Few things to consider before you go too crazy

1. when you say RNA quality is good, what do you mean? 260/280 should be around 2.0 but more importantly 260/230 should be >1.7 if this is low then you are having some organic carryover from isolation.

2. If 260/230 is good then run the RNA with a bioanlayzer to look at the integrity and make sure you are not getting hydrolysis

3. did you prep all of the cDNA at the same time with the same master mix of dNTPs, primer, and enzyme? this could account for delay in Ct. also was the RNA prepared similarly? is it from tissue - ensure proper technique/timing so as to keep the RNA intact

4. the normal housekeeping genes are not good for your samples/model. Try RPL0 or RPL19 as HKG

5. ensure you are only getting one PCR product with the melt curve, non-specific products could be sucking down your primers early on and delaying detection of the real target.

Nanodrop measures only the concentration and the purity of nucleic acids. It does not distinguish between DNA and RNA, and it does not indicate the integrity of the molecules.

Hence, it is possible that you measured considerable DNA contaminations (and if your primers are intron-spanning, such contaminations will -ideally- not produce PCR products) or that your RNA was strongly degraded.

DNA contaminations might be avoided/reduced using a DNAse treatment during RNA purification. You can measure the content of specifically DNA and RNA using fluorimetry (e.g. Qubit). RNA integrity can be checked using (capillary-)electrophoresis. In eukaryotes you should see a clear 28S and 18S rRNA band, with the 28S band being about twice as intense as the 18S band.

Another reason could be that this sample contains some kind of PCR inhibitor. In this case you should check dilutions of the cDNA. Every 1:2 dilution should increase the Ct value by one cycle. If it contains inhibitor, the inhibitor is diluted, too, and you will see a shift of less than one cycle per 1:2 dilution.

Those Ct values are just fine. It just means you have less RNA in some biological samples compared to others. The entire reason you include the housekeeping genes is so you can normalize data and compare.

Katie A S Burnette , no I think that this is not "just fine". You have similar samples and extract comparable amounts of RNA of comparable purity and you process all samples in the same way, and you get one with "outlying" Ct values. This is absolutely worth to be investigated. I don't think that a good scientist should simply go over such overvations. You are certainly correct that we use reference genes (not housekeeping genes!) to correct for differences in sample concentrations. That would be adequate when there is random variation,and normalization will reduce this random variation. But here is one sample misbehaving. This might include a considerable shift between gene of interest and reference gene, so you cannot even assume that normalization will work in this instance. It may be a problem of this particular sample, but it might be a problem of the assay as well. It seems very relevant to me to know or investigate what the source of the observed discrepancy is.

I agree with Jochen Wilhelm , those Ct values differ by 7-8 that is 1000-fold difference. if the RNA quality, integrity and all those other things check out then those genes are just not good for housekeeping. Hossein Nasr Azadani you could spike in an exogenous RNA in your isolation to track if you are losing RNA along the way.

Hello Hossein Nasr Azadani

First, if you are using a technical triplicate and there is a wide difference between their Ct values, this would mean a mistake in nanodrop itself. if that is not the case, try to check the efficiency of this sample reaction. If it is not ideal, please add DMSO at a range up tp 10% (you have to optimize that).

best....

Did you start with similar amounts of biological material? The easy answer is that one sample is simply less mass.

Thank you all Dear Researchers for your kind and excellent comments and guides. I did all steps of experiments for all samples including RNA extraction, cDNA synthesis, qRT-PCR and etc in the same conditions in details but on different days for each sample. Dear Julie Ann Dougherty, the ratio 260/280 for this sample was 2.06 and the ratio 260/230 was 1.3, BUT I'm working on end-stage cirrhotic liver tissue that totally loses its function, Moreover about all my samples have the same situation regarding 260/230 ratio but just this sample doesn't work well in qRT-PCR. Dear Jochen Wilhelm I used DNase treatment on the column during the RNA extraction process by RNeasy Micro Kit (Qiagen), but the presence of some inhibitor in the sample is possible, however, I try 2 dilutions of cDNA (50ng and 100 ng) but the results didn't differ. Anyway, ideas and comments from all of you are considerable and appreciable.

