RNA Amplification Kits

Whether you are using it as a tool or an object of study, you may require significant quantities of RNA for your project—often far more than can be obtained from sources such as laser capture microdissection (LCM), needle biopsy, and primary cultures. Thus, it’s often necessary to amplify the RNA that is at hand, making enough copies to use as probes for Northern blots or ribonuclease protection assays, to create libraries, to inject into tissues or animals, or to hybridize to microarrays.

What was once an onerous task, requiring baked glassware, DEPC-treated reagents, and dedicated work areas, has become a relatively simple task. There are a host of kits available for amplifying RNA—some fairly generic, others designed for more specific applications. “You can essentially open up a box and say, ‘Oh, I’ve got everything I need’,” notes David Kerry, product manager for gene expression at Stratagene (which was purchased by Agilent in June). With simple things like RNAse-free, barrier tips and good, standard lab procedure, he adds, “it’s more error-proof now than it used to be way back in the day.”

In their essence, most (if not all) commercially available kits amplify in the same, relatively straightforward way. The process involves an initial reverse-transcription (RT) step during which the source RNA is copied to form a cDNA intermediary. Following this RT step, the cDNA is then transcribed back into RNA. Thousands of RNA transcripts can be sequentially produced from a single cDNA strand during the in vitro transcription (IVT) step. In the case of Invitrogen’s kits, for example, it’s actually the complement of the original RNA—termed aRNA, for “antisense RNA”—that’s produced, notes Jason Johnson, Invitrogen’s senior product manager for DNA microarrays.

Differences among kits—and thus their utility for any given lab’s specific needs—are found in a host of areas. Enzymes, for example, both native and engineered, may affect the length or quantity of aRNA produced—as can whether multiple amplification steps are called for. Primers may affect what type(s) of RNA are amplified. The format may determine how many samples can effectively be amplified at a time. Incorporated labels can determine how they are to be assayed. And so forth.

It's not the purpose of this article to rank the kits that are on the market, to give a listing of their claims or counterclaims, nor to try to evaluate individual products. And because each consideration described below can, to some extent, be applied independently, they should not be thought to constitute a decision tree. Rather, this article will give you some ideas of what to look for when assessing whether a particular kit might meet the needs of your lab.

Kits using spin columns, for example, are generally designed to eliminate unbound primers and other oligonucleotides. “Essentially, you’ll lose RNAs under 200 bases in length,” notes Kerry. And even if they do make it though the spin column, most kits won’t prime shorter oligos. Kits designed specifically to amplify microRNAs are available, as are kits that will ligate a tag onto shorter RNAs to allow priming.

Leaving small oligos aside, the field is essentially split between kits that will amplify the whole transcriptome using random primers, and those that focus only on mRNA by using oligo dT primers, notes Robert Setterquist, senior manager and scientist of research and development, for Ambion Inc., an Applied Biosystems business. Random primers will amplify all the RNA in the sample—including that from messenger RNA (mRNA), transfer RNA, ribosomal RNA, and other non-mRNA. The resulting transcripts exhibit a bias toward the 5’ end of the source RNA, since transcription can begin anywhere on the cDNA. On the other hand, oligo dT primers hybridize with the 3’ poly-A tail of mRNA, lead to aRNA mostly from mRNA, and exhibit a 3’ bias.

IVT is a linear method of amplification, so the aRNA product should exhibit the same ratio as the source, explains Johnson. “It should give you a very accurate representation of what the original mRNA was.” This is as opposed to subjecting the cDNA to PCR amplification, which will preferentially amplify shorter products, as well as those that are in high abundance in the genome.

Microarrays are the largest—but not the only—consumer of amplified RNA, notes Setterquist. “And the large majority of those would be for GeneChip analysis using Affymetrix GeneChips®.” Expression studies such as these use oligo-dT priming, and will often incorporate labeled nucleotides—thus it’s important to know whether your application requires biotin or dyes like Cy3 or Cy5, for example. The latter, says Setterquist, is popular among the “self-spotters” and “homebrew microarray” community, as well as users of Agilent microarrays. Many companies offer kits targeted to specific platforms.

The market for whole transcriptome RNA is growing, he adds, and the principle application is Affymetrix’ exon arrays. For these, a random priming technique is used to query splice variants that would not be as easily detectable if only poly-A priming were used.

Some users amplify RNA from a small number of cells (using a random priming approach) to construct libraries for sequencing or cloning, Setterquist says.

