reactions: KanR forward, GATCATCCTGATCGACAAGACCGG; KanR reverse, CTGCCGAGAAAGTATCCATCATGGC; rpoB forward, GAGTACCGGAATCTCGTCAAAGCC; rpoB reverse, CGAAGATCTCTACGGCAACTACCTGG.
Growth rate quantification
The growth rates of all WT and mutant strains were evaluated in both a rich LB medium and a minimal salts medium (MOPS minimal + 10 mM glucose)52. For each, a glycerol stock was used to inoculate a 10-ml culture of LB that was grown overnight at 28 °C shaking at 200 rpm. This saturated culture was spun down and resuspended in either MOPS minimal salts + 10 mM glucose or LB medium at a concentration that was 1:200 of the original saturated stock. Next, 150 μl of this bacterial culture was added to each well of a 96-well clear plate in quadruplicate (Falcon, 353072) and the cultures were grown shaking linearly at 1,000 cpm at 28 °C in a BioTek Synergy H1 plate reader. The OD600 was measured every 6 min for 24 h; using the data from the plate reader, the growth rate for each culture was calculated using the GrowthCurver package in R. The growth rate in this study is reported as doublings per hour, which is 60 divided by the doubling time.
A.
thaliana stable transformation
A. thaliana was transformed using the floral dip method of transformation as previously described53. Arabidopsis seeds were sprinkled over the surface of wet Sunshine no. 4 growing mixture and allowed to grow for 12 days. After this period, 18-pot flats were prepared with wet Sunshine no. 4 growing mixture with Osmocote and five Arabidopsis seedlings were transplanted per pot in an X pattern. Flats were grown at 22 °C under short-day conditions (8 h of light, 16 h of darkness) of 150 μmol m−2 s−1 PAR. After 3 weeks, the flats were moved to a 22 °C chamber with long-day lighting conditions of the same intensity. The first floral meristem from each plant was excised to promote axillary shoot growth. After three more weeks, all plants began flowering with multiple inflorescences and these were used for the floral dip procedure.
The top-performing origin, pVS1, and lower-expressing origins RK2 and pSa were selected for an analysis of impact of enhanced binary vector copy number on stable transformation. The same plasmids used in the tobacco screen were used for this experiment as they contained a 35S::KanR component that allows for the selection of T1 plants with kanamycin. These plasmids were electroporated into the A. tumefaciens strain GV3101 and selected on solidified plates of LB + rifampicin, kanamycin and gentamicin grown for 3 days at 30 °C. Then, 300-ml cultures of each strain were grown overnight at 30 °C shaking at 200 rpm in LB + rifampicin, kanamycin and gentamicin. These cultures were centrifuged for 20 min at 3,200g and the supernatant was discarded and replaced with 300 ml of floral dip medium consisting of 5% sucrose with 0.02% Silwet (50 g of sucrose and 200 μl of Silwet brought to 1 L with water).
Pots containing five A. thaliana plants were gently inverted and the inflorescences were submerged into the Agrobacterium solution for 15 s with light agitation. All pots dipped in the same construct were then laid sideways in an empty planting tray and were covered with a plastic lid that had been misted with water. Plants were left at room temperature overnight in this humidity chamber before being returned to an upright position in the 22 °C long-day chamber. Then, 1 week after the initial floral dip, this process was repeated with the same plants to further expose newly grown buds to the bacteria. For each construct, four pots containing a total of 20 plants were dipped; following the second dip, a single stake was placed into one pot and the numerous inflorescences were gently tied together in a single mass. Next, 2 weeks after the second dip, a paper bag was used to cover the inflorescence heaps and the plants were moved to a drying rack to die and desiccate.
Following desiccation, seeds were sifted multiple times to separate them from any silique debris. Then, 200 mg of seeds per sample were measured and placed into tubes for the plating experiment. Seeds were then sterilized by first submerging in 70% ethanol for 1 min followed by shaking in a solution of 50% bleach from a 12.5% sodium hypochlorite stock with a drop of Triton-20 surfactant for 7.5 min before rinsing with sterile water three times. A moderately dense layer of seeds that were distributed in close but not overlapping proximity to each other was spread onto the surface of selection plates consisting of Murashige and Skoog salts + 1 g L−1 MES brought to a pH of 5.7 along with 50 mg L−1 kanamycin and 200 mg L−1 timentin to kill residual A. tumefaciens and 8 g L−1 agar for solidification. Each 200 mg of seeds approximately covered 6 5-inch plates. The plates were then wrapped in vent tape, placed into a long-day 22 °C chamber and allowed to grow for 3 weeks before quantification of successfully growing nonchlorotic plants.
R.
toruloides transformation
To construct a R. toruloides expression vector, the plasmid JPUB_013523 (ref. 54) was modified to contain the WT or mutant forms of the pVS1 or RK2 ORIs (R106H and S20F mutants, respectively). Origin variants from this base vector were created using Gibson assembly and were electroporated into the EHA105 strain of A. tumefaciens. Single colonies were used to inoculate LB + kanamycin and rifampicin for overnight growth at 28 °C shaking at 200 rpm. AMT of R. toruloides was conducted as previously described55.
Selection was conducted on plates containing nourseothricin; for each transformation, plates with a 1:10 dilution along with a full-concentration plating of remaining cells were made. Plates were incubated at 30 °C for 2 days and then imaged using a AnalytikJena UVP GelSolo followed by colony count quantification using Fiji (https://imagej.net/software/fiji/downloads).
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
