Strains and culture conditions

Electrocompetent competent E. coli DH10β was used for cloning and purchased from Biomed Co., Ltd. (Beijing, China). The antibiotics of ampicillin were used at 100 mg/mL. Unless otherwise noted, the cells were cultured in Luria-Bertani (LB) broth with shaking at 220 r.p.m. or on LB agar plates at 37 °C.

Library construction

For the assembly of 509 sequences, the oligo pool was synthesized and the lyophilized pool consisted of 11,776 oligos of 192 nts (synthesized by Twist Bioscience), which included the 152 nts payload in each oligo. The pool was resuspended in 1× TE buffer for a final concentration of 2 ng/μL. One of the files, 509 oligos, was flanked by binding sites for the premixed primers F02/R02. PCR was performed using Q5® High-Fidelity DNA Polymerases (NEB #M0491) and primers F01-F04/R01-F04 (10 ng oligos, 2.5 μL of each primer mix (100 mM), 0.5 μL Q5 High-Fidelity DNA Polymerase, 4 μL 2.5 mM dNTPs in a 50 μL reaction). Thermocycling conditions were as follows: 5 min at 98 °C; 10 cycles of: 10 s at 98 °C, 30 s at 56 °C, 30 s at 72 °C, followed by a 5 min extension at 72 °C. The library was then purified using Plus DNA Clean/Extraction Kit (GMbiolab Co, Ltd. #DP034P) and eluted in 40 μL ddH2O. This library was considered the master pool and run on a 2% agarose gel to verify the correct size. For the assembly of 11520 sequences, the synthetic DNA pool consisted of 11520 oligos of 200 nts (synthesized by Twist Bioscience), which included the 155 nts payload flanked by binding sites for the primers F1/R1 (Supplementary Fig. 3). The lyophilized pool was rehydrated in 1× TE buffer and the above protocol was used to amplify the file.

Computational strategy for primer homologous arm design

The primers were included three parts, one was the homologous arm (Homo-arm) which used for gibson assembly, the other was Not I which endonuclease recognition site and the last was address sequence. The primer design algorithm was implemented to codify rules for assembly fragments production using PCR. Primer homologous arm was designed using the NUPACK (http://www.nupack.org). Sets of primers are identified satisfying the following rules for homologous arm design: (1) homologous arm length 25 nt; (2) guanine-cytosine (GC) content between 40 and 60%; (3) no interaction between each homologous arm. Selected primers passing all of the aforementioned tests are provided as output as correct pairs that would yield the product of the defined size (Supplementary Figs. 1 and 2).

Assembly experiment

For the backbone preparation, the pUC19 plasmid was used as templates for PCR. The PCR was performed using PrimeSTAR® Max DNA Polymerase (Takara #R045Q) and primer set PCR-vactor-F/R. Thermocycling were amplified for 30 cycles, involving denaturation for 15 s at 98 °C, annealing for 5 s at 55 °C and primer extension for 20 s at 72 °C. Then the PCR products were purified by gel cut using Plus DNA Clean/Extraction Kit (GMbiolab Co, Ltd. #DP034P) and eluted in 30 μL ddH2O.

For the 509 oligos pool assembly fragment preparation, we started with the master pool as described above. The fragments were prepared with different homologous arms using Q5® High-Fidelity DNA Polymerases and the corresponding primers (the sequence was showed in Supplementary Table 1). The template used 0.2 ng of oligo input from the master pool into 50 μL PCR reactions. Initial denaturation was carried out at 98 °C for 5 min. Follow this by 20 PCR cycles, involving denaturation for 30 s at 98 °C, annealing for 30 s at 56 °C and primer extension for 20 s at 72 °C. Finally, the PCR reaction was terminated by incubating the solution at 72 °C for 5 min. The library was then purified using Plus DNA Clean/Extraction Kit (Gmbiolab, #DP034P) and eluted in 30 μL ddH2O. The final library was run under the same conditions as described. Then the Gibson Assembly® Master Mix—Assembly (NEB, #E2611) was used according to user’s manual.

For the 11520 oligos pool assembly fragment preparation, we started with the master pool as described above. The fragments were prepared with different homologous arms using 2× EasyTaq® PCR SuperMix (AS111, TRANS) and the corresponding primers (the sequence was showed in Supplementary Table 1 and Supplementary Fig. 4). The template used 20 ng of oligo input from the master pool into 50 μL PCR reactions. Initial denaturation was carried out at 94 °C for 2 min. Follow this by 10 PCR cycles, involving denaturation for 30 s at 94 °C, annealing for 30 s at 53 °C and primer extension for 20 s at 72 °C. Finally, the PCR reaction was terminated by incubating the solution at 72 °C for 5 min. The library was then purified and eluted in 100 μL ddH2O. NEBuilder® HiFi DNA Assembly Cloning Kit (NEB, #E5520) was used according to user’s manual.