50ng and 100ng of cDNA are both MASSES of cDNA.

I use ~8ng per well, and I work with some low abundance transcripts. For many of my genes I could easily use less (4ng or even 2ng).

Too much cDNA can be inhibitory to your PCR reaction, so using 50ng (or 100ng) might result in very inefficient (and inaccurate) amplification.

You'll also run out of cDNA very quickly this way.

Dilute your cDNA at least 1/5, ideally 1/10, before use (for reference, I dilute 1/20 and use 2ul per well).

Also, re: 260/230 ratios, these reflect chaotropic salt contamination (guanidium salts, essentially) and are expressed as ratios not least because even pure RNA will absorb at 230, but also because the absolute amount of salt isn't usually as useful as the amount of salt relative to your RNA.

In other words, two samples with the exact same salt contamination, but with very different amounts of RNA, will give you markedly different 260/230 ratios, potentially one acceptable, one unacceptable. And this is working as intended, because for downstream applications using these samples, you'll usually be adding absolute amounts of RNA, and thus from the high [RNA] sample you'll only need to add a tiny bit (and thus not much salt) while from the low [RNA] sample you'll need a much larger volume of sample (and hence greater salt), even though both samples have the same initial salt contamination.

BUT

Because for most kits there is an absolute limit on volume of RNA you can add, there is concomitantly a point at which salt contamination becomes less relevant.

Consider: a sample with 400ng.ul-1 RNA and 260/230 of 1.4 is not something you should use in cDNA synthesis, because that's a lot of RNA and thus also a lot of salt. Even adding 2.5ul needed to make 1ug total RNA means adding a lot of salt.

Conversely, a sample with 40ng.ul-1 of RNA, 260/230 of 1.4 is...yeah, probably ok to use, because that's not a lot of RNA and thus also not a lot of salt. Even if you wanted to, you couldn't add the 25ul needed to make 1ug, and in reality you'll probably only be able to add maybe...8-10ul (depending on kit), meaning you'll also add less salt (ratio, remember).

It may be that some of your samples are very low [RNA] (you don't specify) and poor 260/230, while others are high [RNA] and poor 260/230. The former samples might work, while the latter might not.

Also, "end stage cirrhotic liver" is going to be heavily fibrosed and fatty, I would assume? These are still transcriptionally active cell types (fibroblasts and adipocytes don't do a lot, but they are transcriptionally active) so you should still get RNA. Just because the tissue has lost all liver function, it's still viable tissue.

Dear John Hildyard, Thank you much for sharing your precious experience and great explanations. YES, the samples are fully fibrotic and fatty tissue. So, as you said, I'll try with more dilute cDNA (less than 10ng/reaction) or re-synthesize cDNA from a less RNA elution of this sample even though involving less amount of RNA than the constant amount (1ug/RT reaction) applying for all samples to study gene expression status.

Well, you could also just clean it up, if it's too salty.

Make vol up to 500ul with nuclease free/DEPC-treated water (to make downstream volumes more manageable).

Add 50ul of 3M sodium acetate, pH 5.5

Add 10ug of glycogen (optional, increases recovery)

Add 500ul of room temp isopropanol (room temp is important, here)

Mix well

Incubate 20mins at room temp

Spin at max speed for 15 min (chilled centrifuge is ideal, but room temp also works)

Remove supernatant

Wash pellet gently with ice cold 75% ethanol, spin 5-10 mins and remove supernatant

repeat ethanol wash (above)

remove all supernatant as completely as possible

air dry pellet for ~15mins

redissolve in nuclease free water.

Usually this reduces salt content considerably, while retaining most of the RNA.