Assembly PCR can be used to assemble two heterologous pieces of DNA into one piece by simply designing primers a little differently and then doing an extra PCR step. This is essentially just for ease of cloning. Instead of trying to PCR two separate pieces and then assemble them by endonuclease digestion and ligation, it can be easier simply to PCR the first piece w/ a reverse primer that overlaps with the forward primer of the soon-to-be adjacent piece, and then use the product of the first PCR reactions as a template for the assembly reaction. If the overlap of the reverse primer for the 5' piece and the forward primer for the 3' piece overlap by ~20bp, the product of the first PCR reactions should anneal in the overlapping region and create a full length product. Using the Fw primer for the first piece and the Rev primer for the second piece in the assembly reaction then amplifies the desired full-length product.
Use only if: You want to assembly in series two pieces of DNA from PCR product (or you have one piece you want to cut out and assemble in series w/ a PCR product; in this case it's easier just to do the PCR for the piece you already have. Believe me, the assembly reaction product is well worth the cost of the extra primers). 1) Design the reverse primer for the DNA that will be 5' primer w/ significant overlap w/ the forward primer for the 3' piece. Essentially, as long as one of the primers has ~20bp overlap w/ the reverse complement of the other primer, the products should anneal in the assembly reaction. 2) Do PCR as normal for th 5' and the 3' pieces using the longer primers. 3) Check on a gel to make sure you got product from the first PCR reaction. Some people like to cut out the product band and use the purified products as template for the next reaction. I just use the PCR product from the first reactions; the logic is that the undesired primers will be in such low concentrations that the intended reaction will be highly favored. 4) Set up the assembly reaction like a regular PCR, except as template use equal amounts of product from the first reactions (I use 45µl of Invitrogen Taq HIFI supermix, 2µl of 5µM primer each, and 0.5µl of each pcr product as template). Cycle like you did for the first reactions. 5) Run the product on a gel. If the reaction worked, you should see a band the size of the sum of your two templates. 6) Purify the product, cut w/ desired endonucleases, and clone away! The quality of the product from this reaction is usually very good and I can get up to >100 transformants. Felix Moserfelix moser