3
$\begingroup$

I have tried several housekeeping genes – Hprt, β-actin and GAPDH, to analyze the relative expression of a cytokine for measuring the inflammatory local response in mice ears. However, all these housekeeping genes show significant variation in both within and between different experimental conditions.

I really don't know how to solve this. I take the same amount of RNA (quantified by nanodrop) to make cDNA so I guess I have the same concentration in all the samples.

Any suggestion of what can happens and how to solve it? Could I use something different to measure the relative quantification? (My cytokine of interest is TNF).

$\endgroup$
  • $\begingroup$ When you say, "[they differ] from each other significantly", do you mean between replicates under the same experimental conditions or between conditions? It wouldn't be useful to normalize to a housekeeping gene expression level if it doesn't change between conditions. $\endgroup$ – Douglas Myers-Turnbull Jul 13 '16 at 19:05
  • $\begingroup$ They differ between diferent conditions and also under the same experimental conditions. I have checked the DNA concentration of the samples and they are the same. $\endgroup$ – Bio Jul 14 '16 at 11:16
2
$\begingroup$

You don't say how you're measuring but I'll guess that you're using quantitative PCR. A typical approach to handling housekeeping gene normalization in qPCR experiments is described in

Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes.

We outline a robust and innovative strategy to identify the most stably expressed control genes in a given set of tissues, and to determine the minimum number of genes required to calculate a reliable normalization factor. We have evaluated ten housekeeping genes from different abundance and functional classes in various human tissues, and demonstrated that the conventional use of a single gene for normalization leads to relatively large errors in a significant proportion of samples tested. The geometric mean of multiple carefully selected housekeeping genes was validated as an accurate normalization factor by analyzing publicly available microarray data.

There is an R package called NormqPCR that implements this approach (publication here).

Briefly, the approach required measuring multiple housekeeping genes and iteratively removing the ones that are most variable, ending up with two genes that are jointly least variable, and making a sort of pseudo-housekeeping gene from them with which to normalize your other genes.

This doesn't magically turn bad data into good data, but it helps clean up some intrinsic problems with housekeeping genes.

$\endgroup$
1
$\begingroup$

House keeping genes are not necessarily constant. Another way to measure gene expression change is RNA-seq. It normalizes counts against the whole transcriptome instead of against certain housekeeping genes.

$\endgroup$
0
$\begingroup$

If there was no established robust protocol, you could follow Eissa et al. 2017, Scientific Reports. They faced the same problem in another inflammatory model system in mice, and outline a strategy on how to find robust housekeeping genes (and then wrote a publication about this).

$\endgroup$
  • $\begingroup$ Can you include a brief description of the strategy? $\endgroup$ – Michael_A May 31 '18 at 23:59
  • $\begingroup$ I believe to possibly do more harm if compressing the full paper into a paragraph. $\endgroup$ – tsttst Jun 1 '18 at 17:02
  • $\begingroup$ It's hard to tell if OP is running an array or qPCR with carefully selected targets. If it were an array I'd suggest the OP take the genes across all samples, calculate %CV and use the lowest CV genes as normalization genes. He'd just have to cut them from the analysis. It's harder to determine what you should use one at a time. Perhaps some RNAseq, microarray or Nanostring data exist in his model, somewhere, he can use to do this. $\endgroup$ – CKM Jun 1 '18 at 19:10
  • $\begingroup$ Why downvote? The above answer is correct. The publication describes a general protocol to deal with the question of OP. $\endgroup$ – tsttst Jun 6 '18 at 18:11
0
$\begingroup$

Finding a suitable reference gene is tricky. It depends on your prior knowledge about the system. You may start with the same amount of RNA but errors can accumulate many other steps. That is why you use a reference gene to minimize the effect of those errors accumulated during sample processing. So unless you are very precise, you cannot say that everything was constant and the expression is the source of variation. Once you test a putative reference gene thoroughly you can "believe" that it would not change when you repeat the experiment.

Having said that, I would insist that a much better (and quite easy) method is to use a "spike in". It is a synthetic nucleic acid that you add to your samples. Similar concept can be used for proteins and metabolites too. You can spike in the same amounts of the synthetic RNA in your different RNA samples. If you keep the volume constant then the concentration of the spike-in would be the same in different samples. So, instead of a housekeeping gene you would now use the spike-in as a reference. Spike-ins have been found reliable for even single cell RNAseq studies (Lun et al., 2017).

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.