Library prep for Illumina RNA-seq with 0.5-3ng total RNA

[DarkQ]

Hi All,

I'm new to the forum so I'm sorry if I missed similar questions asked about low RNA input for library prep.

I'm trying to do expression analysis on around 10k mouse bone marrow cells, where I can purify about 0.5-3ng of total RNA using RNeasy plus micro kit in 12ul of RNase water. The bioanalyzer results look good for all samples with RIN>8.

Our lab used to use the Nugen Ovation RNA-seq system V2 to RT and amplify cDNA (>500pg according to the kit), but the problem is over-representation of rRNA reads in the final sequencing results. Also it seems a lot of waste when I amplify to get micrograms of cDNA and only use 100ng for the library prep (according to the next library prep kit from Nugen).

I recently found a kit from Clontech (SMARTer Stranded Total RNA-Seq Kit - Pico Input Mammalian) that requires only 250pg-10ng total RNA input to generate sequencing directly, without amplification steps in between cDNA and adaptor ligation (only 5 cycles for adaptor PCR ligation). It seems to be a more cost-effective option and I should in principle avoid the amplification bias. I'm wondering if anyone has experience with this kit or how it is compared to the Nugen kit.

Thanks!

[NextGenSeq]

How much rRNA? We typically see about 25% if the RNA is not degraded.

[jteeee2]

We have had a similar experience with the Nugen Ovation RNA Seq system V2. For good quality material, the Clontech SMARTer kits work quite well. We routinely use the SMARTSeq Ultra Low input RNA Seq kit and have been very pleased with the results of inputs down to 100 pg and as high as 1 ng. This kit does not maintain strand specificity, though. If this is important to your experiment, I would try out the SMARTer Stranded Total RNA-Seq Kit. We have been meaning to do a direct comparison between this and the Nugen Ovation kit, but just haven't had the free time to do R&D. It would be great to see these two kits head to head. Best of luck!

[DarkQ]

Thank you guys for the reply. Right now I'm proceeding with the ovation kit because my PI thinks it's not worth it to try a completely new kit. However I'm still facing some technical problem with the Nugen kit.

I started with about 600pg of high quality RNA (RIN>9 purified with RNeasy plus micro kit and on-column DNase digestion) and did RT and amplification using the Ovation RNA-seq V2 kit (7102). I began with 4 samples just to test if the kit works in my hand. From 600pg total RNA I got 3-4ug of cDNA consistently across 4 samples measured by nanodrop (260/230 >2.3, 260/280=1.9). This yield seems to be OK according to the kit manual, although it's about half as compared to the yield obtained by my labmates, who did neural stem cells with similar or less RNA to start with. When I measure yield by Qubit dsDNA HS kit, it gave me half the concentration as nanodrop consistently across samples. As a result it is lower than the yield suggested by the kit (1.8-1.9ug as compared to 2.5ug according to kit manual). I was told by Nugen people that this difference is due to presence of ssDNA.

Afterwards I tried to QC my cDNA by doing some qPCR. Actin and GAPDH looks good and almost equal across samples. However when I check some low abundance transcripts (7-10 Ct value higher than actin), 2-3 cycles of difference starts to emerge even between biological replicates, and some transcripts failed to amplify in certain samples (2 transcripts failed in 2 different samples). I used the same primers before on these similar samples (not amplified) and they normally gave quite similar Ct values. The transcripts of the knockout gene was gone as expected.

I'm wondering if I should proceed or I should continue trouble shooting. The kit is not cheap so everytime I try it costs me 100 dollars per sample. The nugen people couldn't provide any more information other than what was suggested by the kit manual.

Thanks!

[jteeee2]

Hi DarkQ-

Your yield of 3-4 ug from 600 pg RNA sounds very reasonable to me. One thing I would mention is that we've found purifying the post-SPIA product with Qiagen MinElute Reaction Cleanup columns is the best way to go. Also, measuring the product with the HS DNA Qubit assay will most definitely give erroneous readings due to the ssDNA. I generally just do a BR DNA Qubit reading which gives me a ballpark concentration before heading into shearing/library prep. Your Nanodrop reading is likely a pretty good assessment of quantity as well. Do you have access to a Bioanalyzer? If you haven't done this workflow before, I would recommend running out your cDNA on a fragment analyzer of some sort to see that you are getting an expected size distribution similar to the one Nugen has in their kit handbook.

[DarkQ]

Hi Jteee2,

Thanks for your reply! Yes I used Qiagen cleanup column to purify. I wanted to use Qubit HS assay because only dsDNA would be useful for adaptor ligation during library prep right? Anyway I'm probably paranoid since it's such an expensive experiment. I was worried mainly because the lower yield as compared to other people with similar input, as well as variations in qPCR.

I have access to a tapestation and I can run a D1000 tape to QC. One thing I'm not sure is whether the size distribution is cell-type dependent. Anyway I should expect similar pattern across my samples.

[jteeee2]

You are correct that the Qubit assay is the best way to determine input into library prep. However, I've found that using the Broad Range DNA Qubit assay gave me more reliable results for this type of cDNA product compared to the High Sensitivity DNA Qubit assay. I have had really good success taking 500 ng of sheared cDNA into the Kapa LTP DNA library prep kit. I've recently started to switch over to using the Kapa HyperPlus protocol (uses enzymatic-based shearing) but don't have any sequencing results to share.

I'm not sure what to make of your variable qPCR data. Generally, if we see consistent cDNA traces post SPIA, the sequencing data is very good.

Comparing your yields to other people is not always wise. Your initial quant of the RNA could have been more or less accurate than the previous folks, but you can't really be sure unless you have some of their leftover material. A consistent cDNA yield across different samples of equivalent quality is a better metric.

[DarkQ]

That's a relief. The main reason I wanted to check qPCR was some problems I had last time where cDNA yield was similar according to nanodrop, but Ct values for actin/GAPDH were everywhere. I kinda figured out the problem last time, which was due to the salt contamination leftover in RNA extraction because I didn't use column purification and the DNase buffer leftover was inhibitory to RT according to the manufacturer. Anyway since I'm using column purification and the Ct values for actin/GAPDH were essentially equal across the board, I'm much more comfortable to proceed I guess. I'll post the update how it goes afterwards.