We have had a lot of experience with SOE and it's generally been very reliable in our hands. First, you do not need to generate single strands from each amplicon; it is absolutely fine to mix the purified double-stranded PCR product for the "overlap PCR" in your link. We have done this many, many times.

In your case, have you run your initial "extension PCR" products out on a gel to confirm that you have the correct 400 and 90 bp fragments? If there are contaminating bands I would recommend that you gel purify the bands; if the products are clean then column purification is sufficient. We quantitate the PCR products (Nanodrop is fine) and mix equimolar amounts for the "overlap PCR" without the end primers in a 25 µL volume. We use Phusion routinely without issues (despite what the protocol that you linked to says). Double-check the predicted Tm of the overlapping 35 bp; in our case we generally have a high enough Tm that we can use a 2-step PCR for the "overlap PCR" (i.e. 72˚C for both annealing/extension). The 60˚C annealing temp listed in that protocol is probably too low for a 35 bp overlap.

For the "purification PCR", we take 1/5 of the "overlap PCR" reaction (i.e. 5 µL) and use that to set up a 50 µL reaction with the addition of reaction buffer, end primers, more polymerase, dNTPs, etc. For the "purification PCR" we use an annealing temp calculated from the flanking primer Tm, not a fixed 72C temperature, and we cycle for 30 cycles. I usually column-purify the PCR product, digest, gel purify, and ligate. If you are not sure whether you have the correct product, though, it'd make sense to run a small amount on a gel to check.

Are you working with ds DNA? The complementarity is 100% with both strands present from each amplicon but only 10-30% with 2 different strands.Self hybridisation will overwhelmingly predominate. You need to generate single strands form each amplicon with single primers first (eg 10 cycles of single primer amplification) on opposite strands. Then mix the 2 reactions, allow to anneal and extend then amplify.

Ian Teo

I've not read about that in my source articles, but I'll try your suggestion and let you know if it worked for me. Thanks again!

Are you sure about the sequence of your fragments, it is really what do you expect? Conversely, you may check by sequencing.

However, I have done similar amplifications with less homology (around 20bp) for site-directed mutagenesis experiments on a previously cloned cDNA. I suggest to mix your external primers directly with the purified amplicons while preparing the PCR reaction. Put it in the termalcycler and set at least 35 cicles of amplification at the appropriate annealing temperature of your primers (the 72°C indicated in your protocol sounds quite high!). Apply an extension time of at least 1 min.

Let us know how it goes

We have had a lot of experience with SOE and it's generally been very reliable in our hands. First, you do not need to generate single strands from each amplicon; it is absolutely fine to mix the purified double-stranded PCR product for the "overlap PCR" in your link. We have done this many, many times.

In your case, have you run your initial "extension PCR" products out on a gel to confirm that you have the correct 400 and 90 bp fragments? If there are contaminating bands I would recommend that you gel purify the bands; if the products are clean then column purification is sufficient. We quantitate the PCR products (Nanodrop is fine) and mix equimolar amounts for the "overlap PCR" without the end primers in a 25 µL volume. We use Phusion routinely without issues (despite what the protocol that you linked to says). Double-check the predicted Tm of the overlapping 35 bp; in our case we generally have a high enough Tm that we can use a 2-step PCR for the "overlap PCR" (i.e. 72˚C for both annealing/extension). The 60˚C annealing temp listed in that protocol is probably too low for a 35 bp overlap.

For the "purification PCR", we take 1/5 of the "overlap PCR" reaction (i.e. 5 µL) and use that to set up a 50 µL reaction with the addition of reaction buffer, end primers, more polymerase, dNTPs, etc. For the "purification PCR" we use an annealing temp calculated from the flanking primer Tm, not a fixed 72C temperature, and we cycle for 30 cycles. I usually column-purify the PCR product, digest, gel purify, and ligate. If you are not sure whether you have the correct product, though, it'd make sense to run a small amount on a gel to check.

forgot to mention that for the "overlap PCR" we use 100 ng of the longest fragment and equimolar amounts of the smaller fragments (so in your case 100 ng of the 400 bp fragment and 22.5 ng of the 90 bp fragment).

