Protocol Online logo
Top : Forum Archives: : Molecular Biology

to purify these PCR products or not? - preparation for restriction digest (Dec/05/2005 )

I am using PCR to incorporate some restriction sites to flank my gene of interest, with the goal of using those sites to eventually digest and ligate into a given expression vector. A colleage of mine suggested that I do NOT try to purify the resulting PCR products (with QIAquick kit), but rather to proceed straight with the restriction digest and then ligation. His rationale was that after 18 rounds of such a PCR reaction, there is not enough product to purify effectively and I would probably lose most of my fragment of interest in the column. Is this true? Without purifying the PCR fragments, how do I quantify how much I have? Or is there a rule of thumb that people generally use? I was also told that after 18 rounds of my PCR reaction I may or may not be able to visualize the products on a gel. Is that a fair statement? I would appreciate any help or comments on these issues. Thank you very much.


to be honnest i would go up to 25cycles.
For efficient digestion you can se at neb websites the enzymes that are capable to cut in a primer extension mix.
see here :

additional info may be founded here :

for instance if you want to remove pcr buffer i would strongly recommend you to add 1/10 vol NaAc, and 1vol of isopropanol. Eventually you can use a carrier as glycogen but not ssDNA or tRNA.
spin 30' 4° and wash in ethanol 70%


After trying to do a direct PCR->restrictoin digest unseccesfully I would recommend that you either go the few extra cycles (As suggested by our illustrious Fred 33) or perhaps go through a T-vector intermediate. In the long run it can save a heck of a lot of time, especially if you do not have large amounts of PCR product. I routinely use the pGEM-T-EASY system from promega and have had only good experiences with it.


Unless you eliminate the PCR enzyme, the enzyme will fill in a restriction site cut by your restriction enzyme. This will happen for the 5' overhang created by most REs, since it looks exactly like a primed DNA fragment. 3' overhang fragments created by pstI or similar enzyme will not behave this way, but this is rare.


To beat a dead horse, doing extra cycles will pay off dividends. By doing 18 cycles, you will have 2^18=2.6x10^5 copies. By doing 7 more cycles, 2^25=3.35x10^7 copies or 129 times more DNA than 18 cycles.

With more cycles, you'll be able to visualize (and if you're real good, eyeball the concentration based on band intensity) your pcr on a gel and if there are multiple bands or the template, you can gel purify the right one and quantitate it. You're just setting yourself up for a lot of headache by cutting corners. "Maybe digest didn't work because of the PCR didn't work, the enzymes, template, voodoo, etc"

But if it's an "established" procedure and people do it in their sleep, then I suppose you could go with the flow but when things don't work, you gotta go back and do things by the book until they work.


vsoy: your calculations are not entirely accurate. By the thirtieth cycle, PCR synthesis has already begun to plateau, thus yielding decreased DNA with each additional cycle. If you ever perform real-time PCR, you will realize this after observing your data.

As for the original question, I would run 35 cycles and purify the product/cut or clone into a TOPO vector then cut. You will only lose 5% of your product (see Qiagen manual) through purification of PCR product which will not effect the outcome of the digest/ligation. Make sure to leave four to six bases on the ends of your primers so that the restriction enzymes cuts effectively.

I like cloning into TOPO because you know that your restriction enzyme is cutting when you see two separate bands. This is merely personal preference. People cut PCR product with high success rates all the time.