Protocol Online logo
Top : Forum Archives: : Molecular Biology

Difficult PCR amplification of 4kb genomic DNA - (Mar/30/2001 )

I am trying to amplify a fragment (about 4Kb) form genomic DNA. Initially there was no amplification even from the positive control which used plasmid as template. After some troubleshooting, I found the 3' part of the fragment was GC rich (70% in 400bp). I add 5% DMSO to the reaction and got the product. However, the same condition failed for the PCR using genomic DNA as template. The template quality is good and I have tried different dilution. I also tried GC-Melt kit from Clonetech and Finnzyme, but both end up in failure. Anyone can give ne some recommendation?



It seems that you have tried every possible alternatives. The only thing I may suggest is designing some new primer pairs. Some times, one pair of primers doesn't work, other may work.

I once tried GC amplification kit from Clonetech to clone a GC rich region and found it really offered some advantages over regular PCR. You may also try different cycling program like two-step or three step.

Good luck!


I have successfully used this kit to introduce more than 20 mutations. It seems so far I have had to overcome every possible problem with this method - hopefully I can help. The first point you need to check is are your primers working? After running the pfu PCr and dpn I digest run 5 ul aliquot of the 50ul PCR on an agarose gel. If no band then your primers could be a problem. If you are sure that you primers have no inherent problem - hairpins, false priming site, contamination etc then you could try diluting the primers down. With this kit I have found that high primer concentrations lead to formation of sense/antisense duplex that traps all the polymerase. try serial diltuion of the primers. Sometimes you have to go to 0.2 micromolar final concentration of each primer inorder to get it to work.<p>If your PCR is working and you get transformants but they are all wild type then either the DpnI enzyme is bad or you are using way too much template (I also had this problem). Setup three prallel PCR reactions with 10 fold serial dilutions of the template. Follow the protocol and tranform the three into your bacteria. You wil see that the plate of the PCR mix with the greatest amount of template has the most colonies. Pick colonies from the plate that has the least colonies on it - this way you minimize the background and achieve close to 100% success rate.

Good Luck.


Hello, Is there anywhere I can find complete protocols for doing the type of genetic tests that forensic labs do for uniquely identifying individuals from blood samples and cheek cells? I know that it is usually RFLP analysis, but what are the regions of the human genome that these tests work on? Do companies sell the need primers? How do you extract the DNA from the cheek cells? Any help with details much appreciated.

Thank you.