Smearing in "Overlap Extension PCR" - (May/02/2006 )

Hi,

I have been trying splicing by overlap extension PCR. I have obtained three products of approx 900 bp, 100 bp and 500bp which have to be joined to give a spliced product of 1500bp. Getting the three products is not a problem but joining them is.

I am using a two step process. In the first step the three products are mixed (in equimolar amount) and a PCR is run for ten cycles without any oligos. The overlapping region of the 100 bp has a Tm of 51 and 53 degrees with the 900 and 500bp. Hence i am using 45 degrees as the annealing temp.
In the second step, i am taking 2 ul of the PCR product from the first step and doing a PCR for 30 cycles with the external oligos. The PCR protocol used is the same, which gives the amplification of the full length gene without splicing. However i am getting a smear throughout and faint amplification at aorund 100 and 500bp.

The internal oligos are of of ~90bp length, PAGE purified and 20 bp overlapping.
Please help?

I have tried this type of PCR and I must tell you that putting all 3 together is a trouble. Of course, you designed internal primer but I had to use the forward and reverse primers of the two products.

What I would recommend is to set up 2 PCR reactions, I presume that you are looking for 900-100-500bp sequence. So first setup a PCR using forward primer of 900 and reverse primer of 100. You are right there may be an issue with annealing temperature as Tm of the fwd primer and rev primer may be different in this case as they are for two different DNA regions, so you may have to do gradient PCR to confirm the temperature. Make sure you increase your extension period as well to about 2 minutes.
So now you will have 900-100 product.

Setup another PCR after preferably gel purification of 900-100 product with gel purified 500 product exactly same way as the first step.

In theory, you are doing right of running 10 cycles without primers but I have noticed that it doesn't make any difference even if you don't run these cycles, so you may want to skip it.
By this method, I managed to stick together 2.5kb and 1.9kb successfully with load of product.