RiboZero - Disappearing mRNA?

[rndouglas]

I'm having some trouble with Epicentre's RiboZero Plant Leaf Kit, and I'm hoping someone here might be able to point me in the right direction.

I have been isolating total RNA from maize leaves using a standard Trizol procedure. I DNase-treat that RNA, then use the RiboZero kit.

The starting RNA and the DNase-treated RNA look fine on the Bioanalyzer (see attached photo, the top two frames of the figure).

However, after I run the RiboZero kit and I send my sample to the Bioanalyzer there seems to be nothing but tRNAs left (bottom frame of the attached figure), and I'm really not looking to just sequence tRNA.

I've tried this on four separate samples, and I keep getting the same Bioanalyzer result. The first two times I followed the RiboZero protocol for using the Qiagen RNeasy MinElute Cleanup kit, and the last two times I tried Zymo's RNA Clean & Concentrator kit.

Our DNA Core facility hasn't had anyone else use an rRNA removal kit, so they haven't been much help.
Am I reading my Bioanalyzer results correctly? I believe our DNA core ran the RiboZero on a separate mRNA Bioanalyzer chip instead of a total RNA chip.

I know there's at least some mRNA present in my total RNA extractions (before using RiboZero, after DNase-treatment) because I'm able to make a pretty decent batch of cDNA from them.

Any suggestions? I'm at my wit's end!

[pmiguel]

I don't see anything necessarily wrong here. Our experience with maize (and other plants for that matter) is that rRNA seems to be >99% of the total RNA. Either that, or we just suck at obtaining high yields of non-ribosomal RNA from plants.

So, lots of rRNA, low percentage of non rRNA, what would you expect from successfully removing all the rRNA? Basically the picture you show. If you are doing normal RNA seq, I would suggest using the Zymo protocol that washes away the small RNA, then ask your core to run a pico RNA chip on your RNA. That would give you a profile of the length distribution of your resulting RNA without a giant small RNA peak squishing your mRNA down to the base line.

--
Phillip

[MPPFig]

I am also having trouble with this RiboZero plant leaf kit, I am posting bioanalyzer traces of one sample after I depleted the rRNA (notice a peak at 1300 nt). I continued with the construction of the library (ScriptSeq™ mRNA-Seq Library Preparation Kit) and this is the profile I obtained (second figure) in the bioanalyzer (HS-DNA chip). My concern is that my library pretty much contains this abundant fragment that was not clean up by the RiboZero kit. Has anybody seen this before?

[rndouglas]

I'm in the process of trying Phillip's suggesting right now. And by that I mean I'm still waiting for our DNA core to run the Bioanalyzer now that I've tried to remove all the smaller RNAs with the Zymo kit.

[pmiguel]

Originally Posted by MPPFig

I am also having trouble with this RiboZero plant leaf kit, I am posting bioanalyzer traces of one sample after I depleted the rRNA (notice a peak at 1300 nt). I continued with the construction of the library (ScriptSeq™ mRNA-Seq Library Preparation Kit) and this is the profile I obtained (second figure) in the bioanalyzer (HS-DNA chip). My concern is that my library pretty much contains this abundant fragment that was not clean up by the RiboZero kit. Has anybody seen this before?

What species of plant is this?
--
Phillip

[MPPFig]

Phillip:
This is Brachypodium, what do you think? Epicentre took the rRNA seqs and blasted those to their probes, they said RZ should pick up my rRNAs...

[pmiguel]

1.3 thousand nucleotides would make it a chloroplast "16S" (small subunit) rRNA, most likely. Possibly your sample had a very high amount of chloroplast in it and RiboZero could not deplete it completely. But it does suggest that your 16S rRNA may be divergent enough in sequence from the consensus sequence Epicentre likely used to produce their binding oligos. Either that or your 16S has temperature-stable secondary structure that is interfering with oligo binding.

You could ask Epicentre for help. They might be able to suggest a tweak to your conditions (like starting with half the amount of total RNA) that would result in a better depletion.

--
Phillip

[rndouglas]

Finally got my new Bioanalyzer results back, and I can't say I'm encouraged.

I used the Zymo kit to remove the big tRNA peak and wanted to see if I had any mRNA left.

Should I just trash these guys and start over?

[pmiguel]

How much did you get (total) and how much do you need?
Was this a nano RNA chip or a pico?
To me they look fine. Well, a little weird you have so much stuff at >4 kb, but maybe that is expected in your species.

--
Phillip

[rndouglas]

I only need 50 ng for my library prep, and I have about 300 ng, so that's not an issue.

This was done on a pico chip.

Shouldn't I see a little bump for mRNA on the X-axis somewhere?

This is maize...I have no clue if the bump at 4kb is expected or not.

[NormSci]

FYI, from the Epicentre website FAQ RE: Ribozero:

After I used a Ribo-Zero Kit, I noticed a peak at 100-140 nt in my Bioanalyzer traces (but no 16S/23S rRNA). Why is that?

