mtDNA to nDNA Ratio gDNA - qPCR - (May/12/2013 )

I am trying to measure the ratio of mtDNA/nDNA from a set of cells. The problem is that when I run the final qPCR (with SybrGreen) I get wildly different results for the same cell lines (supposedly with the same ratios).

Briefly, I isolate total DNA using a genomic DNA isolation kit from sigma, quantify it using a spec and dilute it to 50 ng/uL and then dilute it again to 2.5 ng/uL and add 2 uL of this to my 25 uL qPCR reaction for a total of 5 ng in the reaction.

I optimized these primers using serial dilutions of a mixture of gDNA from all my cell lines (2 fold). I got great efficiency for my mtDNA gene (slope = -3.3228, R2 = 0.988) but for my nDNA I got poor efficiency (slope = -4.3011, R2 > 0.99). These primer sequences have been used and published many times by other labs, albeit typically using TaqMan technology.

Despite this, I continued and ran my samples and am getting pretty crazy variability between samples within the same cell line and same condition.

My questions are:

Could this variability within the same cell line and same condition be due to the poor efficiency of my nDNA primers? I've read that you shouldn't compare reactions that have large differences in efficiency? If so, how can I increase my efficiency or is there a way to compare reactions with different efficiencies by using a correction? I'm currently using the delta delta Ct method.

Should I be using more gDNA?
Can this be because of contamination? It seems like it shouldn't be since I get fine amplification with my mtDNA.

Since gDNA is much larger and more complex than mtDNA could this be part of the problem? Should I make changes to my running method? My current run method consists of a 10 min holding period at 95ºC, then 40 cycles of 60ºC for 1 min and 95ºC for 15s, and then a melting stage of 95ºC for 15s, and 60ºC for 1 min and 95ºC for 15s).

Thanks to everyone in advance for any help they can give!!!

I'd suggest cutting your gDNA with a frequent cutter that fails to cut within the amplified region to reduce viscosity. Accurate dilution of high MW DNA is almost impossible.