Try to decrease the overlapping region sometimes works

Assemblies with many pieces work more reliable in yeast, if you can afford this more elaborate method time wise.

Hi Peter

I have bought the assembly myself but have not used it yet. However the rep from NEB told that they always suggest to go increase the overhangs and that normally solves the problem. If that does not help the rep said that the people at NEB are happy to help with finding out the problems. I was in contact with techsupport and they replied fast and with helpful tips.

Good luck

Barbara

You could also use the flanking primers and PCR the whole thing up using a HF DNA polymerase like phusion or q5.. Also try to transform your product into low recombination event cells like SURE cells from stratagene to avoid rearrangement in your E. coli if possible (homology/secondary structure might cause problems .. Alternatively Gibson up part 2345 and then throw in part 6 and 1... Let us know how it goes..

I never useI never used that technique, so im not sure if you can do this, but could you try do it with a little bit of DMSO, that will help you if there is some problem with secondary structure.

Regards

I agree with Karl.While you sort out the issue you can PCR the whole mix (Fusion PCR). The overlapping homology between fragments should be OK. Phusion works and also Platinum Pfx. Don’t know about q5

Many thanks to all for your valuable contributions.

Karl: Happy to try your suggestion to Gibson up the middle bits first and then the two flaking fragments. Do you know of any kinetic/thermodynamic issues to assemble two bigger fragments that would render your approach more feasible?

Gonzalo: We use DMSO for PCR, but haven't done so for Gibson assembly. Thanks for suggestion.

Abdelhalim: So you are suggesting an overlap extension PCR for the two big fragments rather than another Gibson round? We are using Accuprime Pfx for PCR.

Give us a week or two and I will come back to you.

I made some overlap PCR of 3.5kb using Hi-Expand fidelity of Roche, for 2,4 kb probably will be fine.

Good luck

I am suggesting a fusion PCR with the 6 fragments in one go. I done it with up to 5 fragments to generate cassettes to be integrated into the genome and up to 5 Kb in lenght. In my opinion there shouldn't be any problem with 6. I always had around 40 bases homology between each two fragments.

@peter: I've been doing Gibson with multiple +5kb fragments and it did work but the efficiency went down (~10 col, whereas with regular 3 piece assembly I would get ~200 or more depending on the marker and cells used). I suggest you try Abdelhalim approach, just put all purified fragments into a single reaction and use the flanking primers.. make sure you use a low cycle number (~18-23) in your first PCRs to keep potential mutations at bay.. just gel purify and EtOH concentrate a 100ul reaction and you should be good to go..

I completely agree Karl and Abdelhalim. DMS0 (as suggested by gonzalo could be useful too).

Coming back to the initial feed and our experience with Gibson assembly. After we could not Gibson up a gene of interest (with any of the suggested methods), we designed new gene strings (G blocks) with a 20 bp overlap. This time it worked on the first go: synthesis error? GC content? We remain puzzled. We are now checking for GC content of the overlap (20-40 bp is a recommendation) and try to keep it below 50% if possible. PCR amplification of Gibson products seems reasonable and plausible to 'dilute out' by-products before cloning. We have good experience with overlap extension PCR as suggested by Abdelhalim.

I am very interested in yeast assembly as suggested by Tobias. Any suggestions, literature and applications for yeast assembly? I will come back to this with a new question, but would be grateful for some input from you.

Hi Peter,

Did you try the Gibson assembly with DMSO as suggested by gonzalo. Now I am having a failed Gibson reaction. One of my fragment is >10 kb while the rest 2 fragments are within 2kb. I had to go with 10 kb Gibson fragment as the original plasmid (Vector for amplification of fragments) sequence is GC rich(75%) and has many tandem repeats which are too hard to amplify as individual fragments. Previously I succeeded with Gibson fragments ~ 10 kb. I am planning to try  DMSO in Gibson reaction as this could avoid secondary structure formation of my fragment. 

I would be interested as well if there are any other suggestions for the big fragment Gibson reactions. Thanks

Hi Ramakanth, Did you find any improvement after trying DMSO in the Gibson reaction