Mystery in my PCR - (Jul/09/2013 )

Hi all,
Please help me figure out what happened with my PCR.
I used the same master mix, same DNA template, same condition of thermocycles for 4 reaction tubes (10ul/tubes). However, I just only got the band in 1 tube when I checked on agarose gel. I repeated the experiment, but got nothing. I tried one more with 2 tubes and new aliquoted primers got from stocks, and I got products in both. So, I tried again with 4 tubes (because I need high quantity of PCR products), and got nothing...
I dont know what was the problem. I repeated several times, but the PCR just worked in some times and some reaction tubes...
I am so thankful if you give me some suggestion...
Best.

PCR is very sensitive, and will amplify very small differences. There may be a very low amount of template present for your primers in some samples. More likely, the primers are either mismatched or the cycling temperatures are far from what would normally be required, and once, rarely, things amplify. Once they start in a tube, they will go all the way. I'd suggest verifying your primer design (ideally by sequencing your template), and using gradient PCR to check primer activity. Depending on your template GC content, you may need to add betaine or DMSO (high GC), or to lower the extension temperature (very low GC). 10 ul samples are quite small, and a heated lid will be required to keep the sample at the bottom of the tube.

Thanks Phage434.
At first, I used gradient PCR to optimize the annealing temperature (from 60-68oC) and used Pfu DNA Pol. I chose 63oC to use as the annealing temp for my experiment, others also showed the bands but of course not specific binding. Thus, I don't think the annealing temperature was the problem.
Next, the primers did work because whenever I tested with gradient PCR or test with 1~2 tubes, I got the bands. I wondered about the secondary structure of Primer or primer dimer, so I used the tool in idtdna.com the analyze my primer, and it showed 2 hairpin structure formed at 47 and 48oC.
I still worry about the primer, maybe it work unstably... How can I check the primer if they can work well or not?
Thank you again for suggestions.

The major problem with primer dimers, either hetero or homo, is the extension of primers at the 3' end following binding of the 3' end to a 5' strand. This is supposed to happen if it is your template that is bound, but if it is either another primer, or a portion of the same primer (a hairpin) then the extended primer will never bind to your template. So hairpins or matches at the 3' end, which leave a 5' overhang on the other strand are very very bad. The tools, unfortunately, do not distiguish this case from the benign cases, such as binding in the middle, or binding with not 5' overhang.

Thanks Phage434, I tried once again with two different genomic DNA as templates in the same concentration. I could get the product from 1 template but not from the other. Is there any chance of the template problem? They are all genomic DNA extracted from human white blood cells.

PCR inhibitors in samples can definitely be a problem. Phenol is a particular problem. Make sure the volume of template you are adding is very small relative to other components of the PCR reaction, which will minimize sample contamination effects. Is it possible one of your samples has mutations in the primer binding region?