Golden Gate Assembly

Golden Gate Assembly is a one-pot cloning method that uses Type IIS restriction enzymes — most commonly BsaI, BsmBI or BbsI — to join multiple DNA fragments in a defined order. Type IIS enzymes cut at a fixed distance outside their recognition sequence rather than within it, so the 4 nucleotide overhang they leave behind is sequence you choose, not sequence the enzyme dictates. Because the recognition sites sit on the fragments being discarded, they are removed during assembly and the final construct is scarless and cannot be re-cut, which lets digestion and ligation run simultaneously in a single tube.

Why do Type IIS enzymes matter?

A conventional Type II enzyme such as EcoRI cuts inside GAATTC and always leaves the same overhang, so every fragment cut with it is compatible with every other and the assembly order is random. A Type IIS enzyme such as BsaI recognises GGTCTC and cuts one nucleotide downstream, leaving a 4 nt overhang whose sequence comes from the adjacent DNA. Different fragments therefore carry different, non-interchangeable overhangs, which is what makes ordered multi-fragment assembly possible.

How do I choose the overhangs?

Overhang choice is the single biggest determinant of whether a Golden Gate reaction works. Potapov et al. (2018) measured ligation fidelity for all 256 possible 4 nt overhangs on T4 DNA ligase and showed that some pairs mis-ligate at appreciable rates. Modern design tools score a candidate overhang set against that empirical data.

  • Use overhangs that differ from each other by at least two nucleotides.
  • Avoid palindromic overhangs — they ligate to themselves.
  • Avoid overhang sets that are all-AT or all-GC.
  • Do not reuse the same overhang twice in one reaction unless you intend those junctions to be interchangeable.
  • Verify that no internal BsaI or BsmBI site exists in any fragment; domesticate the sequence first if one does.

What does the reaction look like?

Digestion and ligation are cycled in the same tube. A typical programme alternates 37 °C (enzyme cuts) and 16 °C (ligase joins) for 25–35 cycles, then a final 60 °C step to inactivate the enzyme and drive the reaction toward the assembled product. Correctly assembled junctions destroy the recognition site, so the product accumulates while unassembled fragments keep being re-cut.

When should I use something else?

Golden Gate requires that no fragment contain an internal copy of the recognition site, so a sequence full of BsaI sites needs domesticating first, which costs a round of mutagenesis. If you are joining a couple of one-off PCR products and do not need a reusable part library, Gibson Assembly avoids that constraint entirely and needs no enzyme-site planning.

Common Type IIS enzymes used in Golden Gate Assembly

EnzymeRecognition siteCut offsetOverhang
BsaIGGTCTC1 nt downstream4 nt, 5′
BsmBI (Esp3I)CGTCTC1 nt downstream4 nt, 5′
BbsI (BpiI)GAAGAC2 nt downstream4 nt, 5′
SapI (LguI)GCTCTTC1 nt downstream3 nt, 5′
AarICACCTGC4 nt downstream4 nt, 5′

Frequently asked questions

What is the difference between Golden Gate and Gibson Assembly?
Golden Gate uses Type IIS restriction enzymes and defined 4 nucleotide overhangs, so it excels at combinatorial assembly from a standardised part library and runs as a one-pot digest-ligate cycle. Gibson uses 15–40 bp sequence homology and three enzymes at a constant 50 °C, so it needs no enzyme sites at all and handles arbitrary one-off fragments better. Golden Gate requires that no fragment contain an internal recognition site; Gibson has no such constraint but struggles with repeated sequence.
Why is Golden Gate Assembly called scarless?
Because the Type IIS recognition sites are positioned on the portions of DNA that get cut away, they are not present in the assembled product. The junction contains only the 4 nucleotide overhang sequence you designed, which can be chosen to be part of the intended construct — for example a codon boundary — so no extra bases are left behind.
How many fragments can Golden Gate assemble at once?
With well-chosen overhangs, Golden Gate routinely assembles 10 or more fragments in a single reaction, and published high-fidelity overhang sets support assemblies well beyond that. Fragment count is limited mainly by how many mutually distinguishable overhangs you can design, which is why empirical fidelity data matters.
What does BsaI recognise?
BsaI recognises the non-palindromic sequence GGTCTC and cuts one nucleotide downstream on the top strand, leaving a 4 nucleotide 5′ overhang whose sequence is determined by the adjacent DNA rather than by the enzyme. This is what allows a designer to specify the assembly order.

References

  1. Engler C, Kandzia R, Marillonnet S. A one pot, one step, precision cloning method with high throughput capability. PLoS ONE 3(11): e3647 (2008). https://doi.org/10.1371/journal.pone.0003647
  2. Potapov V, Ong JL, Kucera RB, et al. Comprehensive profiling of four base overhang ligation fidelity by T4 DNA ligase and application to DNA assembly. ACS Synthetic Biology 7(11): 2665–2674 (2018). https://doi.org/10.1021/acssynbio.8b00333

Related

Design your Golden Gate Assembly construct in Talindrew.

Free to use. No download required. Works in any modern browser.

Start for free