Hi Rob,

If you want to assemble DNA fragments into one linear fragment, you can achieve this by doing a simple overlapping PCR. Make sure that your two fragments have compatible overhangs, and mix them together with your usual PCR mix, but without any primers. As the strands separate then hybridize, each fragment can serve as a primer for the other as illustrated in the attachment. Calculate the elongation time so that complete elongation can occur. Ideally, make sure you have the same molarity for both fragments in the initial mix.

Might work, but it's less than ideal: 

(1) You will carry some percentage of incorrect assemblies (that necessarily occur) into your PCR.  I would prefer to get rid of those by cloning right after the assembly.  With some bad luck, a short incorrect product might amplify much more efficiently than your desired construct.  Then you'll lose your desired construct at the PCR step and will never know it was even there.

(2) What could possibly be the downside of assembling a circular product (with stuffer sequence if needed), purifying the correct form by cloning, and then linearizing it by PCR?  Why do you want to do what you want to do?  

I'm actually a little less than convinced that the exo is actually going to be dead anyway...apparently it is pretty stable of an enzyme...

Hi Rob,

If you want to assemble DNA fragments into one linear fragment, you can achieve this by doing a simple overlapping PCR. Make sure that your two fragments have compatible overhangs, and mix them together with your usual PCR mix, but without any primers. As the strands separate then hybridize, each fragment can serve as a primer for the other as illustrated in the attachment. Calculate the elongation time so that complete elongation can occur. Ideally, make sure you have the same molarity for both fragments in the initial mix.

Hi Robert,

It worked for me pretty well. I wanted to join two fragments without doing the whole cloning by Gibson to then reshuffle stuff on the plasmid easily. And I also wanted to try how Gibson works as it was the first time when I used this method.

T5 exonulsease is not very stable at 50C and that's why it doesn't keep on chewing your DNA (http://www.srcf.ucam.org/~wac26/gibson/). If you're not sure, you can incubate the tubes at 90C for 2 min after the reaction. I got nice PCR products that I used for my clonings and sequencing confirmed everything was all right.

Best wishes,

Michał

Thanks Daniel for your genious idea. it works very well for us.

Best

I thought that I should follow up since this question receives a lot of views. While NEB claims that linear assembly is possible, it only works if the primers used to PCR amplify the linear fragment are very far from the end (>100 bp). This is because the exo destroys priming sites. In practice, I've had a lot of luck creating circular Gibson assemblies into a cloning vector or even a closed circle of the fragment.

Dear all,

I have a similar work. I have 3 fragments which I wanted to assemble into pUC19 vector. I felt the overlapping region between the first fragment and the vector is less than the optimum and that's why the assembly did not occur in the hifi dna assembly kit.

Does Daniel's idea of joining the 2 fragments without primer works for 3 fragments also? Now, I want to assemble the linear fragments before I put it into the vector. Does anyone have any input? Please help.

Hi all,

I enjoyed reading your discussion. I`m wondering if a verdict has been reached between linear Gibson assembly and overlap extension PCR?

I would like to assemble three PCR products. Is this feasible in a single overlap extension reaction?

I like Robert`s comment about the destroyed primer sites in Gibson assembly. Could I just heat inactivate the exonuclease, add back some polymerase and dNTPs and perform some PCR cycles to fix this before I proceed with the PCR of the assembled fragment?

Michael,

Since it is a 5' exonuclease, you would need to add in primers as well... It should work but I would use the gibson assembly to close the circle instead.

Hi Rob,

you`re right. It`s a 5' exonuclease. I guess, I get now what you`re saying: The first round of PCR will still work, but as the template is chewed up, the primers won`t bind in the second round, right?

In this case I would really need to amplify larger pieces for the PCR fragments and then place the primers further inside to amplify the assembly as you suggested. Would you say about 100 bp is enough?

I don`t really want to close the circle as this adds another day for bacteria trafo and MiniPrep..

At least 100 but closing the circle doesn't have to mean with a plasmid. A closed circle is protected from the exo and perfect for PCR.

That`s really smart. I`ll try it this way.

Thanks!