We are new to Gibson assembly and are using the NEB Gibson assembly master mix. Reactions with vector alone (no insert) give no colonies. Adding an inset of 660bp gave 10s of colonies. But a large insert of 1200 and another of 3700 gave nothing. NEB suggested not to gel purify the PCR products being inserted. So we simply spin-purified the PCR reaction. We obtained 6 colonies in that case after plating 80% of the cells (NEB suggests plating 10% of the cells). My impression was Gibson assembly for inserting a single piece of DNA to a vector should be more efficient than this? And why does a gel-purified small insert work but not a gel-purified large insert? Should we be doing something differently or using another source of reagents?
Some details: we do the assembly for 60 min at 50C with 40-50 ng of a 4790bp vector with a 3-5 fold molar excess of each insert. In some cases the insert is slightly diluted (1:1 or 1:2) before adding a microliter to the reaction, in other cases we add undiluted, purified insert (up to 1.5 ul).
Thanks for any advice and sharing your experience.
For 4-6 fragments being assembled to achieve optimal assembly efficiency, you should design 25-30 bp overlap regions between each fragment with equimolarity of all fragments. I used 0.05 pmol each and it worked!
However, for 2-3 being assembled the optimized cloning efficiency is 50–100 ng of vector with 2-fold excess of each insert . (You can use 5-fold molar excess of any insert(s) less than 200 bp). Total volume of unpurified PCR fragments in the assembly reaction should not exceed 20%.
In addition, you can extend the incubation time by up to 4 hours to optimize the reaction.
Hi, the Gibson reaction is usually straightforward and should give more colonies than what you are seeing. The amounts you are using look correct and the protocol looks pretty good as well. I do admit that it did take a few times to get it down pat, but once you have it, it works very well, so I would still try to get it to work.
For me personally, the two areas that I have found problems are with the proper design of the overlapping regions and the volumes and method of transformation.
The first question would be how long are the overlapping regions and are they in the proper orientation and location? This also comes up when the vector is digested versus amplified as well....this is where I had some problems initially. At first I tried to use the minimum overlapping length, but I found that if you can go longer, it works better.
The second question is how much of your reaction are you using to transform? The Gibson mix itself is not very amenable to transformation....I find that if I take 3-5ul of mix and dilute it with water, it works a bit better.
I find that when I do my gibson reactions, I go for a volume of 10uL. If the volumes become greater, it should be no problem, I do a quick micro-PCR cleanup afterwards to bring the volume down and the transformations work a lot better.
- Badges
- Science topic
- Similar topics
- Vectorization