Troubleshooting Golden Gate Assemblies

Tips

* Screen Inserts for internal BsaI/BsmBI sites! Reactions with single off-target sites can work, but the last step of the assembly protocol should be a ligation. You may need to edit these internal sites out.

* Clone your parts into vectors. This is especially useful for inserts < 250 bp or > 3kb, or inserts containing repetitive elements that might accumulate errors during PCR amplification. For other modules, PCR amplicons can be used in place cloned parts to save time.

* A 2:1 insert:destination plasmid ratio is recommended, although the Golden Gate Assembly process is robust enough that 1:1 ratios also can be used. For PCR amplicons, the amount of each insert to be added can be calculated by molar calculations or relative length calculations.

* Single insert cloning is more efficient than multiple insert cloning. Assembly efficiency decreases as the number of fragments increases. The presence of repetitive sequences in an insert will also decrease efficiency. For inserts < 250 bp or > 3 kbp, pre-cloning these inserts into an entry vector increases efficiency.

* The normal restrictions on overall plasmid size to allow transformation and stable maintenance in E. coli apply to Golden Gate assemblies. Efficiencies are highest when the product plasmid is < 12 kb. Larger assemblies can be made but may require larger numbers of colonies to be screened for the correct full length assembled products.

* Cycling assemblies using 5-10 min temperature steps work well for larger scale assemblies (>10 inserts) and for any assembly for which maximal assembly yields and transformation levels are desired.

* Scale down reaction size from 20 µL to 10 µL - this will increase the proportion of restriction enzyme to the rest of your components and can help with efficiency (can be done on spreadsheet)

Troubleshooting

• No colonies on selective plate - This can be caused by many issues. Check the following:
  • Are your competent cells very competent? Low efficiency competent cells cause a working assembly to appear broken. Electrocompetent E. coli should be of ~1e4cfu/ng puc18 efficiency.
  • Are you plating the entire recovery mixture? Even with high efficiency cells, the number of successfully transformed cells can be quite low with multipart assembly (6+ parts). Spin down, resuspend, and plate the entire transformation on your selective media.
  • Are your parts correct? Over time plasmids can mutate, be mislabelled, or suffer from DNase contamination. Sequence all of your parts, and check for degradation.
  • How much of your assembly did you transform? Use 2µL to transform electrocompetent cells, and more if using chemically competent cells.
• High number of fluorescent colonies - When using a fluorescent dropout vector (pYTK001, pYTK095, pBTK403, pBTK402), the majority of your colonies should be non-fluorescent. A high number of fluorescent colonies indicates these dropouts are not being cut efficiency, and is often caused by the following:
  • Expired BsaI or BsmBI enzymes. As with other enzymes, BsaI and BsmBI can lose their potency over time or with extensive handling at room temperature. Perform a diagnostic digest, or order and use fresh BsaI / BsmBI.
  • Wrong cycling conditions. Use 37 deg for BsaI assemblies and 42 deg for BsmBI assemblies. Cutting times can be extended from 1:30 to 3:00 or 5:00 per cycle. Increase number of cycles to 30 or higher.
• Repeated mutations in assembled plasmids or part plasmids - Caused by costly, toxic, or unstable inserts. Try the following:
  • Pick and screen more colonies.
  • For part plasmids, use a stable backbone. pBTK001 is a p15A origin gfp drop-out vector. Ask Peng or Sean about versions of pTYK001 with flanking terminator sequences.
  • For first-stage plasmids, use an inducible promoter to minimize expression of toxic inserts.

Alternate Thermal Cycler Conditions

* If you're not using a pre-screened parts kit (YTK, MoClo, etc), occasionally your PCR amplified parts will contain internal BsaI or BsmBI sites. In these situations, it's critical you end the assembly on a ligation step. Even with 2 internal BsaI sites, I've effectively used the following two-step, non-cycling protocol:

• For these reactions, add only BsaI initially (total volume 19µL)
• Step 1: 37°C for 30 minutes (optimal cutting temperature for BsaI)
• Step 2: 65°C for 20 minutes (heat inactivation of BsaI)

• Step 4: 25°C for 30 minutes (optimal ligation temperature for T7 ligase)
• Step 5: 65°C for 20 minutes (heat inactivation of T7 ligase)
• Step 6: Electroporate 1-2µL reaction into appropriate electrocomp cells (larger volumes inhibit transformation). If using chemically competent or Z-comp cells, you can add arbitrarily large amounts of reaction.

* If you feel like your efficiency of assembly is still extremely low, sometimes using a longer thermocycling protocol can help (BsaI/BsmBI):