To be sure that your housekeeping gene is long-lived in your cell type of interest (I assume you're treating cells with actinomycin D), I would recommend doing a time course. Treat cells with actinomycin D, then harvest RNA at 0-24 hours. Measure levels of your housekeeping gene at the different time points.

One big issue here is how much RNA you input into your QPCR reactions, given that actinomycin D may decrease the total amount of RNA in your cells over time. If actinomycin is decreasing the total amount of RNA over time, then by inputting the same amount of RNA in your QPCR reactions, you're actually adding too much RNA in the later time points. So to cover this issue I would recommend doing the QPCR two ways. 1. Input the same number of ug into your QPCR assay. 2. Input the same fraction of total RNA yield from each treatment into your QPCR assay (e.g. add 10% of the total yield from each treatment sample).

Also be sure to do a no actinomycin D control. In some primary cultures, the levels of "housekeeping genes" (beta-actin) can increase over time.

I wouldn't predict a reference gene to somebody else. Everybody is doing some other experiments, using other cell lines,...

The best way is to test different genes and then use min. 2 genes which have the the lowest difference in your treatment(s).

Cheers, Nadine

Thanks for your answer. But, I'm not asking for particular gene for my experimental set, but rather for a gene which is known to has long mRNA halflife (among other widely known reference genes). In the lab we are using actin, b2microglobulin, GAPDH, RPL29, RPL30 but I can't find the information which one has the longest halflife of mRNA.

Cheers,

Kamil

I used GAPDH in my qPCR as it consistently stable and less affected compared to other internal reference house keeping genes! But I am not sure about its half life exactly!

given that reference genes itself varies across systems being assayed, i would expect that the same holds true for their properties like stability as well. So even though you manage to find a reference gene with a relatively longer half life, it has to first be validated for your system/cell line... my advice would be pick out a set of commonly used reference genes and perform a simple rna degradation assay... have have rna preps and incubate it for a day or more (depending on your interest) and at regular intervals quantify the rna present for your gene of interest... so when you plot the quantity of template as a function of time of incubation, you may get a negative slope and depending on the half life of the template, the slope may be more negative or less negative.... That would give you a decent estimate of its half life as well i believe !!!!

As far as I know the mRNA of ribosomal proteins is the most stable so you could use RPL29 or 30, whereas the mRNA of Proteins involved in metabolic processes, like GAPDH, seem to be less appropriate (also due to quite high variability of the metabolic status of cells). Still probably it's best to use at least 2 or more reference genes. But for mRNA stability you could also use a set of primers for the 3' end of your reference gene and a set of primers for the 5' end of your reference gene. This 3':5' assay can be used to measure the integrity of your reference mRNA. Also described by Nolan et al. (2006) Quantification of mRNA using real-time RT-PCR; Nature Protocols

No, no. You need an RNA that is exceptionally stable. i have done this before and found that 18S rRNA has a REALLY long half life. You can graph its delta Ct over the course of the experiment and should find that it almost does not change for short time courses. This is because it is 1) highly abundant and 2) tied up as an essential component of ribosomes, which are stable and essential RNPs. If I had to recommend one, it would be a rRNA like 18S, which is already considered a normalization gene. But the sad truth is there is no RNA that will not change in these sorts of experiments. We are now turning to pulse-chase labeling with nucleoside analogs, like 4sU, that can be fed to cells at doses that are low enough to not affect RNA metabolism, but can be affinity captured later and RNA levels determined by qPCR or sequencing. In these experiments the half-lives determined were quite similar though, suggesting that actinomycin D may not be that bad afterall. So go with 18S. We use a taqman probe set from ABI.

Another vote for 18S. In our experiments in telomerase immortalized fibroblast lines, 18s varies less than 25% variation over 48 hours (at which point our cells were mostly dead). p16 (CDKN2A) ais another very stable gene that we used due to our lab's interest in the gene, but it is also stable out to 48 hours if you want a poly-adenylated gene as well.

Myc was our control for a short half life gene, and plumets to about 1/64th the concentration after 6 hours.

Hope that helps!

I Agree with Keith and William, 18s could be your best reference. Sometimes, you can also amplify it in oligo-dT cDNA in some invertebrates as molluscs and isects (see my last paper) and also in mice (I have some references). Other is to quantify against total RNA isolated, but you should be able to quantity your RNA by fluorometric assay (as PicoGreen specific-RNA or any other RNA assay kit). Best luck!

Thank you all for your advices !!! I will order primers for 18s and p16 to have two very stable control genes. Thanks you cordially !!! :)

I think that when you use oligo dT cDNA and amplify with 18S primers, then you are actually detecting processing intermediates of the 18S rRNA that are polyadenlyated but not very abundant. In this case I would rather prime the cDNA synthesis with randomers or a mixture of randomers and dT. Good luck with your experiments!

I looked at the reference genes used by array companies. They have a vested interest in good choices. I tested a bunch on an array used and all worked well individually. Calling Tech Support at the company might help you narrow down reference genes that are good in your system.

Use the 18S or 28S rRNA as reference, but not the external or internal part of the precursor rRNA.

you can also use a mixture of oligo-dT and random 6mers during RT. Jiradet's response is very accurate, and I can only add that autopriming of ribosomal RNA can occur (there is a reference that amplify it well without using any primer in the RT).

beta actin has more than 24 hrs half life. 18S can be a good candidate also

The 18S primers I designed were identical in human, mouse and rat. This is an advantage ...

To be sure that your housekeeping gene is long-lived in your cell type of interest (I assume you're treating cells with actinomycin D), I would recommend doing a time course. Treat cells with actinomycin D, then harvest RNA at 0-24 hours. Measure levels of your housekeeping gene at the different time points.

One big issue here is how much RNA you input into your QPCR reactions, given that actinomycin D may decrease the total amount of RNA in your cells over time. If actinomycin is decreasing the total amount of RNA over time, then by inputting the same amount of RNA in your QPCR reactions, you're actually adding too much RNA in the later time points. So to cover this issue I would recommend doing the QPCR two ways. 1. Input the same number of ug into your QPCR assay. 2. Input the same fraction of total RNA yield from each treatment into your QPCR assay (e.g. add 10% of the total yield from each treatment sample).

Also be sure to do a no actinomycin D control. In some primary cultures, the levels of "housekeeping genes" (beta-actin) can increase over time.

Good evening,

I tried to use 18S RNA and CDKN2A as a reference genes in my setting. While 18S RNA was abundantly expressed and very stable, CDKN2A seems not to be appropriate in my experriment. Designed primers generate dozens of products and it is maybe caused by existance of many isoforms of this gene, or mutations in my lymphoma cells (Raji). But still I am satisfied with the result from 18S.

Thank all of you for relevant help.

Cheers,

Kamil

Hi,

maybe a little bit old-fashioned ;-) I would try to validate qRT-PCR results from ActD assays by northern blot to get some hints about the full length transcripts. Normalisation by total RNA load on gel. Furthermore I think RNA stability can depend on used cell type!

Dear Kamil, you'd better check this out http://miqe.gene-quantification.info/

From figure 1 you'll notice that multiple reference genes are required for accurate normalization, meaning you have to assess several ref genes which prove stable over your experimental samples including controls, and conditions under study. You want to publish at the end, right? The fact that 18S RNA seems to you very stable reference gene - which I don't question - is not enough, you have to prove it.

An example here http://www.sciencedirect.com/science/article/pii/S0378432007001455

what about EIF4 gene? it has been used for Cordycepin assay.