PBGD or HMBS have worked well in a variety of hypoxia experiments in our labs (mouse, rat and man).

I would highly suggest that you get assays for a range of different control genes and see how they are affected by hypoxia. This paper has 10 options for you and shows how to go about selecting the best one: http://genomebiology.com/2002/3/7/research/0034/

I do agree with Michael. The best is to pick up several putative reference genes for your condition taken either from literature or from other sources and check them by qPCR. Do a box plot with the Cts and pick up the one(s) that have best behaviour. when compared under your conditions. If taking several use an average of the Cts for referencing.

For human cells, we've also used 18s, seems to work ok.

But I use ovarian cancer cells.

I have used beta actin and Hprt to compare hypoxic and non-hypoxic samples. Both worked.

Thank you very much, your comments have been really helpful!

Hi Carina, We also used 1% hypoxia in our experiement with endothelial cells and I came across the same problem with Reference genes; beta actin and gapdh did not work for us. Ideally you should run various housekeeping genes and figure out the best. I went thorough some references with hypoxia and endothelial cells and I found that HPRT1 and Beta-2-Microglobulin were found as most stable. I ran both and found that they were stable. Usually I run 2 reference gene all the time with my RTqPCR now.

I would use three reference genes for this type of analysis (e.g beta actin and lamin A/C...etc.). You may find this reference helpful (http://link.springer.com/article/10.1007%2Fs13353-013-0173-x).

Good luck

The stability of the reference genes varies greatly upon different perturbation experiments. According to our recent study, Alu-repeats show relatively stable expression regardless of the experimental conditions, and we recommend an Alu repeat assay as first candidate for normalization of RT-qPCR data. We also recommend to test the stability of several other reference genes, and use two or more assays for the normalization. Please have a look at this paper:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0071776

Hi Carina,

Normalisation of adequate reference genes is very important in your study. Did you know that more than a single reference gene must be selected? Reference genes differ from tussue to tissue, from healthy or diseased, at different developmental stadia. A set of stable reference genes (mostly 4 or more) must be validated in both normal and experimental conditions. Validation also implies stadard curves )at least 6 dilutions) and efficiency calculations. Moreover a reference gen must be chosen in the same Cq range that can be expected for the gene of interest. This because of the fact that abundant axperssion of a reference gene suppress the synthesis of the amplicon at interest. GeNorm is an eample of a tool that can be used. Look for instance at http://www.biogazelle.com/four-tips-rt-qpcr-data-normalization-using-reference-genes

we use RPL13A as a loading control. Actin and GAPDH are all regulated by hypoxia.

The perfect housekeeping it is not simple to find, by definition a good housekeeping should maintain a constant expression in all the cells and conditions.

But what is a housekeeping gene? Every single cell need a set of genes on charge to keep the cell alive and healthy, this Housekeeping genes are responsibly for maintain the cell cycle control, structural integrity, cell death control, metabolic process an so on.

However as you know the expression of thisgenes can be altered under no physiological circumstances like disease.

For this reason a HK gene optimal for same particular case can be not suitable for a different scenario.

I would recommended to you don’t use genes involved in metabolic process control when you are working in hypoxia like is the case.

I think this papers can help you

http://www.gmo-qpcr-analysis.com/rubie-norm-2004.pdf

http://www.sciencedirect.com/science/article/pii/S0168952513000899

Kind Regards

Sonia

We use 5 reference genes for exps at 5%O2, ACTB, B2M, GAPDH, HPRT1 and RPLP0.

What tissue types are you examining?

I am quite sceptical about stability of reference genes (especially if you use only one) unless there are really strong evidences that it works. So I usually do like this: I clone the gene I am investigating. When I have recombinant plasmid I measure the concentration and count copy number. Than during QT PCR I am doing serial dilution - standard curve. The result is expressed in direct copies number of your gen. In this way you are 100% sure that non stable expression of your reference gene will compromise your results.

You could use 18S rRNA but for retrotranscription you have to use the random primers

You could use http://medgen.ugent.be/~jvdesomp/genorm/ to determine the most stable reference (housekeeping) genes from a set of tested candidate reference genes in a given sperimental cell conditions:

Hi Carina,

Our lab has established a set of reference genes (in adipose derived stem cells) that are stable under hypoxic conditions These are YMHAZ, TBP and GUSB.

Take a look on this article

http://www.biomedcentral.com/1471-2199/9/98

Good luck!

Alternative is to go for digital PCR which does not require a reference genes and gives you copy number. However it require access to digital PCR machines. you can do it manually but too cumbersome.

best of luck

I use Beta-2-microglobulin for both primary rodent cells and immortalised human cells under hypoxic conditions. I found 18S rRNA would vary with treatment and oxygen.

Ideally, you should select a panel of candidate reference genes. Run real time PCR, and test their expression level under normoxic vs. hypoxic conditions. Then analyze the data and compare expression stability (M) using the Normfinder or SAS qPCR, etc software. Further you can determine how many optimal reference genes are needed by calculating , a pairwise variation Vn/n+1.

