qPCR virgin, What do I do first? - How to set up experiment (Apr/25/2011 )

This is my first time ever doing Sybr Green qPCR. I theoretically understand how Sybr Green qPCR works, but I'm not sure how to set up the experiment properly.

I am using primers for ActB as internal controls. What do I do first? I heard that I have to set up a standard curve for each set of primers. How do I do this? Do I need to do a temperature gradient? Do I make my standard curve with regular PCR or in the qPCR machine. Can I use any cDNA or do I have to use the actual cDNA that I will test? How many replicates are needed? Do I have to run ActB each time I actually want to test a new gene of interest.

I've looked through a lot of the other discussion topics on qPCR and it seems everyone at least understands how and why a standard curve is needed and is useful, but I'm not sure why or how to do this.

Please help. I think I'm lacking the basics here. I was in another field before I stared in science. I have a lot of techniques under my belt but with this one I am,

Thoroughly confused

What you should do very first:
1) If you never run these primers (for Actb and your genes of interest) you should test/optimise reaction conditions on a positive control i.e. some cDNA you're sure to contain these genes. Not necessarilly the one you need to quantify, actually it would be better if it was some other cDNA you have in abundance and should contain enough of all your genes. Start with usual conditions (Ta = 60), if you have dimers on your melting curve then increase the temperature, if you see no ampification try to lower it. Depending on the mix you use you can play with magnesium concentration too. Theoretically you can find optimal temperature on classic cycler (doing a gradient), but in my experience there is a difference between the results from classic cycler and real-time one, so I'm usually doing it all in the qPCR machine.
This way you know the primers work.

Now you need to decide what are you going to use for calculation of relative quantity, there are several options, simple delta-delta Ct (not recomended because it doesn't calculate with different efficiencies), Pfaffl method (similar to delta-delta but efficiency corrected, my choice) or a standard curve method, when you calculate quantities from the in-run standard curve every time. For the later two you need efficiency/standard curve, classic method is to do a dilution series.

You also need to decide what reference gene (housekeeping, control gene) it best for your aplication, you talk about ActB which is widely used, but there is evidence that ActB may not be that stable in some aplications. You can either test several reference genes on your samples, or look up papers doing the same type of experiment (same cell lines, tissues, blood, ...). But many people just stick with ActB and don't care.


2) When you have working primers and picked referenc(es), you can do a standard curve for calculation efficiency for each of your genes. You could use any cDNA again, but it's better that you have one with high expression, because you going to dilute it (10x dilution, at least three times (like nondiluted, 10x diluted, 100x diluted), ideally 3 replicates each). Run it on same PCR conditions like your going to use for your samples. Each 10x dilution increases the Ct value about 3.3 cycles. You should make dilution such that they cover the range where you expect your samples (if you expect your gene to be 1000 fold downregulated do more dilutions to cover this range). My qPCR machine calculates the efficiency automatically from dilution series standard curve (you put values 1, 0.1, 0.01,... as a known concentration for dilutions described earlier), if yours doesn't, you need to create a regression curve in Excel and calculate it from displayed regression equation.
(if you decide for standard curve method the situation is different, you run standard curve (dilutions) in every run and use it to calculate quantities, which are efficiency corrected, this again may do your cycler automatically or you have to do it manualy in Excel)