When to dilute samples for qPCR - (Sep/15/2014 )

The normal protocol in my lab is as follows:

1. RNA extraction
2. RNA quantification
3. Reverse transcription of 500ng RNA in 20 uL reaction volume (each sample is diluted uniquely at this step)
4. Dilution of each sample 1:25 to yield 1 ng/uL of cDNA
5. qPCR of 2 ng of cDNA in 5 uL reaction volume

I thought I'd be real clever by consolidating the dilution steps in #3 and #4 into a single step by modifying the protocol as follows:

1. RNA extraction
2. RNA quantification
3. Reverse transcription of 10 uL of RNA sample in 50 uL reaction volume (varying quantities of RNA input per sample)
4. Dilution of each sample to yield 1 ng/uL of cDNA (each sample diluted by a different factor)
5. qPCR of 2 ng of cDNA in 5 uL reaction volume

Does anyone see potential problems with this? I'm using the High Capacity reverse-transcription kit in a 50 uL volume, so I should be able to fully reverse-transcribe 5 ug of RNA, and none of my samples have so much RNA as input.

However, my reference genes have Ct values that are up to 2-3 cycles different between samples. I think the reference genes should normally only vary up to 1 Ct between samples, otherwise it would not make a good reference gene, right?

Does anyone have an explanation for the seemingly non-stability in my reference genes based on my protocol?

Thanks!

In order to do RTqPCR, you must determine the efficiency of the RT and of the PCR.  Without these controls, your results are uninterpretable.  Please google the MIQE guidelines to learn how to do these experiments correctly.

My qPCR efficiency, depending on the target, is between 95-110%. My RT duplicates are within 1 cycle of each other. My melting curves reveal specificity of the primers (one peak only).

What other MIQE guidelines are you referring to, mlomonaco?

I do see in the MIQE guidelines that it is recommended to use the same mRNA input into the reverse-transcription reaction regardless of sample. Perhaps this is accounting for the variation I've mentioned. Is this strictly due to differences in efficiency of the RT reaction given different input amounts?