This is normal and expected. The peaks observed in the 100- to 140-nt region on a Bioanalyzer trace correspond to tRNA, small noncoding RNAs, and small amounts of 5S rRNA (in cases where 5S rRNA removal is less efficient). These small RNAs are available for inclusion in a sequencing library if desired, but can be readily removed by gel or other size-selection techniques that are designed to remove small RNAs, either before or after library synthesis.

[lindseykelly]

I used the RiboZero Gold kit for Humans on human tissue RNA also extracted with Trizol. I don't have a picture right now, but our bioanalyzer looked the same as rndouglas' picture. We had a large peak at the beginning and then pretty much nothing. I talked to Epicentre and sent them my picture and they said it looked exactly like it was supposed to. We created the library with the Illumina TruSeq RNA sample prep kit and it looked great. We are still working on the data analysis, but so far they look as expected.

so basically, mine looked the same and my library passed all the QC before I ran it. I say, make your library and go from there.

lindsey

[MPPFig]

Thanks Phillip! I tested a sample using MicroPoly(A)Purist™ Kit, 50ug total RNA as the input, this is how it looks when ran in the Pico Chip, the bioanalyzer calls a 3.5% rRNA contamination. It does look better than the RiboZero output, doesn't it?

[pmiguel]

Originally Posted by MPPFig

Thanks Phillip! I tested a sample using MicroPoly(A)Purist™ Kit, 50ug total RNA as the input, this is how it looks when ran in the Pico Chip, the bioanalyzer calls a 3.5% rRNA contamination. It does look better than the RiboZero output, doesn't it?

Probably, but those are two different types of depletions. The RiboZero clearly did not remove the 18S. Your polyA results are consistent with either: (1) effective removal of rRNA or (2) degradation of rRNA into a trace no longer identifiable as such. You won't know which until you sequence it or test it with qPCR. That 3.5% figure is obviously completely meaningless. You could have zero % rRNA or 100% rRNA (but degraded) and Agilent would have no idea.

--
Phillip

[rndouglas]

Talked to Epicentre today and learned that my traces are exactly what I should expect to see. Wish I had just called them a couple months ago! Full speed ahead with library construction now...

[whw]

I saw a similar loss of mRNA with the RiboZero Human/Mouse/Rat kit. I wasn't too concerned with the loss of Bioanalyzer peaks but when I looked for hGUS mRNA by qPCR I found a 10-13x fold reduction. Any thoughts on this?

[pmiguel]

Originally Posted by whw

I saw a similar loss of mRNA with the RiboZero Human/Mouse/Rat kit. I wasn't too concerned with the loss of Bioanalyzer peaks but when I looked for hGUS mRNA by qPCR I found a 10-13x fold reduction. Any thoughts on this?

This really isn't enough information about your experiment to comment intelligently on.

Do you want to tell us 10-13x fold reduction means? Taken on face value it would mean that the absolute concentration of hGUS in your post-ribozero sample is ~10x lower than in your pre-ribozero sample. That could mean anything, though. It could mean your pre-ribozero RNA is in 200 ul whereas your post-ribozero is in 20 ul. Or they are in the same volume and therefore 90% of your mRNA is lost during ribozero isolation.

Or you could mean something else entirely. Please, if you are going to post here, give enough information to make it useful to answer you.

--
Phillip

[whw]

Yup, sorry about that was in a bit of a rush yesterday.

I started with about 800ng of total RNA in 28uL of water and performed RiboZero as written. My final volume after concentration was 40uL which accounts for a small dilution.

I used nanodrop, bioanalyzer, and qPCR for the housekeeping gene hGUS to assess the quality of my ribo-depleted RNA. Nanodrop was very low between (0.5ng/uL and 1ng/uL) which was fine, but my qPCR showed a 3+ cycle difference between pre- and post-depletion samples. That should correspond to about a 10-fold drop in the amount of hGUS mRNA present. It sounds like some people had a similar experience with the RiboZero plant kit. Bioanalyzer trace post-depletion was flat except for a very small tRNA peak.

[pmiguel]

Originally Posted by whw

I started with about 800ng of total RNA in 28uL of water and performed RiboZero as written. My final volume after concentration was 40uL which accounts for a small dilution.

What method did you use for concentration?

When we tried using ethanol precipitations it was hit or miss as to whether we recovered any RNA or not. (Which, by the way, is pretty typical for ethanol precipitations of RNA/DNA at low concentrations.)

When we switched to Zymo columns for RNA concentration, things were much more stable.

Also, did you do a dilution series on your qPCR? Could be you brought down something that inhibits qPCR, or the qPCR fluor.

--
Phillip

[whw]

I've tried both Zymo columns and ethanol precipitation with similarly disappointing results.

I did run a dilution series on the qPCR but there was no sign of inhibition.

Thanks for your help, I just spoke with RiboZero and they sent me a replacement kit. I'll see if that fixes anything.