You should measure reference gene for Z in both systems.

You have to either have a standard curve with known concentrations OR measure the expression of a housekeeping gene in both bacteria. Otherwise, you cannot compare the expression between the two types of bacteria.

If I'm understanding your results, you did include a standard curve. But your efficiencies are too low to be accurate. 95-105% is the acceptable range or you cannot publish these data.

I agree with Katie A Burnette , you should either measure absolute gene expression or relative gene expression against a reference gene such as 16S (in bacteria). using the reference gene method, you would then use the 2^-(delta delta Ct) method. See below:Article Analysis of Relative Gene Expression Data using Real-Time Qu...

Briefly, you would need to determine the Cycle time required to reach the threshold value (Ct) for each sample, then normalize to Ct of 16S of that sample, then normalize to either Group A or Group B (depending on which is your control). you can then further express results as a fold change relative to your control group.

Hi

If ur using abi machine vi7 and above upgrades,they give u the column to put efficiency and correct ct value automatically ,otherwise u have to use ct correction manually.still if ur efficincy falling 95 to 100 go ahead and use ddct.

I suggest u can try different concentration of primers say 1,5,10 uM to do std curve and use one that gives u good efficiency in range of 15-35 ct /cq.then directly u can go ahead for the ddct

If you compare gene expression by absolute quantification, you have to estimate it in relation to the number of bacterial cells, what will require separate PCR for bacterial DNA quantitation. This method produces also additional statistical bias. Use of relative comparison by reference gene expression is more efficient.

Thanks for all your suggestion. I find that using the relative quantification to compare genes expression is easier than absolute quantification.

Dear Jack Chen , I am going to use the Pfaffl method in order to compare genes expression in my experiment. In my opinion, The good point of this method is that It does not require the similarity of efficiency between the target gene and reference (While the Delta Delta Ct method requires the efficiency of both target gene and reference gene is 100%. I).

Dear Katie A Burnette , Could you send the paper, which shows the requirement of real-time PCR's efficiency for publication?

Here, the paper for Pfaffl method:

Article A new mathematical model for relative quantification in real...

Here's the paper with the MIQE standards:

http://www.rdml.org/clinchem.2008.112797v1.pdf

Thank you very much Katie A Burnette

When you say bacteria A and B, what do you mean? Are they different species of bacteria or are they just two samples of the same bacteria that you have grown under different conditions? If the latter (same bacteria), the problem is a bit less complicated. All you really need to do is add the same amount of RNA to the reverse transcription reactions and then do the qPCR. If you think about it, investigators have always done pretty much exactly this when they measure enzyme activity. They just do the assay and then they measure the total protein in the sample and then they correct the assay result for any difference in the protein concentration--so enzyme activity is usually expressed as activity/mg total protein. In the qPCR case, you can measure the RNA first and then just add the same amount of RNA to each qPCR reaction--and, of course, perform several biological replicates by repeating the whole experiment a few times. Of course, you might have some reason to think that the RNA content per cell is radically different in the two samples/growth conditions. In that case, you need to think about how you can more directly measure differences in gene expression per cell. So you would need to extract RNA from the same number of each type of cell (measured by calibrated A600 for instance) and then do the qPCR assay on that RNA. If you do it that way, then you would need some kind of spike-in control to make sure that the efficiency of RNA extraction is not different between the two samples (it is not clear that 16S rRNA would be a reliable reference in such a situation.) This more complicated approach of measuring per cell is also what you would need if the bacteria are actually different species because there is really no reliably suitable "reference" gene shared between two species.

Thank you very much David F. Barker . I have done several experiments as your suggestions above.

In my experiment, A is group A streptococcus and B is group B streptococcus.

I used the same amount of bacteria cell at exponential phase (measured by calibrated A600) for RNA extraction and then Treated RNA by TURBO DNA-free™ Kit (Thermo Fisher). After that, I used the same amount of RNA for cDNA synthesis (ReverTra Ace -α-® kit, TOYOBO). Finally, I did qPCR with cDNA samples (Quantitect SYBR green kit). I used plasmids containing the target sequence as the standard for the qPCR. Do you think my strategy for absolute quantification real-time PCR is ok or not?

Now my problem is that when I try qPCR with standards I get a good result (efficiency ranging from 98% to 100%, R2 >0.98), but for the cDNA samples, the efficiency is about 80%. Do you know what reason are and How to improve the efficiency of my qPCR from samples?

An Nguyen

Using a plasmid DNA clone as a reference is really not sufficient for "absolute" quantitation. There are too many differences between the processes of reverse transcription (which has unknown efficiency and generates a linear RNA/DNA hybrid as template for PCR) vs PCR of a plasmid (which is purified, double-stranded DNA and probably supercoiled). Considering all of these differences, I think that you are still left with a "relative" quantitation at best.

To address the efficiency problem with your cDNA, I would suggest trying to dilute the cDNA by an additional 1/2 and 1/4 and 1/10 or even more if your Ct values are low enough. If the efficiency of the PCR improves as the dilution factor increases, then you probably have some kind of inhibitor in your cDNA. Your reverse transcriptase manufacturer probably recommends that you dilute the cDNA at least 1/10 before use in a qPCR reaction. If you are already doing that, I would still recommend testing dilutions more than 1/10. If that does not improve the efficiency, then there is likely some issue with the primer sequences or PCR thermal profile.

Dear David F. Barker

Thank you for your suggestions. I will try qPCR again with diluted cDNA samples and let you know result.

An Nguyen

In my response above, I am assuming that you have already compared the melt profiles of the products from your cDNA and plasmid and that they are identical. Also that gel analysis of the plasmid/cDNA products shows the same pattern--not just the band size but also the strength of any "primer-dimer" bands (primer dimer would also probably show up on the melt analysis as a strong peak at low temperature). Full melt curve analysis and possibly a gel of your cDNA dilution tests along with a plasmid "control" could provide some clue about what is happening.