Check out the picture below. This is a series dilution of genomic DNA. The top bands are primers for GAPDH gene, the middle band is actin, and the bottom is my gene of interest (HCN1). Why on earth does it constantly produce a band? I have done negative controls for the PCR with no DNA, and I get nothing. It seems like even if there is a molecule of DNA in the tube, I get this band?
It seems like it even goes up after getting to a low enough amount of starting DNA ?

You have mystery on your hands. There is definitely something weird going on from lane 4 to 5. My first thought is, if this is human DNA, that you have contamination of skin cells, but then you would also see the top two bands amplify past lane 4. So, the other possibility is that you have HCN1 DNA contamination from a vector used it your lab. Not sure why "no DNA" gives you no bands.
Has this happened more than once?
dan

Oh yeah, this has happened more than once... it's an ongoing plague! It's actually rat brain tissue, but I am pretty sure it's not contamination. The really weird thing is that it's not just a primer problem, I pick different primers in the same general area of the gene and I see the same weirdness?

Is the gene cloned in your lab? If so, that could be the source of the band. You just need to figure out what solution or pipet is contaminated . If the gene is in one particular vector, you could try using vector specific primers to look for contamination.

No, it's not cloned. Plus I have used brand spanking new primers, polymerase, and water.

good mystery!

interesting....
did you use the exact same aliquot of water (or buffer) for your negative control and your DNA dilution?

Yeah, actually I purposely used water that wasn't as 'clean' in some reactions in the hope that I would see the product in the non dna runs . By less clean, I just mean from the distilled tap. Usually I use PCR grade water for everything. Anyhow, no product in the non DNA samples?

Maybe if we figure this out we'll get a Nobel prize or something!

The appearance of the very large band in the righthand lanes is also interesting.
Do you also get this result when GAPDH and Actin are not co-amplified? It could be a band that just happens to co-migrate with the HCN1 product. It could be that the primers (to GAPDH, Actin, HCN1, or a combination) amplify a high copy number (but less complementary) sequence when the genomic DNA concentration is lowered. Try higher hybridization temps, fewer cycles, and lower primer concentrations.

That's true, I meant to point that out as well, this does occur when I don't multiplex (just HCN1 primers alone) and that larger band (which oddly is almost twice the size) also appears.

I'm trying different temperatures and enhancers (DMSO, betaine...) just because I'm curious.