UniClo technology exploits methylation for universal scarless DNA assembly

TL;DR

UniClo is a new DNA assembly method that allows scarless and flexible DNA construction by using methylation and CRISPR-dCas9.

Contribution

UniClo introduces a methylation-based system with CRISPR-dCas9 protection to enable scarless DNA assembly without sequence domestication.

Findings

UniClo allows assembly of DNA fragments with internal type IIS sites without requiring sequence domestication.

The method enables fully scarless hierarchical DNA assembly using only three vectors and two universal overhangs.

UniClo provides a detailed protocol for methylation-based DNA assembly and scarless vector trimming.

Abstract

Several Golden Gate-based DNA assembly techniques have been developed previously with different limitations including the requirement for domestication of sequences with internal sites for the type IIS restriction enzymes used, insertion of persistent scars in assembled DNA and the need for multiple assembly vectors and overhang sequences. We developed UniClo, which overcomes all these problems. Sequences with internal type IIS sites can be assembled, it allows fully scarless hierarchical assembly and requires only three assembly vectors and two universal overhang sequences. This is achieved by three key elements: (i) Recombinant methylases are used in vitro to methylate, and thus inactivate, any sites in fragments to be assembled for the type IIS restriction enzyme used in the assembly as well as the outer sites of the assembly vectors. (ii) A CRISPR-dCas9 molecule is used to protect the type IIS restriction enzyme sites required for the assembly from methylation, thus preserving the activity of these sites. (iii) A set of engineered vectors is used to trim overhangs that would otherwise generate scars. Here, we present a detailed protocol for performing DNA assembly using UniClo and describe the methylation-protection of the fragments to be assembled, methylation of the scarless vectors, the assembly reaction, and analysis of the final assembled DNA molecule. UniClo offers substantial flexibility in the assembly design and enables the assembly of any DNA molecule regardless of its sequence, nature and application.

Figures

Overview of DNA assembly using UniClo. In the first round of assembly, a start fragment [fragment F1] is in a VL donor plasmid, a middle fragment [fragment F2] is in a VM donor plasmid, and an end fragment [fragment F3] is in a VR donor plasmid. The start and end fragments of the sequence to be assembled include ‘u’ and ‘v’ as flanking overhangs, respectively. These overhangs are compatible with those of the scarless assembly vectors (VL, VM, and VR) and are trimmed during the assembly process. If the fragments [F1 + F2 + F2] are assembled in a VL assembly vector the right overhang (‘v’) is trimmed, if they are assembled in a VR assembly vector the left overhang (‘u’) is trimmed and if they are assembled in a VM assembly vector both overhangs are trimmed (‘u’ and ‘v’). This overhang trimming process happens during the subsequent rounds of assembly until a VM assembly vector is used in the final round to obtain a fully scarless assembled DNA. MP-dCas9/sgRNA: methylation-protection dCas9/sgRNA molecule.

Schematic representation of the steps involved in UniClo DNA assembly. MP-dCas9/sgRNA: methylation-protection dCas9/sgRNA molecule.

DNA fragment design in the VL, VM, and VR donor plasmids for the first round of UniClo DNA assembly. (a) Start fragments are inserted into VL plasmids including the CCTC adapter at their left end (overhang ‘u’=CCTC). (b) Middle fragments are inserted into VM plasmids without any additional adaptors and their overhangs are derived from the sequence to be assembled (‘b’ and ‘c’). (c) End fragments are inserted into VR plasmids including the TGAGAG adaptor at their right end (overhang ‘v’=AGAC).

UniClo three-vector set consisting of the (a) VL, (b) VM, and (c) VR scarless assembly vectors. POC1518, POC1519, and POC1520 are kanamycin-resistant forms, while POC1553, POC1525, and POC1554 are the chloramphenicol-resistant forms—the respective VL, VM, and VR plasmids differ only in the selection marker. The overhangs of each of the three assembly vectors when cut during the assembly will always be ‘u’ = CTCC and ‘v’ = AGAC. These overhangs are generated by cutting directed from the inner bsa I sites of the assembly vectors while the outer bsa I sites are methylated during the assembly and so are not cut. The outer bsa I sites on the right of the VL and VM assembly vectors are generated during the assembly [14].

Overhang compatibility between the fragments to be assembled in VL, VM, and VR donor plasmids and the assembly vectors to achieve a fully scarless DNA assembly. (a) A simple assembly scheme for two fragments. (b) A simple assembly scheme for three fragments. Light blue letters (‘a’, ‘b’, ‘c’, and ‘d’) represent different four base-pair sequences derived from the final sequence to be assembled. The green letters ‘u’ and ‘v’ represent the adaptor overhang sequences required for the start and end fragments for compatibility with any of the assembly vectors. Depending on the type of assembly vector (VL, VM, or VR) used for the assembly, the adaptor overhangs ‘u’ and ‘v’ will be trimmed during the rounds of assembly.

Tables

Table 1. DNA methylases, assembly vectors, and donor plasmids used in UniClo.

Plasmid code	Addgene code	Description	Selection marker
Recombinant DNA methylases
POC1471	213759	Switch methylase M.Osp807II	Ampicillin
POC1464a	213736	Non-switchable methylase M2.Eco31I	Kanamycin
POC1466a	213754	Non-switchable methylase M2.Eco31I_2
POC1468a	213756	Non-switchable methylase M2.BsaI
Scarless assembly vectors (three-vector set)b
POC1518	221579	VL assembly vector for assembly of start fragments	Kanamycin
POC1519	221580	VM assembly vector for assembly of middle fragments
POC1520	221581	VR assembly vector for assembly of end fragments
POC1553	238470	VL assembly vector for assembly of start fragments	Chloramphenicol
POC1525	226496	VM assembly vector for assembly of middle fragments
POC1554	238471	VR assembly vector for assembly of end fragments
Donor plasmids for the first round of assembly
POC1535	226476	VL donor plasmids for start fragmentsc	Chloramphenicol
POC1539	226480
POC1542	226483
POC1544	226485
POC1536	226477	VM donor plasmids for middle fragmentsd
POC1537	226478
POC1540	226481
POC1538	226479	VR donor plasmids for end fragmentse
POC1541	226482
POC1543	226484
POC1545	226486

Table 2. Examples of alternative UniClo assembly strategies for the same sequence.

