Golden Gate Assembly¶

In previous tutorials, we built the pET-INS plasmid using both traditional restriction enzyme cloning and Gibson Assembly. In this tutorial, you'll use Golden Gate Assembly to build the same construct—while learning how this method enables precise, scar-controlled, multi-part DNA construction.

What is Golden Gate Assembly?¶

Diagram of Golden Gate Assembly showing a DNA fragment with BsaI recognition sites flanking the sequence. After digestion, the enzyme produces non-palindromic sticky ends which guide correct ligation orientation. Final product is a seamless joint between two DNA fragments.

Golden Gate Assembly is a method for joining DNA fragments using Type IIs restriction enzymes like BsaI, BsmBI, BbsI, and SapI. These enzymes cut a fixed number of bases away from their recognition sites, which allows the creation of custom 4 bp overhangs that control exactly where and how parts join together.

Golden Gate is powerful because it allows:

Custom-designed sticky ends for precise, seamless (or intentionally scarred) ligation
Single-pot digestion and ligation, increasing efficiency
Automatic removal of restriction sites from the final product

These features make it ideal for modular cloning, where standardized parts and repeatable junctions are critical.

The reaction is designed so that digestion and ligation happen together, recognition sites are eliminated, and sticky ends control the final sequence junctions.

⚠️ This means:

The recognition site must not appear in the final product, or it will be re-cleaved.
The method can be iterated using the same enzyme again and again, since the site is removed during assembly.

Designing Golden Gate Oligos¶

Let’s now walk through a complete Golden Gate Assembly plan to make pET-INS, just like we did in the Gibson version.

Step 1: Define Your Product¶

Start by constructing a model of the final pET-INS plasmid, just as you did with Gibson.

Open the pET28a vector sequence and convert it to UPPERCASE.
Open the insulin cDNA sequence and convert it to lowercase.
Paste the INS cds into the intended insertion site as done in the basic cloning tutorial.

🔗 Downloads:

Step 2: Annotate Junctions and Annealing Regions¶

Once you’ve created a model of your final product, define these regions:

For each junction where two fragments will join:

A. Choose the Sticky End¶

Choose 4 bp from the existing sequence at the junction to serve as the sticky end. These should be non-palindromic (e.g., avoid AATT, GATC, etc.) to ensure correct orientation.
Create a feature annotation called "sticky end" over those 4 bases.
In the visualization below, these are highlighted in orange as “sticky end 1” and “sticky end 2”, corresponding to the two junctions in this construct.

B. Mark the Forward Anneal Region¶

Starting at the junction and extending downstream, choose 20–30 bp that follow standard primer design rules. Label this forward anneal.

C. Mark the Reverse Anneal Region¶

Identify 20–30 bp upstream (5′) of the junction. This sequence lies on the coding strand, not its reverse complement. Label it reverse anneal.

Step 3: Design Oligos¶

Design the junction + annealing region for each oligo first. These are the 3′ ends that will bind to your template.

A. Forward Oligo¶

From the forward strand of your product, copy the 4 bp sticky end plus ~20 bp downstream (i.e., to the right).
This is your junction + anneal region.

B. Reverse Oligo¶

From the reverse strand of your product, copy the 4 bp sticky end plus ~20 bp upstream (i.e., to the left).
Reverse complement this sequence. This is your junction + anneal region for the reverse oligo.

C. Add the Golden Gate Prefix¶

Add the following prefix to both the forward and reverse oligos:

ccataGGTCTCa

This includes:

ccata = arbitrary 5' tail
GGTCTC = BsaI recognition site
a = one-base arbitrary spacer before the sticky end

Final orientation of Golden Gate oligos (both forward and reverse):

5' tail - BsaI - spacer - sticky end - annealing region - 3'

⚠️ Important: For the reverse oligo, only the junction + annealing region is reverse complemented. The prefix remains in forward orientation.

🧬 5′ tail design tip: NEB data shows BsaI cuts efficiently with as little as 1 bp upstream of its site. While 5 bp is a safe general rule, it's often more than necessary.

Do the same for each junction. Realistically, you can do this PCR-based variant of Golden Gate for up to 4 fragments. You can also do golden gate with clonal plasmid DNA. With the higher quality DNA and the restriciton sites being further from the ends, plasmid-based golden gate is much more efficient for multi-fragment assembly than the PCR variant.

Here is an example solution:

PCR         insF2       insR2       insulin_cds      ins_pcr
PCR         vecF2       vecR2       pET28a           vec_pcr
GoldenGate  vec_pcr     ins_pcr     BsaI             pET-INS

oligo       insF2       ccataGGTCTCaatggccctgtggatgcgcctc
oligo       insR2       cagatGGTCTCaCGAGctagttgcagtagttctccag
oligo       vecF2       ccataGGTCTCaCTCGAGCACCACCACCACCAC
oligo       vecR2       cagatGGTCTCaccatGGTATATCTCCTTCTTAAAG

Step 4: Simulate It¶

You can simulate the PCR steps of the construction file as you've done previously with your sequence editor or an automation tool. ApE and Benchling both have visual tools to simulate the Golden Gate step as well. C6 can also simulate your entire construction file, and you can try it out in the box below.

💡 Tip: If you’re unsure about orientation after simulating, remember there are only two possibilities—it’s either the sequence you expect or its reverse complement.

🧪 Try It: Swap the Promoter¶

In this quiz, you will replace the T7 promoter in pET-INS with the AraC_Pbad promoter using Golden Gate Assembly.