Some tips for troubleshooting GoldenGate TALEN assembly

[banni...@gmail.com]

Hi there!

We have been using the GoldenGate Assembly method for making TALENs in our lab for the past year and a half. Seeing that there are often questions regarding the use of this protocol posted on these forums I thought I would share with you all some of the problems and solutions we have encountered while using this system. I hope that our experience is helpful for some of you who may be new to the technology, or experiencing similar problems. I have to say that some of our solutions have been found through suggestions posted on these forums and to all of you who are active in discussions a big THANK YOU. Let's keep up the great community spirit.

Ok here goes:

Most common causes of our GG reaction failures:

Bad ligase buffer:
We always aliquot ligase buffer (only NEB T4 10X LB) into small aliquots and always use fresh aliquots for step 1 reactions.

Mutations in pFUS-A:
Once we found very strange but strong amplification of ~700bp bands and other weird laddering patterns in step 1 colony PCR products. We found out that the problem was a mutation in one of the BsaI cloning sites in the pFUS-A vector. This we figured out by sequencing the vector, and by doing a test digest of the pFUSA vector alone with BsaI. Making a new plasmid prep from the original glycerol stock got rid of the problem. We don't know how or why this mutation suddenly cropped up. Given that it was also just before we figured out using BsaI-HF doesn't work (see below) - it could also be that BsaI-HF was involved and partially to blame for this too.

Using BsaI-HF:
We recently had 2 weeks of failed step 1 reactions and thanks to previous posts on these forums figured out that it's because we were using BsaI-HF (NEB) instead of the normal BsaI. Short story is DON'T USE BsaI-HF!!! We know this wasn't any other problem because two different people had the same results trying to make different TALENs. When we repeated the same failed reactions this last week with normal BsaI, everything worked fine. It does work with very low efficiency for step 1 reactions to assemble short RVD arrays into pFUS-B vectors, but never for pFUS-A or B arrays longer than 5 RVDs. We don't exactly know why the HF doesn't work but I strongly suspect this has something to do with the buffer conditions being quite different in the TALEN GG reactions versus the cutsmart buffer that comes with the HF enzyme. The moral of the story: "HF" does not equal better enzyme in all cases.

Low efficiency competent cells:
We use both high competency home made XL1 blue cells or Life TEchnologies TOP10 F' cells for transforming step 1 reactions. Both work pretty well but the TOP10 are always more efficient. Depends a bit on how you do transformations, but we use heat shock, recover cells in a total volume of about 300-500ul of SOC media and get mostly white colonies if we plate upwards of 100ul of transformed cells.

Second step reactions fail because of bad Esp3I
The only times we had trouble with second step reactions failing appeared to be due to Esp3I enzyme not working properly. We could only come up with the explanation that we were frequently using the same large volume aliquot of the enzyme and that perhaps the regular movement in and out of the -20 freezer had affected the activity. We then switched to ordering smaller aliquots of the enzyme and haven't had troubles since. Incidentally we always use Fermentas Esp3I for step 2 reactions.

I have to say that apart from these few minor yet annoying problems, we have had great success using the GG assembly method since the very beginning. I believe that the efficiency of the system does depend on having good quality reagents and making sure that your plasmid preps, glycerol stocks, enzymes, buffers, competent cells, media and antibiotics for selection are all high quality and regularly maintained. We also made working glycerol stocks for all of the plasmids so that we can keep the original glycerol stock plate untouched as a final backup. This was really important when we had problems with the pFUS-A mutation: Re-prepping from the original plate brought everything back to normal.

Just for information we have used both the pTAL3 and the pCS2TAL-DD/-RR vectors for step 2 reactions as well as a home-made pCS2+ backbone vector (essentially the pTAL3 TALEN sequences in the pCS2+ backbone) there doesn't seem to be any obvious difference in the cloning efficiencies using these different vectors, but the -DD/-RR vectors are more useful for in vitro mRNA transcription for microinjection so we use those almost exclusively now.

Good luck to you all with your TALEN assembly!
Steph Bannister

[Luisa]

Hi there,

 

and thanks for sharing your experience.

 

Regarding the BsaI HF issue: as far as I know, the problem is that the HF version of BsaI is not (as) active at 50°C as the 'good, old, normal' BsaI. As the last 5 minutes at 50°C are essential to cut all unspecific ligation products in a GG cloning, if this step fails then the efficiency of the whole reaction drops dramatically.

 

I think at 37°C both enzyme versions work just fine (and the buffer probably also doesn't matter...).

 

If someone else could comment on this, it would be interesting....

 

Thank you, ciao

Luisa

[alvaro sequeida]

Thnax for your advises Steph !! I've been having trouble assembling into pFUS_A, but not in the pFUS_Bs I use, so I'll try making a new prep from the original glycerol stock. I have some questions, do you use 150 or 75 ng/ul of pFUS vectors in the GG1? Also, do you add BSA in the first GG1? I ask you this because in the manual that comes with the Golden Gate kit, they don't describe using BSA, but I add it anyways, just because I've never tried any enzymatic reaction without BSA. In the GG2, there you need 75 ng/ul of destination vector, right?

Thanxs in advance !

[banni...@gmail.com]

Hi Luisa

Ah thanks for the info on the 50 degree step. So if I understand correctly, it can be that the correct pFUS plasmid is still normally assembled, but without cutting away the non-specific products without a ligation step then this is going to bias the transformation of bacteria with non-specific plasmid ligation products versus the correctly assembled plasmid? This would make sense as to why we were getting colony PCR bands that were just the size of the empty pFUS-A versus the correctly sized bands.

Thanks a lot for your reply!
Steph

[banni...@gmail.com]

Hi Alvaro,

We use the GG protocol that comes with the Addgene kit pretty much exactly as stated. We do use BSA in the step 1 reactions though, because this is listed on the reaction lists that one can generate from the GoldenGate Assembly Excel Macro form. We use 150ng of each pFUS and individual RVD plasmids for step 1 reactions and haven't tried other concentrations because this worked perfectly the first time (and basically every other time if everything is ok with the pFUS and the enzymes/buffer).

Steph

[robr...@gmail.com]

Hi Alvaro,

Just a suggestion, try sequencing your pFUSA plasmids. I was having trouble (I still am), but found out that there was a pretty significant mutation in my pFUSA plasmid which wiped out one of the BsaI sites. I re-prepped from the stocks and the re-prep is completely fine! I've learned it's best to sequence everything with this kit.

Good luck,

Rob