Strategy 1
Construct to be assembled		First round		Second round
Fragment (overhangs)	Fragment in donor plasmid (overhangs)	Assembly vector (overhangs) (u-v)	Assembled plasmid (overhangs)	Assembly vector (overhangs) (u-v)	Assembled plasmid (overhangs)
F1 (a-b)	VL(u-b)	VL	VL (u-e)	VM	(a-l)
F2 (b-c)	VM (b-c)
F3 (c-d)	VM (c-d)
F4 (d-e)	VR (d-v)
F5 (e-f)	VL (u-f)	VM	VM (e-h)
F6 (f-g)	VM (f-g)
F7 (g-h)	VR (g-v)
F8 (h-i)	VL (u-i)	VM	VM (h-j)
F9 (i-j)	VR (i-v)
F10 (j-k)	VL (u-k)	VR	VR (j-v)
F11 (k-l)	VR (k-v)

Full text

Introduction

DNA assembly is central to much of synthetic biology because it is not currently possible to synthesize long lengths of DNA directly. For this reason, there has been considerable research effort to develop methods for DNA assembly that are universal, simple to execute, and allow flexibility in the design of the assembly. There is substantial need because DNA assembly has wide application in different fields for diverse purposes. A number of assembly methods have been developed previously, each with different features and its own strengths and drawbacks. Key drawbacks include the need for domestication (site-directed mutation) of fragments containing internal sites for the type IIS restriction enzymes used during the assembly, the insertion of persistent scars in the assembled fragment, the need for multiple vectors and overhang sequences and in some cases the need for in vivo steps or reactions in cellular hosts [1–5]. We sought to develop an assembly technology that fulfils the desired characteristics above and overcomes the key drawbacks of other techniques. A key element in our approach is the use of methylation; although methylation by DNA methylases has been used in some previous DNA assembly techniques [6–10], methylases have not been deeply exploited to develop simple and efficient DNA assembly techniques [11].

First, we developed MetClo [12], which builds on the Golden Gate approach for DNA assembly, but unlike Golden Gate which requires two type IIS restriction enzymes for hierarchical assemblies, MetClo requires only one type IIS restriction enzyme which can cut both the outer and inner sites in the assembly vector. In MetClo, the outer sites were engineered to be methylated in vivo in Escherichia coli, while the inner sites remained unmethylated. Thus, only the inner sites were cut during the assembly and the outer sites, when not methylated, could be used subsequently in a further round of assembly. We developed a panel of purified recombinant methylases with specificity for sites overlapping in part or fully with sites recognized by specific type IIS restriction enzymes and tested their in vitro activity for the targeted methylation of these type IIS restriction sites [13]. We use the term switch methylase to refer to a methylase which has a recognition site that overlaps only partially with the recognition site for a corresponding restriction enzyme. We use the term non-switchable methylase to refer to a methylase which has a recognition site that fully overlaps with the recognition site for a corresponding restriction enzyme. Switch methylases allow the methylation of type IIS recognition sites that have been engineered to partially overlap with the methylase recognition site, and this methylation blocks the type IIS enzyme activity at the site, but other non-engineered sites are not methylated and so are cut by the type IIS enzymes. Non-switchable methylases methylate all the type IIS restriction enzyme sites. UniClo makes use of both recombinant switch and non-switchable methylase activities [14].

In UniClo, switch methylases are used for the in vitro methylation of the assembly vectors and non-switchable methylases are used for the in vitro methylation of the fragments to be assembled which are in donor plasmids. The assembly vectors are methylated in their outer type IIS restriction enzyme sites, and during the assembly they are cut from their inner unmethylated sites. Prior to the assembly, fragments to be assembled are methylated at any internal sites for the type IIS restriction enzyme to be used in the assembly. However, the outer type IIS restriction sites that flank the fragments must be protected from methylation so that they can be cut during the assembly. To protect these sites from methylation, we developed a CRISPR/dCas9 molecule. The methylation-protection dCas9/sgRNA molecule is guided by a universal RNA that is specifically designed for this purpose and can be used in the assembly of any sequence. Steric blockade by this molecule prevents the non-switchable methylase from methylating and so inactivating these outer sites that flank the fragments to be assembled. Once the assembly occurs using the methylated assembly vectors and the methylation-protected fragments in donor plasmids, the assembled plasmids become new donor plasmids whose flanking sites are the original outer sites of the assembly vector. These new donor plasmids are used in subsequent rounds of assembly. We engineered a unique set of three assembly vectors with different antibiotic selection markers to be used for fully scarless assembly in each round of a hierarchical assembly. Only two universal overhangs are required, one at each end of the assembled fragment and these are compatible with the overhangs of each of the three scarless assembly vectors. These overhangs are trimmed from the assembled fragments during the assembly process to produce a final scarless assembled DNA construct (Fig. 1). By using UniClo, we have achieved the successful assembly of long DNA pieces containing multiple internal type IIS sites for the enzyme used during the assembly [14].

Overview of DNA assembly using UniClo. In the first round of assembly, a start fragment [fragment F1] is in a VL donor plasmid, a middle fragment [fragment F2] is in a VM donor plasmid, and an end fragment [fragment F3] is in a VR donor plasmid. The start and end fragments of the sequence to be assembled include ‘u’ and ‘v’ as flanking overhangs, respectively. These overhangs are compatible with those of the scarless assembly vectors (VL, VM, and VR) and are trimmed during the assembly process. If the fragments [F1 + F2 + F2] are assembled in a VL assembly vector the right overhang (‘v’) is trimmed, if they are assembled in a VR assembly vector the left overhang (‘u’) is trimmed and if they are assembled in a VM assembly vector both overhangs are trimmed (‘u’ and ‘v’). This overhang trimming process happens during the subsequent rounds of assembly until a VM assembly vector is used in the final round to obtain a fully scarless assembled DNA. MP-dCas9/sgRNA: methylation-protection dCas9/sgRNA molecule.

