Plasmid Construction: From Restriction Cloning to Gibson Assembly and the Discipline of "Read the Map Before You Build"

Restriction–ligation · PCR-based cloning · Gibson · Golden Gate/MoClo · Primer design · Mutagenesis · Vector choice · Validation

Published 2026 · Last updated 2026-09-26 · Hailu Zhou

Plasmid construction is the quiet infrastructure of modern molecular biology. People talk about CRISPR and single-cell RNA-seq, but most labs still run on vectors: expression plasmids, reporter constructs, sgRNA backbones, shRNA cassettes, tagging vectors, bacterial artificial chromosomes, and shiny new base-editor plasmids. A plasmid is not "a circle of DNA." It is a legal contract between you and the cell: origin of replication says who copies me, antibiotic gene says who survives me, promoter says when I speak, ORF says what I make, and terminator says where I stop.

Restriction–ligation cloning

The oldest reliable method is restriction–ligation cloning. You cut vector and insert with compatible restriction enzymes, sometimes creating sticky ends (EcoRI, BamHI) or blunt ends (EcoRV, SmaI), then ligate with T4 DNA ligase. Its virtue is simplicity and predictability; its vice is reliance on unique sites. If your insert contains an internal EcoRI site, EcoRI cloning becomes a trap. Experienced builders therefore read the full sequence map before touching enzymes. A good plasmid map lists: ori, selectable marker, promoter, RBS/Kozak, MCS, tags, introns (if mammalian), recombinase sites (LoxP/Flp), and all restriction sites used.

PCR-based and homology-based assembly

PCR-based cloning changed the game. You amplify the insert with primers that carry overhangs, then clone by homology. Gibson Assembly uses a 5′ exonuclease to generate single-stranded overlaps, a polymerase to fill gaps, and ligase to seal nicks; fragments with ~20–40 bp shared ends assemble in one isothermal reaction. It scales beautifully: two fragments or five fragments, linearized vector or PCR-amplified backbone, all in one tube. In-Fusion is conceptually similar but proprietary. Golden Gate / Type IIS cloning uses enzymes like BsaI/BsmBI that cut outside their recognition site, letting you assemble multiple parts in one pot with scar-less junctions and standardized part libraries. MoClo and Golden Braid took this into synthetic-biology territory: promoters, CDSs, tags, and terminators become LEGO bricks.

Primer design for cloning

A non-obvious skill is primer design for cloning. For Gibson, the 5′ tail must match the destination backbone, not the template. For Golden Gate, the tail must contain the Type IIS overhang and the junction sequence you want after cleavage. Tm calculations lie if you ignore tails; the insert-binding region and the homology/adaptor region have different jobs. Adding restriction sites usually needs 4–6 extra bases upstream of the site or the enzyme will not cut efficiently. Adding tags (His₆, FLAG, HA, Myc, GST, MBP, SUMO) means thinking about protease sites: TEV, thrombin, Factor Xa, 3C protease. A bad cleavage site ruins purification later.

Mutation introduction

Mutation introduction is its own subcraft. Site-directed mutagenesis (SDM) by overlap-extension PCR or QuikChange-style whole-plasmid amplification lets you change one residue, delete a domain, or introduce a stop codon. The plasmid must be methylated-template-resistant polymerase (DpnI digestion removes the original dam-methylated template from DH5α). People forget that DpnI only cuts methylated DNA: if you transform without DpnI, colonies are dominated by unchanged template. Sequencing must cover the junction and the whole insert—not just the mutated codon—because PCR polymerases make errors.

Vector choice is a biological decision

pET vectors put expression under T7 RNA polymerase, which works only in DE3 strains carrying lacUV5-controlled T7 pol. pBAD uses arabinose; pTrc/pTac use lac/trp hybrids. Eukaryotic expression needs CMV/SV40 promoters, polyA signals, and often EBNA/oriP for episomal maintenance. Bacterial expression vectors often lack eukaryotic splicing, so cDNA—not genomic DNA—is mandatory. Adding an N-terminal tag changes translation initiation; adding a C-terminal tag can destabilize the protein or block secretion. Every base matters: a missing start codon, a premature stop, a frameshift from a bad primer, and the construct "expresses" only in your notebook, not in the cell.

Validation is not "I saw a band"

A correct pipeline: (1) colony PCR for insert presence; (2) restriction digest of miniprep to check size; (3) Sanger sequencing of insert + junctions; (4) optionally NGS for large assemblies; (5) transformation into expression strain only after sequence confirmation. A classic PhD mistake is transforming BL21 with an unsequenced plasmid, getting a 70 kDa smear, and spending three months blaming induction conditions when the clone had a frameshift from day one.

Plasmid construction is interface engineering

The philosophical point: plasmid construction is interface engineering. You are designing how DNA will be read by polymerases, ribosomes, recombinases, and future collaborators. Sloppy cloning propagates forever—into proteins that misfold, mice that express the wrong isoform, and figures that quietly describe a construct nobody actually built. Good builders are boring: they check maps, align primers, sequence everything, and keep a clean plasmid inventory with version numbers. Boring is reproducible.

References

  1. Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA 3rd, Smith HO. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods. 2009;6(5):343–345.
  2. Engler C, Kandzia R, Marillonnet S. A one pot, one step, precision cloning method with high throughput capability. PLoS One. 2008;3(11):e3647.
  3. Quan J, Tian J. Circular polymerase extension cloning for high-throughput cloning of complex and large DNA fragments. Nat Protoc. 2011;6(2):242–251.
  4. Green MR, Sambrook J. Molecular Cloning: A Laboratory Manual. 4th ed. Cold Spring Harbor Laboratory Press; 2012.
  5. Shetty RP, Endy D, Knight TF Jr. Engineering BioBrick vectors from BioBrick parts. J Biol Eng. 2008;2:5.
  6. Weber E, Engler C, Gruetzner R, Werner S, Marillonnet S. A modular cloning system for standardized assembly of multigene constructs. PLoS One. 2011;6(2):e16765.
  7. Sarrion-Perdigones A, Falconi EE, Zandalinas SI, et al. GoldenBraid: an iterative cloning system for standardized assembly of reusable genetic modules. PLoS One. 2011;6(7):e21622.