One more thing--the 90 bp fragment is short enough that you can simply use one long oligo to fuse the sequence of the 90 bp fragment with the 400 bp fragment in a single PCR reaction. It's relatively inexpensive to order long oligos (e.g. from IDT) and if you're still having difficulty with the SOE PCR approach to generating your 490 bp fragment I'd go this route instead.

I agree whith Andrew. One more thing you might try and which worked for me in the past is touch-down approach. The 35 bp overlap should be the first thing which starts annealing, later on you will have enough of correct template.

I would not disagree that you can do overlap PCR without having to generate single strands. As I understand Matias's query, however, he has tried the "conventional" approach and for whatever reason it has not worked. Like Matias , I too have had problems in the past when sometimes it works and other times it did not. I sumised that the 2 different partially complementary strands were not hybridising sufficiently well / quickly to generate an amplifiable product. The rate of zippering and self annealing and complementary annealing will be dependant upon the structure of the strands and the overlap regions.It can take time to get the right combination of annealing conditions and PCR conditions to get the system to work well. By generating single stranded products in the first instance, you can circumvent this problem,- if it is an issue. The potential for reannealing does not occur as the only hybridisation that can happen is in the region of overlap.

Ian Teo

I'm now executing Andrew's proposal because it's quicker and I haven't tried those exact conditions. However I don't reject Ian's suggestion, but it will take more time because of the extra steps.

Andrew, I'm running the "overlap PCR" as you said, but using Platinum PFX polymerase. I did this because my overlapping region Tm's is 66C and the PFX extension Tº is 68C, while Phusion has a Tº of 72C.

My new doubt is: if I want to change the polymerase in the "purification PCR" (just for testing), should I purify the "overlap PCR" product by gel to get rid of the "old" polymerase buffer and Mg+ form (MgCl, MgSO)? Or should I just take the product volume needed and add the reaction components to it?

Hm--I wouldn't recommend changing polymerases unless you have a good reason to do so. If you do, I'd guess that you should clean up the "overlap PCR" product; wouldn't gel purify though because I'd expect the correct full length fragment to be low in abundance and difficult to see. A column purification should be fine. Again, I probably wouldn't do this myself without a very good reason to do so.

You could also try PCR-ligation-PCR as an alternative to SOE PCR. We developed it back in the 90's as a universal method for joining PCR fragments and generating mutations. See BioTechniques 18(5)746-50, or my ResearchGate profile.

You can add all the components directly to the PCR reaction without doing some cycling without primers. I would try a gradient PCR. Also, Takara PrimeStar and Phusion are very potent enzymes that should take care of your issue.

Re Stuarts point, I have done a lot of blunt end ligation PCRs. They work fine using a promega rapid ligation kit with extended ligation times - dilute out the ligation and PCR with the external primers. Make sure you PCR products are blunt ended (use a proofreading Taq) and kinased

Ian.

Hi! I've tried Andrew's protocol without success, but I do achieve my objective through a common PCR protocol using GoTaq polymerase (Promega) with long annealing and extension times (3 min) in each cycle. As I kept close to recommended reaction conditions I didn't reach any result, so my director an I decided to try longer times just in case, and we get the desired amplification product.

I wouldn't say the protocol was super efficient, but I got enough material to continue my research.

When trying to reproduce the protocol times changing the polymerase for Phusion, it didn't work.

So I'm going to proceed using my GoTaq amplicon, but I have to make other recombinant product, so I'll keep trying conditions and suggestions.

I'm kind of new in genetic engineering field, so I wasn't aware of that "gradient PCR" technique. Reading about it I realized I could've saved much time, so I'll give it a try next time.

Thanks again for all your answers!

Have a nice day!

i think gibson assembly strategy of NEB can be tried in this case, using NEBuilder software primers can be designed for both 90 and 400bp fragments, after amplifying both the fragments separately, they can be fused in a single reaction.

Hello Matias, it's been two years since the last response. I'm having a similar problem with longer bands (400 and 2200), have you refined the protocol? I haven't had any succes, I have TAQ and Woosion polymerase only. I havent tried a longer annealing time, whats the meaning of it?