Overall, the UniClo assembly technique can be used for the assembly of DNA molecules regardless of their sequence to obtain a fully scarless DNA. Here, we describe the experimental procedures involved in UniClo, which uses recombinant methylases for the in vitro methylation and methylation-protection reactions, the dCas9/sgRNA molecule for the methylation-protection itself and the set of three vectors for scarless assembly.

Materials

Plasmids

It is possible to undertake UniClo using just the three types of plasmids, each encoding resistance to one of two antibiotics, kanamycin or chloramphenicol. These core assembly vectors can accommodate large inserts and use the F origin which is a single copy origin derived from the F plasmid. The tight single copy regulation of plasmid number is desirable for stability with large inserts. The assembly vectors are acceptor plasmids into which the assembled DNA is annealed during the assembly. The three-vector set (VL, VM, and VR) of UniClo is available, respectively, as plasmids POC1518, POC1519, and POC1520 for kanamycin selection and POC1553, POC1525, and POC1554 for chloramphenicol selection. These plasmids have all been deposited with Addgene (Table 1). In their starting state, these plasmids contain a relatively small insert that includes a lacZ gene and a ColE1 origin. The lacZ gene allows for blue-white selection using X-gal and IPTG to help identify colonies in which the insert has been removed. The ColE1 origin allows higher copy number replication of the plasmid for easier plasmid preparation in advance of any assembly, but is removed during an assembly reaction.For convenience, we have also developed a set of donor plasmids that have the p15A origin. The p15A origin is a low copy number origin, which allows easier production of higher levels of plasmid DNA than might be obtained with the single copy number F origin. Fragments for a first round of assembly can be cloned or synthesized into these p15A-origin donor plasmids. The cloning can be carried out using Bsa I, Not I or Dra III. These p15A-origin donor plasmids encode resistance to chloramphenicol and they are available with codes from POC1535 to POC1545. These plasmids have all been deposited with Addgene (Table 1). Each fragment to be assembled can be cloned into one of these VL, VM, or VR p15A-origin donor plasmids, depending on whether they are at the start, in the middle or at the end of the sequence to be assembled respectively.

Bacteria

E. coli BL21(DE3)pLysS competent cells, Novagen, Merck, 70236-MNEB^®^ 10-beta competent E. coli, New England Biolabs, C3019INEB^®^ 10-beta electrocompetent E. coli, New England Biolabs, C3020K

Methylases

The recombinant M.Osp807II switch methylase for the targeted methylation of the assembly vectors is expressed using plasmid POC1471 which has been deposited with Addgene (Table 1). The plasmid encodes ampicillin-resistance for selection.The recombinant M2.Eco31I, M2.Eco31I_2, and M2.BsaI non-switchable methylases for the methylation-protection of the fragments in donor plasmids are expressed using plasmids POC1464, POC1466, and POC1468, respectively. These plasmids have been deposited with Addgene (Table 1). The plasmids encode kanamycin-resistance for selection.

Reagents

Tryptone, Fisher Scientific, 12787099Sodium chloride, Scientific Laboratory Supplies, CHE3314Yeast extract, Fisher Scientific, 15815488Bacteriological agar, VWR International, 84609.0500S.O.C. Medium, ThermoFisher Scientific, 15544034Ampicillin sodium salt, Merck, A9518Kanamycin Sulfate, ThermoFisher Scientific, 11815024Isopropyl-beta-D-thiogalactopyranoside (IPTG), ThermoFisher Scientific, R0392Imidazole, Merck, 56750Sodium phosphate dibasic, Merck, S9763Sodium phosphate monobasic, Merck, S3139Glycerol, Merck, G7893Ni-NTA Agarose, ThermoFisher Scientific, R90115Pierce™ BCA Protein Assay Kit, ThermoFisher Scientific, 23227Novex™ Tris-Glycine Mini Protein Gels, 10%, 1.0 mm, WedgeWell™ format, ThermoFisher Scientific, XP00100PK2PageRuler™ Plus Prestained Protein Ladder, 10 to 250 kDa, Fisher Scientific, 1183212420× Bolt™ MOPS SDS Running Buffer, ThermoFisher Scientific, B0001Sample Buffer, Laemmli 2× Concentrate, Merck, S3401-10VLInstantBlue^®^ Coomassie Protein Stain, Abcam, ab119211GeneArt™ Precision gRNA Synthesis Kit, ThermoFisher Scientific, A29377Novex™ TBE-Urea Gels, 6%, ThermoFisher Scientific, EC6865BOXNovex™ TBE-Urea Sample Buffer (2×), ThermoFisher Scientific, LC6876Century™-Plus RNA Markers, ThermoFisher Scientific, AM7145Q5^®^ High-Fidelity 2× Master Mix, New England Biolabs, M0492GeneJET Plasmid Miniprep Kit, ThermoFisher Scientific, K0502Monarch^®^ Spin Plasmid Miniprep Kit, New England Biolabs, T1110Agarose, Merck, A9539SYBR™ Safe DNA Gel Stain, ThermoFisher Scientific, S3310210× TBE Buffer (Tris/Boric Acid/EDTA), BioRad, 1610770Quick-Load^®^ Purple 1 kb Plus DNA Ladder, New England Biolabs, N0552EnGen^®^ Spy dCas9 (SNAP-tag^®^), New England Biolabs, M0652TDiluent B (with rAlbumin), New England Biolabs, B8533SS-adenosylmethionine (SAM), New England Biolabs, B9003Trizma^®^ base, Merck, T1503Ethylenediaminetetraacetic acid (EDTA) disodium salt dihydrate, Merck, E51341,4-Dithiothreitol (DTT), Roche, 10197777001T4 DNA Ligase, New England Biolabs, M0202 Bsa I-HF^®^v2, New England Biolabs, R3733Genomic DNA Clean & Concentrator™ Kit-25, Cambridge Bioscience, D4064NEB^®^ 10-beta Competent E. coli, New England Biolabs, C3019INEB^®^ 10-beta Electrocompetent E. coli, New England Biolabs, C3020KUltraPure™ DNase/RNase-Free Distilled Water, ThermoFisher Scientific, 109770355-Bromo-4-chloro-3-indolyl-beta-D-galacto-pyranoside (X-Gal), ThermoFisher Scientific, R0404Dimethyl sulfoxide (DMSO), Merck, D5879Chloramphenicol, Merck, C0378Ethyl alcohol, pure, Merck, 32221-M Dra III-HF, New England Biolabs, R3510 Not I-HF, New England Biolabs, R3189HiSpeed^®^ Plasmid Midi and Maxi Kits, Qiagen, 12643, 12662, 12663Macherey-Nagel™ NucleoBond™ Xtra BAC, Fisher Scientific, 12713401

