With their ability to post-transcriptionally regulate gene expression, microRNAs (miRNAs) are claiming an ever-greater share of attention in the quest to understand biology and conquer disease. Aberrant expression of these highly conserved, ~22-base genetically encoded nucleic acids have been implicated in a host of disorders. In part because they are secreted into circulation, relatively stable, and can be found in a variety of biofluids, miRNA “signatures” have been proposed as biomarkers for disorders ranging from cancer to diseases of the central nervous system.

While miRNA has traditionally been investigated using RT-qPCR and microarrays, these technologies “are for analyzing current known miRNA species,” and not so much for discovery, points out Xinkun Wang, director of the NUSeq Core Facility at Northwestern University. “With miRNA sequencing [miRNA-seq], if there is something new you should be able to find it.”

While next-generation sequencing (NGS) of miRNAs is in its essence the same as sequencing any other RNA, some of the steps involved in generating the input material do differ. This article will look at some of the latest methods being used to prepare miRNA-seq libraries.

Not Your Messenger’s RNA

Current library prep and PCR technologies are good enough that it’s no longer necessary to enrich a sample for small RNAs (smRNAs), and cores are asking for total RNA, says Jonathan Shaffer, associate director, NGS assay product development at Qiagen Sciences. The biggest requirements are “just a clean RNA prep” and no heparin (which is “a very potent RT inhibitor”).

But it is important to use an RNA extraction protocol that retains the smRNA—standard protocols can be modified with higher salt precipitations to make sure the smaller RNAs can be recovered, says Lutz Froenicke, manager of the genome and biomedical sciences facility at the University of California Davis. For customers preferring kits there are a host of good choices—he recommends those from Zymo Research “because they’re inexpensive, and they work. And what I tend to recommend for situations where the input is very low are the Norgen Biotek kits.”

Messenger RNA library preparation typically begins with fragmenting the (relatively) long transcripts of RNA, followed by random priming to create cDNA by reverse transcription (RT). Since miRNA are so small, fragmentation is not necessary. And “if you worked with random priming, the random priming would happen somewhere in the middle of the molecule—leaving you with about 10 informative bases,” explains Froenicke. “For miRNA you need to sequence the whole thing.” For that reason most miRNA library preps ligate 3’ and 5’ adapters directly at the RNA level —prior to RT —at the same time creating a template to which the primers can bind separate from the miRNA to be sequenced.

Adapter Dimers

Yet the adapter ligation strategy for creating a miRNA-seq library poses problems as well. Among these is that 3’ and 5’ adapters can form dimers with each other. An overabundance of dimers can require deeper (and costlier) sequencing of the library to get the same information, lower-abundance species may not be sequenced, or the entire run may even be lost. “The instruments need to have a very diverse library in there to be sure that they can differentiate clusters and give you the correct sequence back, and adapter dimers are all going to be the same sequence over and over,” says Marta Gonzalez-Hernandez, scientist II at Takara Bio USA.

Several strategies have been employed to prevent or get rid of primer dimers. Perhaps the most obvious is size selection, which doubles as a clean-up step to get rid of other contaminants as well. Size selection can be accomplished in at least three ways. Gel electrophoresis followed by manual excision of the appropriately sized band is the traditional approach, but “can be quite laborious, particularly if you’re preparing libraries for lots of samples,” notes Praveen Sethupathy, associate professor of biomedical sciences at Cornell University.

Sage Science’s Pippin Prep automated DNA size selection system “works very nicely and reliably, but costs at least $20,000” and so is something you would only find in a core or in the lab of somebody who is using it all the time, says Froenicke. The third method, bead-based selection, doesn’t have the resolution necessary to distinguish adapter dimers from library with inserts, but beads are fast and useful for clean-up, and thus included in or recommended by several kits.

The large majority of people doing smRNA-seq are interested only in miRNA.

Several available kits prevent adapter dimers either by including a step to remove excess 3’ adapters before addition of 5’ adapters, or by using chemically modified adapters that block dimer formation.

Another option is to not utilize adapter ligation in the first place. At least two vendors — Takara and Diagenode—offer kits that use a combination of poly-adenylation, reverse transcription, template switching, and PCR to create an RNA-seq library with the appropriate barcode adapters.

Ligation Bias

The Takara and Diagenode strategies are also designed to “mitigate the adapter ligation bias that has now been well appreciated,” says Sethupathy, who has not had a chance to evaluate them but is very interested in doing so.

Another way to mitigate bias is, like Bioo Scientific’s NEXTflex® Small RNA Sequencing Kit, to use a pool of adapters “with more or less the same sequence but with randomized bases at the very ends where ligation will occur,” explains Sethupathy, who says the kit’s resulting library correlates very well with qPCR. “What you’re really doing is diluting the effects of adapter ligation bias across the entire pool of miRNA sequences.”

Qiagen has “integrated UMIs [Unique Molecular Indexes] into our kit, and that can allow you to understand and deal with any biases either during amplification or sequencing,” notes Shaffer.

smRNA

The large majority of people doing smRNA-seq are interested only in miRNA, says Wang. But there are other classes of smRNAs that also have roles in different biological processes, and sometimes people want to specifically look for those.

It’s important to be aware of how the prep kit’s chemistry works with your RNA species-of-interest. For example, many smRNAs have modified bases or protective caps that pose a challenge to ligation-based protocols.

Biocompare’s Library Prep Search Tool
Find, compare and review library prep
tools from different suppliers Search

The flip side, of course, is that using protocols designed to pick up all smRNA means contending with a host of contaminating species (if you’re looking interested in miRNA). The good news is that it’s relatively straightforward to sort that out bioinformatically after the fact.

Image: Shutterstock Images