Perhaps , reading this paper will help

Evidence based selection of housekeeping genes.( 2007)

http://www.ncbi.nlm.nih.gov/pubmed/17878933

I'd start with 18S. Even if you used oligoDT for retro-transcription you can still get it due to missprimings. Since 18S is so abundant, there's very high detection (low CT values) and saturation may be an issue when normalizing it with your genes of interest.

Personally I think that the use of different houskeeping gene is the best way for the normalization, but which one use is the more problematic answer , you must to identify the cell type and also the treatment.

For the right choise do you must use a program like geNORM.

http://www.primerdesign.co.uk/products/9460

Hi,

You can try SDHA, YWHAZ & UBC as a reference gene. Since, I am also working with 1% O2 and they are always consistent with normoxia and hypoxia.

For normalization, I would also strongly suggest to use the geometric mean of at least two genes. I used to test several potential housekeeping genes, load them into GeNorm software to determine those with most stable expression across samples, and normalize results using the geometric mean of the latter.

For original publication of this method:

Vandesompele, J., De Preter, K., Pattyn, F., Poppe, B., Van Roy, N., De Paepe, A., Speleman, F., 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes.

Have you tried to test simple and widely used beta-actin gene? otherwise I do support Muhammad Alam suggestion too.

Try

polar2A

I tested RPS6, GUSB and HPRT against each other in 5% and 21% in murine cells. RPS6 and GUSB were quite stable - HPRT in contrast did show quite some differences to the other two genes.

May I suggest using a set of different housekeeping genes at one test? Eliminate the non-expressed genes. Then try to reduce or increase the concentration of the reference gene as well as the target gen at another test. In this case you will have a better idea.

Good luck

I tried 18S before, and it is a good internal control for hypoxic condition.

Rps2 is stable in hypoxic and normal conditions

Try 18s rRNA.

I disagree with the 18s rRNA as a good internal control in RT-qPCR experiments for the following reasons:

1 - The rRNA constitutes more than 90% of total RNA. It is more or less a constant fraction of the total RNA, and this fact led many researchers to use it as a control for normalization of gene expression data. However, it has been shown that equal fractions of rRNA do not necessarily ensure equal fractions of mRNA, check out this paper (http://www.ncbi.nlm.nih.gov/pubmed/11141312)

2- The issue of imbalance in the ratio of rRNA/mRNA is even more pronounced in cancer cell perturbation experiments. This is due to the fact that perturbations may lead to differential expression of RNA polymerase I and/or II, or differential degradation of the two RNA populations, and consequently would lead to different rRNA/mRNA ratio.

3- The high amount of rRNA as compared to mRNA makes it difficult to substract the background fluorescence in data analysis of RT-qPCR.

More details about this issue can be found in this paper:

(http://www.ncbi.nlm.nih.gov/pubmed/?term=Effective+Alu+repeat+based+RT-qPCR+normalization+in+cancer+cell+perturbation+experiments)

This paper may help

It would suggest that In hypoxic conditions the genes involved in the regulation of glucose metabolism e,g. glucose regulated protein, molecular chaperones in endoplasmic reticulum may be considered.

Could you please read attach pdf

I agree with Luana Schito. We culture chondrocytes under 5% O2 and for the expression analysis we use geometric mean of RPL13A and B2M

Actb was the most stable housekeeping gene under hypoxia....

I agree with Ali , 18sRNA is not a good choice, it is too abundant in cells. I agree to the suggestions that you should try different control genes and reference some publications.

You might wish to look in large scale microarray gene expression analyses for the stability of any potential housekeeping gene. In our experience, different tissues show different HKG stabilities. Microarray data can also be exploited for the selection of HKG with similar expression levels to the gene(s) of interest or, for multiplex HKGs, with expression levels spanning orders of magnitude.

Actin is also stable under hypoxic conditions.

Everybody must keep in mind that nowadays in the good journals 3 internal controls are required for qPCR and these controls may vay according to the cell type or the experimental setting used. For Hypoxic conditions, GAPDH, 18S rRNA must definitly be discarded. cyclophilin can be used confidently. actb can be used too. The best answer has been provided by Dr Pfeffer.

Hi,

You've got a lot of good answers already. I hope not messing up two much, but my experience working with cancer cell lines, SCLY is a fantastic reference gene for rtQPCR both in vivo and in vitro experiments.

Good luck!!

We used 36b4 in an in vitro model (24-48h hypoxia). That worked well at 1-3% O2.

why not just actin, I did it for HIF-1 loading control, it works well

Mol Vis. 2008;14:2575-96. Epub 2008 Dec 31. Analysis of partner of inscuteable (mPins) expression in the developing mouse eye. Raji B1, Dansault A, Vieira V, de la Houssaye G, Lacassagne E, Kobetz A, Arbogast L, Dufier JL, Blumer JB, Menasche M, Abitbol M.

cyclophilin corresponds to a stable internal standard in qPCR in many experimental setting see articles published in Molecular Vision under the name of Marc Abitbol

The Primers for Semi quantitative RTqPCR can be found in

Of Course for qPCR the primers must be redesigned