Preparation of reagents and buffers

Antibiotics are prepared as 1 mL stock solutions, filter-sterilized and stored at −20°C.

• Ampicillin at 100 mg/mL in UltraPure™ DNase/RNase-Free Distilled Water.
• Kanamycin at 30 mg/mL in UltraPure™ DNase/RNase-Free Distilled Water.
• Chloramphenicol at 25 mg/mL in absolute ethanol.The final concentrations used are 100, 30, and 25 µg/mL for ampicillin, kanamycin, and chloramphenicol respectively, so these are 1000× stock solutions. Thus, for 1 mL of LB medium, 1 µL of antibiotic is used. Isopropyl β-D-1-thiogalactopyranoside (IPTG) is prepared as 0.1 M stock solution in UltraPure™ DNase/RNase-Free Distilled Water, filter sterilized and stored at −20°C. The final concentration used is 0.5 mM.Buffers for the recombinant methylases purification:
• Buffer A: 50 mM sodium phosphate buffer, 300 mM NaCl, and 10% glycerol, pH 7.5
• Buffer 1-Equilibration buffer: as Buffer A plus 10 mM imidazole, pH 7.5
• Buffer 2-Wash buffer: as Buffer A plus 50 mM imidazole, pH 7.5
• Buffer 3-Elution buffer: as Buffer A plus 400 mM imidazole, pH 7.5 dCas9: sgRNA: target DNA ratio 10:10:1 equivalent to 8110 fmol: 8110 fmol: 811 fmol:
• EnGen^®^ Spy dCas9 (SNAP-tag^®^) contains 500 pmol protein in a 25 µL vial. For 8110 fmol (8.11 pmol), 0.4 µL of this solution is required. The protein was diluted in Diluent B using 0.4 µL of protein with 1.65 µL of Diluent B to produce a final volume of 2.05 µL which was used in each methylation-protection reaction.
• sgRNA has 100–120 nucleotides, therefore 313 ng are required for 8110 fmol (https://nebiocalculator.neb.com).
• Donor plasmids ng calculations are carried out using https://nebiocalculator.neb.com. The total target DNA concentration (811 fmol) will be divided by the number of donor plasmids to be methylation-protected for the assembly. For instance, if there are four donor plasmids, 202.75 fmol of each one will be required and if all of them have 5000 bp, 625 ng of each one will be needed to have 202.75 fmol. The four donor plasmids will be mixed together for the methylated-protection reaction. 10× methylase buffer: 0.5 M Tris-HCl, 0.1 M EDTA, and 50 mM DTT, pH 7.5 stored at −20°C.5-Bromo-4-chloro-3-indolyl-beta-D-galacto-pyranoside (X-Gal) is prepared as 20 mg/mL stock solution in DMSO. The final concentration used is 50 µg/mL, so this is a 400× stock solution.

Equipment and consumables

New Brunswick Innova^®^ 44/44R—Stackable Incubator Shaker, EppendorfUltrasonic processors, Vibra-Cell™ VCX 500, AvantorAvanti J-15R Benchtop Centrifuge, Beckman CoulterCentrifuge 5424, EppendorfPrism R™ Refrigerated Microcentrifuge, LabnetCLARIOstar Plus Microplate Reader, BMG LabTechPTC-225 Peltier Thermal Cycler PCR 96-well Tetrad 4 block, MJ ResearchXCell SureLock Mini-Cell Electrophoresis System, ThermoFisher ScientificWide Mini-Sub Cell GT Cell, BioRadChemiDoc™ MP Imaging System, BioRadGene Pulser™ Electroporation system, BioRadBINDER™ Series BD Avantgarde.Line—Standard-Incubator, Fisher ScientificEppendorf^®^ ThermoMixer^®^ F1.5, Merck, EP5384000039NanoDrop™ One/OneC Microvolume UV-Vis Spectrophotometer, ThermoFisher Scientific, ND-ONE-WPierce™ Centrifuge Columns, 10 mL, ThermoFisher Scientific, 89898Disposable PD-10 Desalting Columns—30/pk, Merck, GE17-0851-01Gene Pulser/MicroPulser Electroporation Cuvettes, 0.1 cm gap, BioRad, 1652089MCE 0.05U WH PL 25MM 100PK, Merck, VMWP02500

Software

SnapGene, Version 8.2.0

Methods

Planning an assembly experiment by using UniClo

DNA assembly using UniClo involves the following steps: in silico fragmentation of the DNA construct to be assembled, the design and synthesis of the donor plasmids, transformation, amplification, and purification of the donor plasmids and assembly vector, in vitro methylation-protection of the donor plasmids, in vitro methylation of the assembly vectors, the assembly reaction and the transformation and analysis of the assembled plasmid (Fig. 2). Multiple assembly strategies can be designed using UniClo, depending on the number and size of the fragments into which the sequence to be assembled is subdivided. A final construct to be assembled using UniClo can be broken into multiple fragments, which can be inserted into donor plasmids by direct synthesis or conventional cloning. As outlined in the Materials section, the donor plasmids have the p15A origin and are low copy number plasmids suitable for fragments up to 30 kb [15], typically with chloramphenicol-resistance as the selection marker. They can be obtained by synthesis or any cloning technique, but if by synthesis, it is sensible to break the DNA into fragments of up to 3—4 kb due to the time and cost of DNA synthesis. The fragments can be in VL, VM, or VR donor plasmids, depending on the UniClo strategy for the assembly and whether they are at the start, middle, or end, respectively, of the DNA construct to be assembled. Using UniClo, the fragments to be assembled can include multiple internal type IIS restriction sites for the enzyme used during the assembly (Bsa I). Their design in a VL, VM, or VR donor plasmid must fulfil the following requirements:

Schematic representation of the steps involved in UniClo DNA assembly. MP-dCas9/sgRNA: methylation-protection dCas9/sgRNA molecule.

