RNA purification, DNA contamination, and small RNA sizing profiles

[kumard]

Hello all,

I am extracting total RNA from tissues derived from wild animals after 24-48 h of death. I am able to get a RIN of approximately 7, but my samples have DNA contamination. I am using only 2.5mg of tissue (homogenized in 1 mL of Trizol-like proprietary reagent).

I have tried Trizol-based method (without DNase treatment) that claims to remove DNA and a silica column based method (with On-column DNase treatment from two different suppliers). My RNA concentration approximately 150 ng/uL and DNA contamination is approximately 3 ng/uL (2% of total RNA). Is this level of contamination "tolerable" for RNASeq (both large and miRNA).

I know that getting rid of DNA completely is not possible. But, what should I do now?

My second issue is about change in the small RNA profiles as seen in this image below. The column-based purification is certainly changing/reducing the small RNA species. I do not want to use the columns for this reason, but DNA contamination is giving me a headache now.

I want to get feedback and guidance from those who have first-hand experience in extracting RNA, performing QC, and generating cDNA libraries for RNASeq.

It seems to me that the companies advertise false information and claims are generally not true (May be true for RNA extraction from model organisms and cells) and one need to develop a custom-made solution for their use.

Could someone please also guide me to the published literature on such problems in RNASeq?

Have a nice day!

Thanks,
DK

[olafblue1955]

What RNA-seq method are you wanting to use? Typically, DNAse treatment in solution works fine, followed by rRNA depletion. Our library preps are done using a stranded system (any DNA reads would be "unstranded" or map to both strands). I have even taken to total nucleic acids sample, treated with Turbo DNAse, run rRNA depletion and then standard TruSeq Total RNA library prep. This should be done sequencing tomorrow or Monday; then we run RNA-Seq alignment and look at the stranded percentage. If there is excessive DNA in your samples, then percent stranded metric will be lower than 97% or so.

[kumard]

What RNA-seq method are you wanting to use? Typically, DNAse treatment in solution works fine, followed by rRNA depletion. Our library preps are done using a stranded system (any DNA reads would be "unstranded" or map to both strands). I have even taken to total nucleic acids sample, treated with Turbo DNAse, run rRNA depletion and then standard TruSeq Total RNA library prep. This should be done sequencing tomorrow or Monday; then we run RNA-Seq alignment and look at the stranded percentage. If there is excessive DNA in your samples, then percent stranded metric will be lower than 97% or so.

HI! Thanks for your message. I will send the samples to our Core for TruSeq Small RNA library prep or SMARTer smRNA-Seq. Did you perform PicoGreen of QubIt DNA HS assay for your RNA sample? If yes, how much DNA do you see?

[olafblue1955]

Hi,

I only use Qubit for final library quants after PCR and AMPure cleanup; I've not done DNA quants after RNA extraction (our samples are either samples that are acquired from retail sources of from collaborators who use standard RNA purification, with DNAse I/Turbo DNAse treatment. Sample integrity is assayed bioinformatically after sequencing and use of an RNA-Seq aligner program (Bowtie, etc.)

Is the main issue with DNAse removal something to do with the sample types or the data that you are trying to acquire?

[luc]

Qubit DNA assays will also measure Doublestranded RNA, e.g. ds parts of rRNA.
Do you see any signs of DNA on the Bioanalyzer?

[olafblue1955]

I don't, but then again I only use the Qubit High Sensitivity DNA assay for measuring the quants on my dsDNA library outputs.

[luc]

I guess your RNA coul actually be clean, DNA-free.

[torben]

Most likely your RNA is free from DNA. Even though the Qubit DNA assay is highly selective for double-stranded DNA, pure RNA will also give some signal, see figure 1 in https://assets.thermofisher.com/TFS-...S_Assay_UG.pdf.

[kumard]

Hi,

I only use Qubit for final library quants after PCR and AMPure cleanup; I've not done DNA quants after RNA extraction (our samples are either samples that are acquired from retail sources of from collaborators who use standard RNA purification, with DNAse I/Turbo DNAse treatment. Sample integrity is assayed bioinformatically after sequencing and use of an RNA-Seq aligner program (Bowtie, etc.)

Is the main issue with DNAse removal something to do with the sample types or the data that you are trying to acquire?

I am a newbie in RNASeq. Most of the RNA-library prep kits talk about using DNA-free RNA. That is why I am concerned about the residual DNA.

[kumard]

I guess your RNA coul actually be clean, DNA-free.

I hope so. But, then why the instruments like Qubit are so popular when it cannot determine low levels of DNA in the presence of RNA. Is this false advertising or I am missing something here. I guess Qubit DNA HS could be a better job in quantifying the cDNA libraries than residual DNA in RNA sample.

Thank you!

[luc]

I guess Qubit DNA HS could be a better job in quantifying the cDNA libraries than residual DNA in RNA sample.

It will.

Who says molecular physics/chemistry is easy? These signals are likely due to the nature of dyes. Likely the HS-DNA dye will mostly bind to dsDNA to become fluorescent, but also to other ds nucleic acids (dsRNA stretches ? or DNA/RNA hybrids?)