PCR problem: non-specific "fragments"

[Aqueron]

Ok, so here is the thing.
We did a PCR to detect a gene on a couple of samples. So the gel was like that:

c+ s1 s2 c-



_

But we KNOW that the gene was supposed to be in sample 1 and 2. So we took these amplicons and did another PCR (using them as template intead of the DNA), to amplify what supposedly was amplified in the first PCR but didn't show in the gel.

RESULT:

c+ s1 s2 c-
|| || ||
|| || ||
|| || ||
|| || ||

No band appeard, but a large trace. Like if the DNA were totally degraded.

Any thoughts on what could have happened? o.O


PS: The PCR conditions should be ok, as the c+ worked at the first time. I don't know if that's possible (I'm a graduate student from Brazil, so I'm no expert), but what if the PCR mix was contaminated with something? Like an endonuclease, so it degraded all amplified DNA?

[JohnDakota]

for your 2nd round of PCR did you dilute your templates at all? Usually you have to dilute them anywhere from 100-1000x.

Also are you using any nested primers for your 2nd round of pcr?

[Aqueron]

actually, I'm doing the exact same I did on the first one :S same primers and everything.
I heard that some of my colleagues did that and it actually worked :S

[jonoave]

If the large amount of DNA is smearing starting from the well towards 1/2-3/4 of the gel, it shows non-specific amplication due to high concentration of template. Dilute your templates by around 50-100x (more if necessary) before running the subsquent amplification.

You should probably increase the dilution rate for your positive control, which has already shown a band in the previous gel. Your samples can probably be diluted slighly less as there wasn't any band previously (which doesn't mean there isn't any DNA, just not enough to be visibly detected as a band).