Each fragment in a VL donor plasmid must include the 4-bp adaptor CTCC at its left end to generate the CTCC overhang (‘u’) after cutting at this Bsa I flanking site. Likewise, it must not have any adaptor at its right end, and the overhang generated after cutting at this Bsa I flanking site will be part of the sequence of the fragment to be assembled (‘b’) (Fig. 3a).Each fragment in a VM donor plasmid must not include any adaptor at its left or right end, and the overhangs generated after cutting at these Bsa I flanking sites will be part of the sequence of the fragment to be assembled (‘b’, ‘c’) (Fig. 3b).Each fragment in a VR donor plasmid must include the 6-bp adaptor TGAGAC at its right end to generate the AGAC overhang (‘v’) after cutting at this Bsa I flanking site. Likewise, it must not have any adaptor at its left end, and the overhang generated after cutting at this Bsa I flanking site will be part of the sequence of the fragment to be assembled (‘c’) (Fig. 3c).

DNA fragment design in the VL, VM, and VR donor plasmids for the first round of UniClo DNA assembly. (a) Start fragments are inserted into VL plasmids including the CCTC adapter at their left end (overhang ‘u’=CCTC). (b) Middle fragments are inserted into VM plasmids without any additional adaptors and their overhangs are derived from the sequence to be assembled (‘b’ and ‘c’). (c) End fragments are inserted into VR plasmids including the TGAGAG adaptor at their right end (overhang ‘v’=AGAC).

The assembly vectors to be used in the different rounds of assembly are the UniClo scarless three-vector set (VL, VM and VR) [14] (Fig. 4). They can be used in either of their two forms, with kanamycin or chloramphenicol selection markers allowing a hierarchical modular assembly. The two forms are identical, other than the selection marker. They have the high copy ColE1 (ori) in their insert and the single copy F (ori2) replication origin in the plasmid backbone, but after the assembly, only the F is kept. The F is a single copy number origin and this makes it suitable for the assembly of long DNA fragments [15]. In a hierarchical assembly, if the first-round donor plasmids encode chloramphenicol-resistance, then the first-round assembly vectors should encode kanamycin-resistance. Therefore, in the second round, the fragments to be assembled will be in plasmids encoding kanamycin-resistance and the second-round assembly vectors should encode chloramphenicol-resistance. Again, in the third round, the fragments to be assembled will be in plasmids encoding chloramphenicol-resistance, and the third-round assembly vectors should encode kanamycin-resistance.

UniClo three-vector set consisting of the (a) VL, (b) VM, and (c) VR scarless assembly vectors. POC1518, POC1519, and POC1520 are kanamycin-resistant forms, while POC1553, POC1525, and POC1554 are the chloramphenicol-resistant forms—the respective VL, VM, and VR plasmids differ only in the selection marker. The overhangs of each of the three assembly vectors when cut during the assembly will always be ‘u’ = CTCC and ‘v’ = AGAC. These overhangs are generated by cutting directed from the inner bsa I sites of the assembly vectors while the outer bsa I sites are methylated during the assembly and so are not cut. The outer bsa I sites on the right of the VL and VM assembly vectors are generated during the assembly [14].

The simplest construct in a first round of assembly will be formed by the assembly of two fragments from donor plasmids. In this situation, the start part of the sequence to be assembled should be in a VL plasmid and the end part in a VR plasmid (Fig. 5a). For the assembly of three fragments, the start part of the sequence to be assembled should be in a VL plasmid, the middle part in a VM plasmid and the end part in a VR plasmid (Fig. 5b). For the assembly of five fragments, the start part of the sequence to be assembled should be in a VL plasmid, the three middle parts in VM plasmids and the end part in a VR plasmid. In all these cases, the resulting flanking overhangs of the assembled DNA fragment will be ‘u’ (CTCC) and ‘v’ (AGAC), compatible with the overhangs of any of the scarless assembly vectors (VL, VM, and VR), which will also be ‘u’ and ‘v’ (Fig. 5). This will also be the case in any subsequent rounds of an assembly. The final round will be carried out in a VM assembly vector to produce a fully scarless assembled DNA molecule (Fig. 1). A given sequence to be assembled could, in principle, be assembled in multiple different ways from the constituent fragments. For example, a sequenced could be divided into 11 fragments for assembly over two rounds (Table 2). In the first round, the 11 fragments could be assembled into 4 fragments consisting of 4, 3, 2, and 2 of the original fragments respectively. These 4 assembled fragments could then be assembled together into one larger fragment in the second round. Alternatively, in the first round the 11 fragments could be assembled into three fragments consisting of 3, 4, and 4 of the original fragments, respectively. These three assembled fragments could then be assembled together into one larger fragment in the second round. The end result of both schemes would be exactly the same.

Overhang compatibility between the fragments to be assembled in VL, VM, and VR donor plasmids and the assembly vectors to achieve a fully scarless DNA assembly. (a) A simple assembly scheme for two fragments. (b) A simple assembly scheme for three fragments. Light blue letters (‘a’, ‘b’, ‘c’, and ‘d’) represent different four base-pair sequences derived from the final sequence to be assembled. The green letters ‘u’ and ‘v’ represent the adaptor overhang sequences required for the start and end fragments for compatibility with any of the assembly vectors. Depending on the type of assembly vector (VL, VM, or VR) used for the assembly, the adaptor overhangs ‘u’ and ‘v’ will be trimmed during the rounds of assembly.

