Understanding qPCR and gene expression changes - (Mar/31/2009 )

Hi! I'm just starting to learn qRT-PCR. I find the whole thing a bit confusing. I hope maybe you can help me. I've investigated several different genes to find out wether they're up or down regulated. I understand that, after constructing the Ct line, the samples that reach this line with fewer cycles than others will contain more starting cDNA templat, and therefore will be more expressed than the others. Is this correct? Is there any way to calculate the starting amount of cDNA in my samples?. To me,everything seem so relative. that we just have to compare the graph from each gene up against all the others. What tells me that this gene is up- regulated under the present conditions? or down regulated under the present conditions?

Would really appreciate some help

thanks

You are correct, a lower Ct equals to more starting cDNA template. This is because it takes fewer PCR cycles to generate fluorescence signal when you start with more cDNA template. Another thing to consider is that to determine this Ct what you are doing (or the software automatically does for you) is determine at what point the threshold line crosses the amplification curve. The threshold line is a level set by default at 10 standard deviations above the background signal generated during the early (first few) PCR cycles. Another way to think about this is that the threshold line crosses your amplification curve at the exponential stage of PCR amplification.

What tells you if a gene is up or down regulated is your reference gene (also known as housekeeping gene). Likely you are aware about the use of genes like GAPDH, 18S and beta-actin as genes used to normalize the loading of RNA in experiments like Northern blots. qPCR uses genes like this for the same reason. The concept is that these housekeeping genes change very little, or not at all, in expression across your experiment so any change in the signal generated by these genes is due to things like differences in the amount of cDNA loaded in each reaction. By normalizing the signal generated by your genes of interest to the signal generated by a housekeeping gene then you can infer if your genes are up or down regulated.

ivanbio on Mar 31 2009, 06:13 PM said:

Thank you very much I see things more clearly now, I just have a few things on my mind. Let me see if I have understood you correctly. The reference gene represents the standardized expression, and we will therefore compare our genes of interest with that reference gene. I've used the Pfaffl method to calculate the expression ratios. But how can we tell from these expression ratios how a gene is up or down regulated? In this formula we normalize our target gene with the reference gene, and comes out with a number. Would the genes with higher (E^CP) than the reference(E^CP) be up regulated, and those with lower (E^CP) be down regulated?

Sciencefreak on Mar 31 2009, 02:02 PM said:

Yes, the higher the E^CP, the higher the relative expression and vice versa. As you said before, these are RELATIVE estimates of gene expression. qRT-PCR without precise standard curves cannot be used to determine the absolute change in gene expression. Also, keep in mind that mRNA changes do not always equate to changes at the protein level.

I think you've got it. By normalizing the expression of your gene of interest to the expression of the reference gene, you remove most variation due to different amounts of template in each reaction and things like that. For example if for some reason there was twice as much cDNA in one well as compared to the other (and you did not know this), then you would conclude that there was twice as much expression of your gene of interest in one well compared to the other. Now if also ran your reference gene on these samples, you would find that this gene also showed twice as much expression in one well compared to the other. Now, if you take your reference gene into account and use it to normalize the expression of your gene of interest, what you would find is that there is no difference in expression for your gene of interest in these two samples. Hope this made sense .

I am not sure that I follow your explanation of the Ct values. The way you know which gene is up regulated versus down regulated is by comparing your results to a reference sample. Note that this is a sample, not a gene. For example, if you were looking at the expression of your gene of interest across multiple tissues (muscle, brain, kidney, etc), then you would choose one of these tissues as the reference sample (for example liver). By this rationale, any tissue that had a higher Ct for your gene of interest than the Ct for liver would be down regulated (relative to liver), while any tissue that showed a lower Ct for your gene of interest than the Ct for liver would be up regulated.

Hope this helps.

Sciencefreak on Mar 31 2009, 11:02 AM said:

Thanks:) Why is it so important to measure CT values in the exponential phase of the amplification curve? and what is the advantages and disadvantages of using the Fit point method compared to the 2nd derivative method?

thanks

It is important to measure the Ct at the exponential phase of amplification because it is at this stage that the reaction is as close to 100% efficient as it will ever be (no limiting reagents). If you wait until later then your measurements will be off by issues like used up reagents. All qPCR software do this for you automatically.

The fit point and the second derivate methods are just two ways of determining the cycle threshold value. I cannot really think of any advantages or disadvantages for either one: they are just two ways of obtaining very similar results.

Sciencefreak on Apr 19 2009, 12:13 PM said:

ivanbio on Mar 31 2009, 06:13 PM said:

I am just getting started with RT qPCR, but what I understand from my supervisor, many traditional housekeeping genes are actually not expressed as predictably as we like them to be and there is great variations among cell types and tissues. Therefore it should be stressed that a given tissue must be analyzed for proper normalization genes before doing your experiment, or there's a risk that you normalize to a gene that will lead to wrong interpretation of your data.

A former student of the lab did this work which explains the problem in more detail: Bonefeld BE, Elfving B, Wegener G: Reference genes for normalization: a study of rat brain tissue. Synapse 2008, 62:302-309

Hi, just wondering about the use of housekeeper; do you think using Atp5b and Rpl10 are okay?

We found that using qBASE Plus software (from Biogazelle) was very helpful in determining which genes to use for normalization. We started with 10 so-called "housekeeping" genes in our organism (Boophilus microplus) and in the end only found that two of these were sufficiently constant to be used for normalization. Interestingly enough Beta-actin was not one of them (it has been used as a normalizing gene in various other previous publications for B. microplus).

You might like to check the following article - it really discusses why you should always check if the reference genes you have chosen to normalize against can be used under your experimental conditions.

Guénin, S., et al, 2009. Normalization of qRT-PCR data: the necessity of adopting a systematic, experimental conditions-specific, validation of references. Journal of Experimental Botany 60, 487-493.

And the article for qBASE:

Hellemans, J., et al, 2007. qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biology 8.