Why you use very low annealing temperature 40 ? ..... what about 50 c ....... also think about how much template you use from product AB and CD for constructing AD .... do not forget that you did purification from gel at which amount of template is low (according to the expected size of the band) .... you should use 2.5 ul at least from each PCR product to be able to splice them together ....

While for making the products AB and CD may be you used 1 ul or even 0.5 ul template which is enough ..... but it is not enough for splicing the two products to each other.

My first question is how long fragments you should obtain? If I remember well my Phusion polymerase works quite slow and I had problems to obtain products longer than 1kb even in 30s. And second: have you checked presence AB and CD products after I PCR on gel? Ibrahim is right, you should first purify templetes to II PCR. 7 cycles is not enough for that. You can make dose dependent II PCR using increasing amount of templates. A third question: have you checked different buffer? I think there should be two kinds of Phusion buffer. Good luck!

Thanks a lot, i do purify my Dna fragments from gel AB and CD, and mix in gradient in the overlap reaction by 1.5 ul, 3 ul and 5ul. I use buffer with GC. My overlap product should be 500 bp. should i use annealing 50C in the first round and 10 cycles rather than 7 and increase the annealing in the second round to 60?

I guess that you are using a low temp for the first annealing stage because the overlap between the fragments is short. However, this should only be important for the very first cycle(s), where "fused" products are formed when the fragments prime each other. After that, the annealing conditions should be chosen to be optimal for the PCR primers (A and D) that you have added to the reaction which will amplify the full length product. So your basic approach is good. When I have done this, I have not even bothered to gel purify the initial products, if they look reasonably strong and clean on a gel. Just dilute them 1/1000 or 1/10000 into the second reaction. Also, if you do see any of the correct product on a gel after the first A/D amplification, you can do a "poke" PCR, where you just poke the band with a pipette tip and just touch that into a fresh PCR reaction mix. Do this a few times with separate reaction tubes and run them for various numbers of cycles. Hopefully, you will find that it only takes a low number of cycles, say 10 to 15, of this "re-amplification" to get a useful amount of product.

I just did a very similar overlap PCR. In my case, I synthesized AB and CD, purified those fragments, and combined them at the highest possible amounts in a new PCR. I used half of my purified fragments (20 ul of each). Polymerase was Pfu in order to avoid synthesis errors and I ran 35 cycles. Annealing temperature does not really better much in this setting. 55 degrees would be fine. I do not add any outer primers at all. Just mix the two fragments AB and CD and let the polymerase synthesize the rest. I had earlier tried to use outer primers in this protocol and it invariably lowered my yields.

Thanks a lot, i will try as Ralf says, i will run for 35 cycles without adding A and D primers, hoping that AB and CD become ligated and extended:)

But did you clone the product further? the AD, i will do TA cloning first then allelic exchange as a complete deletion of gene.

If you just mix the AB and CD products without adding A and D primers, then what you are doing is not PCR. It is just mutually primed synthesis. Since there will be no amplification, there is no point in running more than a few cycles--although the process may be inefficient, because at high concentration each segment will tend to reanneal with itself rather than at the small overlap shared with the other segment. If you run 35 cycles, you will very likely create illegitimate concatamers which will appear as a large smear on a gel analysis. Also, the most full length product that you can expect to get out is just the sum of what you put in.

Note that if primers A and D are added to the reaction, they will generate additional copies of the strands which can anneal to each other and extend to the full length product. And they can also prime PCR of the full length product.

In my experience with overlapping PCR, you have to do either what Ralf or David mentions above: either purify the products from the first step or dilute them significantly before proceeding with the second PCR rxn.

I think the important thing is to get rid of any remaining B and C primers, otherwise you will likely just make more AB and CD and not enough of the longer product. It shouldn't be necessary to mix the AB and CD products without the AD primers in the second step if you design your B and C primers to have enough overlap to anneal at the same temperature as A and D.

I too have the same problem but mine is a three fragment overlap can you please help me to standardize the procedure.

I have gel purified 2 end fragments of 100 bp each and have synthetically ordered the 79bp middle fragment

I want a complete product of 234 bp

I tried with a 35 C annealing for 19 cycles added end primers and 55C for 30 cycles

I am getting large smears and a 100 bp band

Please suggest how to proceed

I am using a Q5 polymerase enzyme for the reaction

I tried ralphs advice even then i am getting a smear type of band

 Hi Uthra, did you solve your problem, I met the same situation and got a very long smear type of band. I have no clue to solve this problem.

I would definitely gel excise and purify the PCR products obtained by using AB and CD primers. Then for the overlap PCR, I would use 1:1 ratio of both excised fragments (~75ng each). This always works for me.