Once the assembly strategy is designed, the donor plasmids are synthesized or cloned, transformed, purified and methylation-protected to methylate and so inactivate any internal Bsa I sites. Similarly, the assembly vectors are transformed, extracted and switch-methylated in their outer Bsa I sites. The assembly reaction is performed using the methylated-protected donor plasmids and the methylated assembly vector. The assembled plasmids are transformed and the transformants are selected using blue-white screening. Plasmids are extracted from the white colonies, and the correctly assembled plasmids are confirmed by restriction digestion and sequencing. These plasmids are then used as donor plasmids in the next round of assembly until the final desired DNA sequence is assembled.

Production and purification of recombinant methylases

Transformation of the plasmids

Add 5 µL (100–400 ng) of one of the plasmids POC1471, POC1464, POC1466, or POC1468, as appropriate, to 20 µL of E. coli BL21(DE3)pLysS competent cells and mix gently by pipetting. Use a 1.5-mL Eppendorf tube previously chilled on ice.Incubate the mixture for 1 h on ice, then for 1 min at 42°C and then for a further 2 min on ice.Add 200 µL of S.O.C. Medium and incubate at 37°C for 1 h at 300 rpm in a ThermoMixer.Spread 50 µL of culture onto LB agar plates with chloramphenicol plus the appropriate antibiotic for each methylase, then incubate the plates overnight at 37°C in a standard incubator.Pick a single colony and inoculate into 10 mL of LB medium containing 10 µL of chloramphenicol plus 10 µL of the appropriate antibiotic in a 50 mL conical tube. Incubate overnight at 37°C in a shaker at 220 rpm.Prepare glycerol stocks by mixing 1 mL of the overnight culture and 1 mL of 50% glycerol and store them at −80°C.

Expression of the methylases

Add 10 µL of glycerol stock to 10 mL of LB medium containing 10 µL of chloramphenicol plus 10 µL of the appropriate antibiotic and incubate overnight at 37°C in a shaker at 220 rpm.Transfer 150 µL of the overnight culture into 150 mL of LB medium containing 150 µL of the appropriate antibiotic. Use 1 L baffled shaker flasks.Incubate at 37°C, shaking at 220 rpm, until the absorbance at 600 nm reaches between 0.4 and 0.6 (ideally 0.45). Typically, this takes around 2 h.Add 750 µL of 0.1 M IPTG and incubate at 20°C in a shaker at 220 rpm for up to 24 h.Centrifuge the culture at 4000×g for 20 min at 4°C. Use 50 mL conical tubes.Store the cell pellets at −20°C until the purification process is carried out.

Purification of the methylases by gravity-flow affinity chromatography

Pack a 3 mL volume (1 column volume) of Ni-NTA Agarose in a Pierce™ Centrifuge Columns of 10 mL.Equilibrate the column with 5 column volumes of Buffer 1.For each methylase, resuspend two pellets of 50 mL culture in 1.25 mL of Buffer 1 for each pellet, and collect together in a single 15 mL conical tube to have 2.5 mL in total.Sonicate the resuspended cells on ice with 20 cycles of 10 s ON and 50 s OFF at 35% amplitude.Centrifuge at 17 200 × g and 4°C for 30 min and recover the clarified lysate.Load the clarified lysate onto the previously equilibrated column.Wash the column with 9 column volumes of Buffer 2.Elute the methylase using Buffer 3 by collecting the eluted volume in 3 fractions: Elution 1, Elution 2, and Elution 3, where Elution 2 contains the highest amount of methylase. Consider the following elution volumes for each methylase:

• M.Osp807II: Elution 1, 2.5 mL; Elution 2, 3 mL and Elution 3, 14.5 mL.
• M2.Eco31I: Elution 1, 2 mL; Elution 2, 2.5 mL and Elution 3, 5.5 mL.
• M2.Eco31I_2: Elution 1, 2 mL; Elution 2, 3 mL and Elution 3, 5 mL.
• M2.BsaI: Elution 1, 2 mL; Elution 2, 3 mL and Elution 3, 5 mL. Desalt Elution 2 of each methylase using PD-10 Desalting Columns and Buffer A. Discard the first 0.5 mL of desalted enzyme and collect the next 3 mL.Measure the protein concentration using the Pierce™ BCA Protein Assay Kit.Analyse using SDS-PAGE by loading 15 µg of protein and on a 10% Novex™ Tris-Glycine Mini Protein Gel, with 1× Bolt™ MOPS SDS as running buffer and PageRuler™ Plus Prestained Protein Ladder, 10 to 250 kDa. Run the gel at 200 V for 30 min.Aliquot the purified methylases into 0.2 mL tubes and store at −80°C for up to 6 months.

Targeted methylation-protection of type IIS restriction enzyme sites in fragments in donor plasmids

Donor plasmids transformation and extraction

If plasmids are lyophilized, resuspend in UltraPure™ DNase/RNase-Free Distilled Water.Transform donor plasmids by adding 5 µL (100–400 ng) of each plasmid to 20 µL of NEB^®^ 10-beta competent E. coli, and spread onto LB agar plates containing the appropriate antibiotic.Pick a single colony and inoculate into 5 mL of LB medium containing 5 µL of the appropriate antibiotic and incubate overnight at 37°C in a shaker at 220 rpm.Perform plasmid purification using the GeneJET Plasmid Miniprep Kit or the HiSpeed^®^ Plasmid Midi Kit and measure the DNA concentration. Elute in UltraPure™ DNase/RNase-Free Distilled Water.

sgRNA in vitro synthesis

Use the GeneArt™ Precision gRNA Synthesis Kit.Perform the PCR assembly using the primers Forward CO9568360F: 5′-TAATACGACTCACTATAGGTGCAGTACCTCTCACGAC-3′ and Reverse CO9569360R: 5′-TTCTAGCTCTAAAACAGTCGTGAGAGGTACTGCAC-3′. The annealing temperature of the primers is 57°C.Elute the purified sgRNA in 10 µL of nuclease-free water and measure the concentration.Analyse by loading 500 ng of sgRNA on a 6% Novex™ TBE-Urea Gel with 1× TBE Buffer as running buffer. Use Century™-Plus RNA Markers and run the gel at 180 V for 80 min. Stain the gel for 30 min using SYBR™ Safe DNA.Store the sgRNA at −80°C for up to 6 months.

