Hi,I've checked the expression of some microRNAs by qPCR and I've used 2 reference genes(2 snoRNAs) for normalization my data.But I found my Ref genes aren't match completely. It means one of them shows expression 1 when another is as Ref. How can I understand which of my Ref genes have a steady expression than the another one?
Hi Maryam,
first, it is a good idea to use several normalizers. That never harms anybody and provides you with a better idea how well your gene of interest is really differentially expressed. I personally use 2 to 3 normalizers for microRNA qRT-pcrs since I trust none of them completely.
I am not sure what you mean with < and >1. How much do they deviate from each other?
There will always be errors: piptetting, cDNA, RNA concentrations etc etc. Then these small RNAs themselves may not be perfectly kept at an exact level. I would allow for a 20% deviation meaning that normalizer1 is 1 and normalizer2 is between 0.8 and 1.2. There are probably nice mathematical ways to determine the exact level they are allowed to fluctuate until one of them has to be declared a non-normalizer. So, don't expect them to be each within 1%.
Overall your question boils down to the other quetion: what is a good normalizer for microRNA expression in qRT-PCR. That may depend on the species you work in, maybe even the tissue. We use for human microRNAs miR423, miR5701 and U6.
Thanks for your question!
Thomas
Thanks Thomas.
When normalizer1 is 1, normalizer2 is 1.658 and when normalizer2 is 1, normalizer1 is 0.7. our samples were kidney tissue from mouse model. Do you know how I can understand which normalizer has steady expression?
Best,
Maryam
Do you think this is too different? This is an average of several kidneys? I think it is still in a range that is acceptable, expectable. For tissues, I would go with a microRNA to normalize another microRNA because the other small RNAs may degrade differently than microRNAs (that do not seem to be affected that much at all by poor RNA quality). Since you use tissue, the RIN values will never be perfect and the degradation of non-microRNAs may be different from sample to sample. Then, since you use tissue, a mix of many cell types, you may not be able to get the normalizers that close to each other.
Perhaps you can use a mouse microRNA expression dataset (such as GSE48270) and identify microRNAs that are stable between many samples. We did this for human samples and came up with miR-423 (which others also identified as stable) and miR5701. They are not perfect but, I think, they do the trick.
If you look at the options for mRNA normalization in qRT-PCR you have the same problems. We use sometimes 4 different normalizers (Gapdh, RPL13A, UBE2d2 and EIF3d; and frequently one of them is a little bit off; HPRT1 was very often off). But they are rarely all 4 with less than 10% difference, probably never.
It also depends, what level of dysregulation you want to measure. 5-fold, 2-fold, 1.3-fold... The closer you get to the variability to the normalizer(s) the tougher it will be to define dysregulation of your microRNA of interest. And then the question is: is a 1.3 or even 2-fold difference in the expression of a microRNA relevant? Some people will clearly say no. I am not sure, but it makes more sense to go for the strong differences of substantially expressed microRNAs (see the "guidelines" by Competitive endogenous RNAs cannot alter microRNA function in vivo.
Broderick JA, Zamore PD.
Mol Cell. 2014 Jun 5;54(5):711-3. doi: 10.1016/j.molcel.2014.05.023.)
After your explanations, I don't think, this range is different!
Thanks a lot.
Best wishes
Maryam
- Badges
- Science method