The concentration of each fragment was calculated for optimal assembly, based on fragment length and weight, we used the following equation:

$${\mathrm{pmols}} = \left( {{\mathrm{weight}}\,{\mathrm{in}}\,{\mathrm{ng}}} \right) \times 1000/\left( {{\mathrm{base}}\,{\mathrm{pairs}} \times 650\,{\mathrm{daltons}}} \right).$$

(1)

Based on this calculation, the number of molecules per oligo and copy number was determined according to the following formula:

$${\mathrm{moles}}\,{\mathrm{dsDNA}}\,\left( {{\mathrm{mol}}} \right) = \,{\mathrm{mass}}\,{\mathrm{of}}\,{\mathrm{dsDNA}}\,\left( {\mathrm{g}} \right)/\left( {\left( {{\mathrm{length}}\,{\mathrm{of}}\,{\mathrm{dsDNA}}\,\left( {{\mathrm{bp}}} \right) \times 607.4} \right)} \right. \\ + 157.9\,{\mathrm{g}}/{\mathrm{mol}};$$

(2)

$${\mathrm{DNA}}\,{\mathrm{copy}}\,{\mathrm{number}} = {\mathrm{moles}}\,{\mathrm{of}}\,{\mathrm{dsDNA}} \times 6.022{\mathrm{e}}23\,{\mathrm{molecules}}/{\mathrm{mol}}.$$

(3)

We found an excellent agreement concentration between the assembly fragment and the backbone of Gibson assembly experiment. Thus, for 509 oligos pool assembly experiment, optimized cloning was performed using 1011 copy number of vectors and 108 copy number of inserts. For 11,520 oligos pool assembly experiment, optimized cloning was performed using 1010 copy number of vectors and 1010 copy number of inserts. The sample was incubated in a thermocycler at 50 °C for 60 min, respectively. Following incubation, store samples on ice or at −20 °C for subsequent transformation.

DNA storage in living cells

To prepare the fragments for the 509 oligos pool assembly, we started with the master pool as described above. The fragments were prepared with different homologous arms using Q5® High-Fidelity DNA Polymerase and the corresponding primers. Then, the Gibson Assembly® Master Mix (NEB, #E2611) was used according to the manufacturer’s instructions. To prepare the fragments for the 11520 oligos pool assembly, we started with the master pool as described above. The fragments were prepared with different homologous arms using 2× EasyTaq® PCR SuperMix (AS111, TRANS) and the corresponding primers. NEBuilder® HiFi DNA Assembly Cloning Kit (NEB, #E5520) was used according to the manufacturer’s instructions.

Transformation and culture

Electroporation was carried out in 1-mm gap cuvettes with conditions 1.8 kV, 200 Ω, 25 Μf, cells were recovered into fresh SOB medium for 1 h at 37 °C. For 509 assembly experiment, each sample (5 μL) was added to DH10β electrocompetent cells (50 μL) for electroporation reaction.

After recovery, 500 μL cells were plated on the selection medium plates with appropriate selective condition (Amp) and the colonies were counted by ImageJ. For 11520 assembly experiment, each sample (2 μL, totally 20 μL) was added to DH10β electrocompetent cells (50 μL) for electroporation reaction. After recovery, 500 μL cells were plated on the selection medium plates with appropriate selective condition (Amp) and the colonies were counted by ImageJ. The transformation rate was calculated by the following equation:

$${\mathrm{Transformation}}\,{\mathrm{efficiency}}\left( {{\mathrm{cfu}}/\!\mu {\mathrm{g}}} \right) \!=\! {\mathrm{colonies}}\,{\mathrm{on}}\,{\mathrm{plate}}/{\mathrm{plasmid}}\,{\mathrm{DNA}}\,{\mathrm{spread}}\,{\mathrm{on}}\,{\mathrm{plate}}$$

(4)

$${\mathrm{Transformation}}\,{\mathrm{rate}} = {\mathrm{Transformation}}\,{\mathrm{efficiency}}/10^{10}\left( {{\mathrm{cfu}}/\mu {\mathrm{g}}} \right) \ast$$

(5)

*note: 1010 cfu/μg is the theoretical transformation efficiency of DH10β electrotransducer cells.

For 509 oligos pool assembly experiment, another 500 μL cells were inoculated in 5 mL Luria broth (LB) medium plus appropriate antibiotics and grown overnight [37 °C, 220 revolutions per minute (RPM)] to obtain seed cultures. For 11520 oligos pool assembly experiment, 5 mL recovered cells were inoculated in 45 mL Luria broth (LB) medium plus appropriate antibiotics and grown overnight to obtain seed cultures. The seed cultures were then serially diluted (1:10) in 50 mL of prewarmed LB plus appropriate antibiotics followed by OD600 reached 1.2. This consecutive procedure was repeated 5 times (Supplementary Fig. 12).

Data recovery

After liquid and plate culture, the plasmid library was extracted using a plasmid miniprep Kit (TIANGEN, #DP103). Then, QuickCut™ Not I (Takara, #1623) was used for fragments recovery. After gel recovery of the correct fragments using the Plus DNA Clean/Extraction Kit, the samples of the 509 oligos pool (1F, 3F and 5F) and the 11520 oligos pool (passage-1 and passage-5 of 1F and 3F) were sequenced directly. To obtain more complete information, we performed a PCR of the constructed plasmid to amplify the 11,520 oligos pool (passage-1 and passage-5 of 1F and 3F) using Q5® High-Fidelity DNA Polymerases and the primer pair F02/R02. The thermocycling protocol was: (1) 98 °C for 5 min, (2) 98 °C for 30 s, (3)54 °C for 30 s, (4) 72 °C for 10 s, then repeat steps 2–4 five times, followed by a final elongation at 72 °C for 5 min. The products were purified using the Plus DNA Clean/Extraction Kit (GMbiolab Co, Ltd. #DP034P) and sequenced.

Statistics and reproducibility

Experiments with various sizes of DNA oligo pools were performed with at least 3 separated samples. Histograms were generated using Origin software. Quantitative data in figures are presented as the mean and standard deviation from three biological replicates.

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Source