Methylation-protection reaction

Calculate the ng of dCas9, sgRNA, and target DNA to be used for the methylation-protection reaction, applying a dCas9: sgRNA: target DNA ratio of 10:10:1 equivalent to 8110 fmol: 8110 fmol: 811 fmol. Use https://nebiocalculator.neb.com for the target DNA calculations with the tool dsDNA: Mass to/from Moles Convertor, using the donor plasmid length (bp) and mass (ng) to reach the fmol needed. If required, the target DNA can be reduced to 405.5 fmol, especially for larger plasmids from the second round of assembly.Set up the three-step methylation-protection reaction in a 0.2 mL PCR tube on ice (three reactions can be carried out for a higher concentration) and run it in a thermocycler:

• First step: dCas9 incubation with sgRN
• 3 µL of 10× NEBuffer™ r3.1
• 313 ng of the sgRNA
• 2.05 µL of EnGen^®^ Spy dCas9 (SNA-tag^®^)
• Make the final reaction volume up to 30 µL with UltraPure™ DNase/RNase-Free Distilled Water
• Mix gently by pipetting and incubate at 25°C for 10 min
• Second step: addition of the target DNA (donor plasmids)
• Add the donor plasmids mix equivalent to 811 fmol (the total molarity of all donor plasmids together, so 811/n fmol per plasmid where n is the number of plasmids)
• Mix gently by pipetting and incubate at 37°C for 15 min
• Third step: addition of non-switchable methylase
• Add 1 µL of 3200 µM SAM
• Add 500 ng of purified recombinant methylase
• Mix gently by pipetting and incubate at 37°C for 15 or 30 min, then heat-inactivate at 80°C for 20 min Purify one single reaction or three pooled reactions (if a higher concentration is required) using the Genomic DNA Clean & Concentrator™ Kit-25 and elute in 25 µL of UltraPure™ DNase/RNase-Free Distilled Water.Measure the DNA concentration and store it at −20°C until the assembly reaction is performed.

Targeted methylation of type IIS sites in the assembly vectors

Assembly vectors transformation and extraction

Transform the plasmids POC1518, POC1519, POC1520, POC1553, POC1525, and POC1554 by adding 5 µL (100–400 ng) of each into 20 µL of NEB^®^ 10-beta competent E. coli, and spreading onto LB agar plates containing the appropriate antibiotics.Pick a single colony and inoculate into 5 mL of LB medium containing 5 µL of the appropriate 1000× antibiotic stock solution and incubate overnight at 37°C in a shaker at 220 rpm.Perform plasmid purification using the GeneJET Plasmid Miniprep Kit and measure the DNA concentration. Elute in UltraPure™ DNase/RNase-Free Distilled Water.

Assembly vectors methylation

Set up the reaction in a 0.2 mL PCR tube on ice (three reactions can be carried out for higher concentration):

• 2000 ng assembly vector
• 2 µL of 10× methylase buffer
• 1 µL of 3200 µM S-adenosylmethionine
• 200 ng of M.Osp807II
• Make the final reaction volume up to 20 µL using UltraPure™ DNase/RNase-Free Distilled Water Mix gently by pipetting and run the methylation reaction in a thermocycler. Incubate at 37°C for 1 h and heat-inactivate at 80°C for 20 min.Purify one single reaction or three pooled reactions (if a higher concentration is required) using the Genomic DNA Clean & Concentrator™ Kit-25 and elute in 35 µL of UltraPure™ DNase/RNase-Free Distilled Water.Measure the DNA concentration and store it at −20°C until the assembly reaction is performed.

Scarless DNA assembly reaction using UniClo

Set up the reaction in a 0.2 mL PCR tube on ice:

• 60 fmol of methylated assembly vector (30 fmol can be used)
• 60 fmol of each methylated protected donor plasmid (30 fmol of each can be used)
• 2 µL of T4 DNA Ligase buffer
• 2.5 µL (1000 U) of T4 DNA Ligase
• 0.25 µL (5 U) of Bsa I-HF^®^v2
• Make the final reaction volume up to 20 µL using UltraPure™ DNase/RNase-Free Distilled WaterUse https://nebiocalculator.neb.com for fmol calculations. Mix gently by pipetting.Run the assembly reaction in a thermocycler at the following conditions: 37°C for 15 min followed by 45 cycles of 37°C for 2 min and 16°C for 5 min, then 37°C for 20 min, 80°C for 5 min, and hold at 4°C.Add 2 µL of 10× CutSmart buffer and 1 µL (20 U) of Bsa I-HF^®^v2.Incubate at 37°C for 3 h and heat-inactivate at 80°C for 20 min.

Transformation of the assembled plasmids

Assembled plasmids less than 10 kb

Add 5 µL of the assembly reaction to 80 µL of NEB^®^ 10-beta Competent E. coli and mix gently by pipetting. Use a 1.5 mL Eppendorf tube previously chilled on ice.Incubate the mixture for 1 h on ice, then for 1 min at 42°C and then 2 min on ice.Add 915 µL of S.O.C. Medium, incubate at 37°C for 1 h at 300 rpm in a ThermoMixer.Spread 50, 100, and 200 µL of culture onto LB agar plates containing the appropriate antibiotic, 100 µM IPTG and 50 µg/mL X-Gal.Incubate the plates overnight at 37°C.

