Tips for RNA precipitation with linear acrylamide

[Speese]

Hi all,
I am new to the forum, so I will make this my introduction and first question to the forum. I am currently working with the TU-tagging method (TU-tagging: cell type specific RNA isolation from intact complex tissues, Miller et. al. 2009) to isolate RNA from cells in the Drosophila CNS. The protocol has been working when assessed by real time PCR, however there are about 3 times during the isolation of the "tagged" RNA that one has to clean up the RNA. The workflow looks like the following:

1) Isolate total RNA via Trizol and column clean-up or linear acrylamide precipitation -> 2) Isolate mRNA with Dynabeads -> 3) Label Thiol tagged RNA via Biotin-Thiol coupling -> 4) Precipitate RNA -> 5) Purify biotin labelled RNA with magnetic streptavadin beads -> 6) Precipitate RNA again -> 7) Proceed to cDNA amplification step (NuGen) and ultimately on to Seq. library preparation

I cannot afford to loose RNA at each of the clean-up steps so I have decided to move away from minElute columns. In my hands, I find the columns are a little inconsistent and sometimes I will lose up to 50% of my RNA. Thus, I have settled on linear acrylamide, as it supposedly can recover almost 100% of the input RNA and is compatible with all following downstream steps.

I have a few questions regarding the use of linear acyrlamide assisted precipitation

1) In this protocol I have to precipitate the same RNA sample about 3 times in succession. Do I have to add linear acrylamide at all the precipitation steps, or just the initial step? Each step will be precipitating the RNA from roughly the same total volume.

2) Does anyone have any hints for how to not loose/visualize the pellet of RNA? I can rarely see my pellet and on the last step where I isolate the biotin labelled RNA, I may only have 5 ng of RNA (maybe less), so the pellet is going to be really hard to see.

Thanks

Sean Speese
Research Assistant Professor
OHSU - Jungers Center for Neurosciences Research

[protist]

Hi Sean,

In a previous lab I frequently used LPA for low conc RNA precipitations of tRNA substrates and it worked very well - see old paper attached for comparison of carriers for precipitation. In my current position we generate lots of Illumina sequencing libraries. We initially used glycogen as a carrier and that worked fine until we started working with smaller and smaller amounts of starting RNA. For glycogen containing precips GlycoBlue can be useful option as it colours the pellet pale blue so it is easier to visualise . Another option is to use Pellet Paint NF (as suggested in the TruSeqTM Small RNA Sample Preparation Guide) - again this colours the pellet so you can see when you are mistakenly sucking it up in the pipette tip.

However for the last year for all our RNA precipitations we have switched over to Zymo RNA Clean&Concentrate 5ug columns instead of precipitation - it has made a big difference with regard to the yields. For our bacterial RNAseq libraries there are at least two instances of precipitation required - post rRNA depletion and post fragmentation. We did some head to head tests with other columns and precipitation and found the Zymo columns to be the most consistent with regard to minimising loss and maximising yield.

[Speese]

Thanks for the reply protist. I was hoping I would get someone who had compared different columns as I am not opposed to using them, it is just that the other brands I have used are inconsistent. Thanks for the Pellet Paint tip. I have used this product in the past, but I was told by the NuGen rep that I cannot use glycogen in the steps leading up to their cDNA amplification kits. I am pretty sure Pellet Paint has glycogen as the carrier, but maybe I am mistaken.

I am curious, what kind of average recovery did you get from the Zymo RNA clean and concentrate columns?

[kerplunk412]

I have been testing a few RNA purification methods for eventual library prep, so I will share my experience here. MinElute columns actually work very well (comparable to EtOH precipitation) if you increase the amount of ethanol you add to your sample from 250 to 550 ul. However, this will precipitate all sizes of RNA, so you will still have a lot of small stuff, but since I didn't see fragmentation in your protocol this probably doesn't matter much for you. Also, the Ribominus concentration module works well if you want to elute in a smaller volume (it has a much small membrane), and it also has yield comparable to precipitation. I also tired using AmPure XP beads and got good recovery as well as exclusion of small fragments- this is my favorite thus far, and you get to use magnets

If you must precipitate, I would use the suggested amount of LA (I think 10-20 ug) for the first precipitation and half for each successive precipitation. This is based on my experience with glycogen- when I didn't add it for the second precipitation the pellet was still visible but much smaller and not as compact, if I remember correctly. Also, I have tested whether a low-temp incubation is necessary, and if you spin for 40 minutes at 4 degrees after addition of EtOH there is no need to cool at -20 or -80 first. I know the protocols all say to still do this low temp incubation, but I have tested it multiple times and it really does not seem to help much at all, at least for RNAs 50 nt and larger.

My advice for your protocol is to use MinElute columns with a higher ethanol amount, and if having too many small fragments ends up being an issue substitute an Ampure bead purification. Precipitation is fine, but it will take longer and there are more things you can mess up, especially when working with a small pellet that doesn't adhere to the tube well.

P.S. The results I describe here are from running TBE-PAGE gels to check yield, not spec-based measurement, and using a low range (50-1000 nt) single strand RNA ladder as input material.