Xiao,

Describe the melting curve data.

Thanks Michael for responding to my question. The attached is the melting curves. All samples AND the negative control have pretty much the same curve pattern, all with a single peak around 88.56 C.

My interpretation is that the either the water or the primers are contaminated with target DNA. So that negative controls behave a sample with exceedingly low concentration of target.

How do you think?

I'm more concerned with an efficiency of 200%.

Do you mean an efficiency of 2?

Efficiency of 2 = signal doubles with every cycle, and this is considered to be a reaction proceeding with 100% efficiency.

Efficiency of 1.9 = signal only increases by a factor of 1.9 each cycle, thus efficiency of 90%.

Assuming you simply mean your reaction is ~100% efficient, then honestly: I wouldn't worry about your contamination (and it most likely IS contamination) as long as your sample Cq values can never be mistaken for negative control Cq values. You list "17-28" for your sample Cqs, which I hope is a typo (did you mean 17-18?). If it's genuinely 17-28, then yes: you have problems, because a signal of 28 vs a 'negative control' of 29 are easily confused.

If in contrast your sample Cqs are robustly 17-18, while no-template amplification remains 29-30, then this is a difference of ~1000 fold, well below the noise threshold of your assay.

Hi John Hildyard, the slop of the standard curve is -1.99 yielding efficiency of 216% suggesting the polymerase is inhibited according to the user's manual. One possible explanation is that I used PowerUp master mixture which uses a hotstart polymerase requiring initial 2 min at 95 to activate the enzyme; however, I mistakenly used the regular 30 Sec at 95. This is consistent with partial inhibition of the polymerase by the hotstart antibody. This is my best guess

.

I had 30 samples, and they varied quite bit in Ct values. Shown in the attached figure, the red dots are standards; green and blue are samples.

Yep, ok: that really is a 200% efficiency. This is...not good. General consensus is that this is artefactual: you're not actually tripling your transcript level with each cycle, it's simply that your high [template] concentrations are inhibitory somehow, so skew the entire curve. It's not the polymerase, it's your samples.

If you look at your standard curve datapoints, those three in the middle (up until ~Cq 30) look to be showing a curve much closer to -3.32: the more concentrated standards might reasonably be expected to be non-linear, and the very low concentrations are never going to give you a Cq above 30, because even your no-template control gives a Cq of 30.

How are you preparing your samples? And how are you preparing your standard curve?

Hi John, you have sharp eyes! I just re-calculated the standard curve using the three middle data points and found the efficiency became 103%. Using this new standard, I computed the absolute [target DNA] of my samples and found the numbers made much more sense than before.

My samples were prepared as follows: dense E. coli LB cultures were centrifuged; the pellets were resuspended in sterile water (prepared in the lab, not commercial one) and boiled at 95C for 5min; the suspension was then centrifuge to pellet the debris. One ul supernatant was added directly into a 10-ul qPCR reaction mixture.

My standard was prepared simply by a serial dilution using product from a plasmid prep as the starting undiluted material. Sterile water from the same bottle for preparing sample was used here as diluent. Note that the standard tubes sat at 4C for a week before qPCR experiment, which might be a problem.

Okie doke: one thing to note is that "sterile water" may well have residual bacterial DNA. After all, when you boil your bacterial pellets for 5 mins, the resultant mix is certainly sterile (everything is dead), but assuredly contains DNA.

Your preps are also pretty crude: you're not counting cells, I gather, plus 1ul of raw lysate will contain a whole load of crap (salts, residual proteins, etc) that might interfere with qPCR, especially since this lysate represents a 10th of your entire PCR reaction.

Have you considered bacterial genomic DNA isolation kits? Failing that, what about diluting down your lysates substantially?

Remember: diluting by a factor of 10 should only increase your Cq values by 3.3.

Hi John,

While samples were taken from cell cultures for boiling, OD600 was also taken of the same culture so that at the end I can calculate copies of target DNA molecules per cell by dividing the qPCR-derived copy numbers by OD600*a conversion factor. At least, that is the plan.

Because of the numerous (30) cell cultures, I prefer to avoid any extra complex procedures of DNA extraction. In the next experiment, I will definitely 1) use commercial water and 2) dilute the samples 100 fold.

Thanks a lot for your responses. They are very helpful!