I had a similar problem (low efficiency, very low... no choice in primer binding site) : I simply increase the annealing/elongation step from 30 sec to 1 min (the product size was ~230bp) : efficiency has increase from 75-80% (>_

Thanks Thomas,

I decreased the annealing temperature and increased the primer concentration.

It is getting better and some primers reached an acceptable level now. However, I will have to check if they are still specific enough with the new low Tm...

What kind of qPCR are you performing ? unspecific labelling (Sybr) ? Probe labelling (taqman or other) ? Multiplex ?

And what about the template for the serial dilutions ? Is this plasmid or something else ? If you are using guts genomic DNA (spiked with your favorite bacteria), you must know there is often (always ?) qPCR inhibitors even if you purify these DNA with expensive silica columns kits (Soil blabla, Stool stuff and so on). You should dilute your samples (10-fold or even 20-fold).

When I try new primers, I perform gradient PCR on a endpoint cycler (with and without template) then I test the conditions on other bacteria (always in endpoint cycler) and if it results in a good specificity, I perform qPCR standard curves with plasmid/pure culture DNA (100 nM for primers, 30 sec for elongation/annealing step with the annealing temperature as defined previously). If it doesn't result in a 95-100% efficiency curve, I extend elongation step (45 sec, 1 min). Thereafter, I try to maximise qPCR reaction with primers/probe concentrations.

Edit : Maybe your qPCR system (cycler and buffer) needs 3 steps qPCR (denaturation and annealing AND elongation).

this is not only DNA purity problem, but non-universality of universal primers

diluting complex sample one changes  distribution of different species, leading toward non-linear Ct increase