Assembled plasmids greater than 10 kb

Add 50 mL of UltraPure™ DNase/RNase-Free Distilled Water to a 94 × 16 mm petri dish and place a 0.05 µm filter MCE 0.05U WH PL 25MM on the surface, allow the filter to wet completely for around 5 min.Transfer the 23 µL of the assembly reaction to the centre of the filter and leave it there for 1 h at room temperature for drop dialysis.Retrieve the assembly reaction droplet carefully from the filter in a clean 0.2 mL PCR tube.Add 5 µL of the dialyzed assembly reaction to 25 µL of NEB^®^ 10-beta competent E. coli. Use a 1.5 mL Eppendorf tube previously chilled on ice.Mix very gently by pipetting and carefully transfer the DNA/cell mix into a chilled 0.1 cm electroporation cuvette.Electroporate at 0.9 kV 100 Ω and 25 µF, and immediately add 1 mL of NEB^®^ 10-beta Stable Outgrowth Medium pre-warmed at 37°C, then mix by pipetting.Transfer the cells to a 15 mL conical tube and incubate at 37°C for 1 h, shaking at 220 rpm.Spread 50, 100, and 200 µL of culture onto LB agar plates with the appropriate antibiotic, 100 µM IPTG and 50 µg/mL X-Gal.Incubate the plates overnight at 37°C.

Analysis of the transformants and the assembled plasmids

Identify and count the blue and white colonies obtained as a result of the blue-white screening.Pick single white colonies and inoculate them into 5 mL of LB medium containing the appropriate antibiotic in a 50 mL conical tube and incubate overnight at 37°C, shaking at 220 rpm.Prepare glycerol stocks by mixing 1 mL of the overnight culture and 1 mL of 50% glycerol and store them at −80°C.Add 5 µL of glycerol stock to 5 mL of LB medium with 5 µL of the appropriate antibiotic and incubate overnight at 37°C, shaking at 220 rpm.Perform plasmid purification depending on the size of the assembled plasmid using the GeneJET Plasmid Miniprep Kit (up to 10 kb), Monarch^®^ Spin Plasmid Miniprep Kit (up to 25 kb), HiSpeed^®^ Plasmid Midi and Maxi Kits (up to 50 kb) or NucleoBond™ Xtra BAC (up to 300 kb). Elute in UltraPure™ DNase/RNase-Free Distilled Water.Analyse the extracted plasmids by restriction digestion in a 10 or 25 µL reaction using Dra III or Not I, which separate the vector backbone from the assembled construct. Run a 1% standard agarose gel at 120 V to confirm the correct assembly. Use SYBR™ Safe DNA Gel Stain, 1× TBE Buffer and Quick-Load^®^ Purple 1 kb Plus DNA Ladder.For large assembled plasmids or to confirm the results of restriction digestion, use long-range sequencing.

Next rounds of assembly

Add 5 µL of glycerol stock of the colonies containing the corrected assembled plasmids (new donor plasmids) to 5 mL of LB medium containing 5 µL of the appropriate antibiotic and incubate overnight at 37°C, shaking at 220 rpm.Perform plasmid extraction depending on the size and required concentration of the new donor plasmid using the GeneJET Plasmid Miniprep Kit (up to 10 kb), HiSpeed^®^ Plasmid Midi and Maxi Kits (up to 50 kb) or NucleoBond™ Xtra BAC (up to 300 kb). Elute in UltraPure™ DNase/RNase-Free Distilled Water.Follow the procedure described above in targeted methylation-protection of type IIS sites in donor plasmids, scarless DNA assembly reaction using UniClo, transformation of the assembled plasmids and analysis of the transformants and the assembled plasmids for each round to achieve the final assembled construct.

Advantages

UniClo has three significant advantages over other reported Golden Gate-based assembly methods. Firstly, it allows the assembly of any DNA sequence regardless of the presence of internal restriction sites for the single type IIS enzyme used during the assembly. Previous techniques typically mutate these sites (domestication) before the assembly, causing alterations in the original sequence [16]. In UniClo, these internal type IIS sites are in vitro methylated by non-switchable recombinant methylases. During this methylation, the outer flanking type IIS sites required for the assembly are in vitro methylation-protected by a dCas9/sgRNA molecule. In this way, the original DNA sequence can be assembled without any sequence modification. Secondly, a fully scarless DNA assembly is achieved regardless of the sequence being assembled. Earlier techniques generally use assembly vectors with multiple standard overhangs [9, 17, 18]. At the end of the assembly, these overhangs are present as unwanted scars in the final assembled sequence. Although some of the existing techniques have used, for example, start and stop codons for overhang design [3], this is not universally applicable for any sequence. In UniClo, a fully scarless assembly is achieved by using vectors designed to trim the overhang sequences (scars) from the assembled fragments, leaving the unwanted overhangs in the donor plasmid backbone during the assembly process. The VL and VR assembly vectors trim the right and left overhangs of the start and end fragments of the DNA to be assembled, and the VM assembly vector trims both overhangs of the middle fragments or of the final assembled DNA construct allowing its release without any scar. Thirdly, UniClo only requires three engineered vectors and two universal overhangs for any DNA fragment to be assembled in multiple rounds. The three vectors always have the same overhangs (CTCC and AGAC), which are compatible with the two adapter overhangs of the fragments in donor plasmids. These overhangs are trimmed by using the scarless assembly vectors. In contrast, existing techniques typically use a number of assembly vectors and adaptors for each round of assembly, which can make the process extremely laborious and relatively inflexible [3, 5, 9, 12].

Technical limitations

UniClo required recombinant methylases which must be expressed and purified. However, this is straightforward using the protocol described above and they retain their activity for up to 6 months. UniClo shares limitations with other DNA assembly approaches, in particular the need for synthesis of donor fragments, which can be variable in terms of cost and time depending on the size of the fragments and their complexity. A further shared limitation is that as the size of the DNA increases, the efficiency of transformation falls. However, only one colony containing the correct assembled plasmid is required, so this is not crucial for the success of the technique.