Efficient golden gate assembly of DNA constructs for single molecule force spectroscopy and imaging

Abstract Single-molecule techniques such as optical tweezers and fluorescence imaging are powerful tools for probing the biophysics of DNA and DNA-protein interactions. The application of these methods requires efficient approaches for creating designed DNA structures with labels for binding to a surface or microscopic beads. In this paper, we develop a simple and fast technique for making a diverse range of such DNA constructs by combining PCR amplicons and synthetic oligonucleotides using golden gate assembly rules. We demonstrate high yield fabrication of torsionally-constrained duplex DNA up to 10 kbp in length and a variety of DNA hairpin structures. We also show how tethering to a cross-linked antibody substrate significantly enhances measurement lifetime under high force. This rapid and adaptable fabrication method streamlines the assembly of DNA constructs for single molecule biophysics.

. Primer sequences for PCR parts used in 10.1 kbp end-labelled duplex design. The PCR parts of Figure 2A were amplified using λ-DNA as the template. The primer sequences are shown in Table S1. The BsaI recognition site is highlighted in bold. The four base overhang generated by BsaI digestion is shown in colour. The four base sequences in the designs were chosen based on recommended sets from the NEB golden gate assembly tool (https://goldengate.neb.com/). Complementary sequences share the same colour.

Part
Part stoichiometry Biotin handle (860 bp) 8 2.1 kbp section 1 1.2 2.1 kbp section 2 1 2.1 kbp section 3 1 2.1 kbp section 4 1.2 Digoxigenin handle (860 bp) 8 Table S2. Stoichiometry for golden gate reaction used in 10.1 kbp design. In the golden gate reaction shown in Figure 2 in the main text, the six parts were mixed in the above stoichiometric ratios

Part Forward primer Reverse primer
Biotin handle (860 bp) GATGGTCTCATGCCGCAACAGCACAACCCAAACTG AATCTGCTGCAATGCCACAG 1.5 kbp section 1 GATGGTCTCAGGCACCGTGGAATGAACAATGGAAGTC GATGGTCTCAACCTCTGTCAGGTGGCTCAATCTCTTC 1.5 kbp section 2 GATGGTCTCAAGAGGCCACTTCAGCACGAGATG GATGGTCTCAGCTCTCCAGATCTGCAGCATCCTGAA Digox. handle (860 bp) GATGGTCTCAGAGCGCAACAGCACAACCCAAACTG AATCTGCTGCAATGCCACAG Table S3. Primer sequences for PCR parts used in short hairpin design. The DNA hairpin structure of Figure 4 is composed of four PCR parts and one oligonucleotide part. The primers for the PCR parts (using λ-DNA as the template) are shown.

Part
Part stoichiometry Biotin handle (860 bp) 2.25 1.5 kbp section 1 1.5 Oligo part 1 1.5 kbp section 2 1.5 Digoxigenin handle (860 bp) 2.25 Table S5. Stoichiometry for golden gate reaction used in short hairpin design. In the golden gate reaction, the five parts were mixed in the corresponding stoichiometric ratios.

Forward primer Reverse primer
Biotin handle (860 bp) GATGGTCTCATGCCGCAACAGCACAACCCAAACTG AATCTGCTGCAATGCCACAG 1.5 kbp section GATGGTCTCAGCAACCGTGGAATGAACAATGGAAGTC GATGGTCTCATCTGCTGTCAGGTGGCTCAATCTCTTC 2.1 kbp section GATGGTCTCAAGGAGCCACTTCAGCACGAGATG GATGGTCTCAGCTCTCCAGATCTGCAGCATCCTGAA Digox. handle (860 bp) GATGGTCTCAGAGCGCAACAGCACAACCCAAACTG AATCTGCTGCAATGCCACAG Table S6. Primer sequences for PCR parts used in long hairpin design. The DNA hairpin structure of Figure 5 is composed of four PCR parts and two oligonucleotide parts. The primers for the PCR parts (using λ-DNA as the template) are shown.

Oligo3 -connector PhosTTGCGGCGGTTTCAGCTGCCATTTTTTTTTTTTTCGTCTGTGACTAACTG
Oligo4 -connector PhosTCCTCAGTTAGTCACAGACGA hairpin loop PhosCAGATTGCCAAGTGAGTCCGATTTTTCGGACTCACTTGGCAA Table S7. Primer sequences for PCR parts used in long hairpin design. "Phos" indicates a 5' phosphorylation. The three-way connector oligonucleotide is assembled by annealing the four oligonucelotides while the hairpin loop sequence is separately annealed by itself.  Table S9. Summary of statistics on DNA tethers. Three separate batches of the 10 kbp dsDNA design were used ( Figure 2). For this design a "good" tether is one which is torsionally constrained. For the hairpin designs, a "good" tether is one for which there is the expected signal due to unfolding of the hairpin. Figure S1. Time dependence of golden gate DNA assembly. The DNA was mixed as in Table S2 (except a stoichiometric ratio of 2.4 was used for the handles). (A) The time indicates the length of assembly incubation. The reactions were quenched at each timepoint with 1x SDS loading dye (NEB).

FIGURES:
(B) Quantitation of band intensity (as % of total signal from all bands at a given timepoint). The % of the final product (10 kbp) at t= 24 hrs is 6%.

Figure S3. Test of non-specific attachment on flow-cell surface.
A flow-cell surface was prepared as described in the Materials and Methods using full length anti-digoxigenin and a glutaraldehyde fixing step. The top image shows three fiducial beads in the field of view which are used for drift correction. For the test, 10 µL Dynabead M280 stock was mixed with 10 µL 2 mg/ml BSA, 2 mg/ml βcasein in 20 mM Tris-HCl (pH 7.5), 150 mM NaCl and 5 µL of this mixture was added into the flow-cell (middle image). This mimics typical conditions used with DNA coated beads. After incubating for five minutes, ~20 channel volumes of 1 mg/ml BSA, 1 mg/ml β-casein in 20 mM Tris-HCl (pH 7.5), 150 mM NaCl was passed through the channel. The bottom image shows the same field view after washingthere was one non-specifically attached magnetic bead which is marked in red. From five such fields of view we measured a total of three non-specifically attached beads. The scale bar is 20 µm.
(C) Oligo annealing: Oligos were purchased from Integrated DNA Technologies and resuspended in IDT DNA duplex buffer (30 mM HEPES, pH 7.5; 100 mM potassium acetate). A particular oligo part was assembled by mixing all oligos together at equal stoichiometry with IDT DNA duplex buffer such that the final concentration of each oligo was 10 μM and then heating to 70 °C for one minute before linearly cooling to room temperature over 30 min.
(D) Golden gate assembly: The purified PCR amplicons and annealed oligo parts were mixed together at the ratios given above (Tables S2, S5, S8) together with BsaI-HFv2 (NEB -R3733S), T4 DNA ligase (NEB -M0202S) in 1x T4 DNA ligase buffer (NEB -B0202S). The assembly was carried out in a single tube of 20-40 µL volume with final concentrations of 1 U/µL BsaI-HFv2 and 20 U/µL T4 DNA ligase. The total DNA concentration in the golden gate reaction was 30-100 ng/µL and was kept as high as possible to maximise the ligation efficiency. The mixture was incubated for 3 hrs cycling between 37 °C and 16 °C every 5 min.