Even arrays that use labeled cDNA—rather than RNA—make use of RNA intermediates.

Applications that require amplification of smaller RNA—siRNA or microRNA, for example—probably need a kit specifically designed for such applications. According to their website, “the only commercially-available kit for global amplification of small RNAs, including microRNAs for expression studies of small tissue and cell samples,” is made by System Biosciences (SBI). But Invitrogen offers the NCode miRNA Amplification System, which adds a poly-A tail and produces sense RNA. In lieu of these, there are kits for enrichment of small RNAs, as well as labeling kits and PCR- and microarray-based assays.

In addition to being subject to degradation, RNA from samples that have been fixed in formaldehyde and embedded in paraffin (FFP or FFPE) is often cross-linked to proteins and to itself. Reverse transcriptase has great difficulty reading across such cross-links. Yet even if some of the RNA could be read some of the time, because it’s often not known what protocol was used to fix the samples, or whether each sample was identically prepared, the data may not be meaningful.

“Purity is absolutely key,” argues Kerry. “If you’ve got lousy RNA to start with, you may not be able to count on the results.” He recommends that RNA be both quantified and qualified—using instrumentation such as the Bioanalyzer, a microfluidics-based analysis tool manufactured by Stratagene’s parent company, Agilent—prior to its use in any downstream application.

There are suppliers, such as NuGEN, that make kits specifically for FFPE RNA. Yet many manufacturers simply don’t recommend them. The variation found may end up being due more to sample preparation than to the biology itself, cautions Setterquist.

“Is this the type of experiment in which you would isolate samples every week?” asks Setterquist. “Or can you do it all at once—one experiment where you isolate your RNA after some treatment, and all the RNA is isolated in one day, followed by amplification and arrays? Or is this an ongoing effort?”

One or a few samples are probably best accommodated by spin columns, points out Kerry. Yet the amount of time to prepare and run 500 samples would be virtually impossible logistically using spin columns. The cost of consumables, too, becomes an issue as the number of samples increases.

Several manufacturers offer RNA amplification kits designed to accommodate large numbers of samples. Ambion, for example, has a kit in which samples can be processed in a 96-well plate using magnetic beads. “It makes it much easier than if you were going to do manual column processing,” notes Setterquist.

But, says Kerry, while cell culture is amenable to high-throughput methods, whether manual or robotic, many other types of samples are not: “Tissue is limited what you can do,” he cautions. Even for large numbers of samples, “you’re still going to have to use old-fashioned spin columns.”

Just as kits are available to process microRNA, total RNA, and FFPE RNA, manufacturers also market a variety of kits designed to make the most out of whatever small amounts of starting material are available.

Many tout kits able to amplify RNA from as little as 10 cells (like Arcturus’ RiboAmp® HS RNA Amplification Kit) or 100 pg (like Ambion’s MessageAmp™ II aRNA Amplification Kit). Epicentre Biotechnologies even claims to be able to “produce microgram amounts of aRNA (cRNA) from 10 pg (1 cell) of total RNA using the TargetAmp™ 2-Round aRNA Amplification Kits.”

How much starting material is necessary may depend on the downstream application that the RNA will be used for. “What’s the realistic lower specification for input amounts of total RNA into each of the kits?” asks Setterquist. “We’ve shown consistently that we can go down to 25 ng of total RNA with the [Ambion] Illumina® TotalPrep™ Kit.” For Affymetrix platforms, on the other hand, while the kit recommends an input level of about 1 microgram RNA, “we can typically do 50-100 ng,” he notes. “That has to do with both the kit and their array hybridization technology.”

The choice of label will affect the yield of the amplification process, warns Setterquist. “That just has to do with the inhibition properties of the nucleotide and the RNA polymerase.”

Other factors affecting RNA amplification kit yield may include the amount of input RNA, the length of time the IVT is allowed to proceed, and how many rounds of amplification the RNA undergoes. For example, the RiboAmp® online literature notes that, “messenger RNA is amplified up to 1,000-fold in one synthesis round and up to 1,000,000-fold in two rounds.”

All outputs—even those outputting similar amounts of aRNA with the same label—are not necessarily the same, though. Enzymes, buffers, and other conditions can affect the fidelity and the length of transcript that is produced. Kits designed for the highest fidelity and longest transcripts will likely tout that in their literature. Whether this makes a difference in results will depend on the specifics of the downstream application.

Another consideration that may be important (depending on the downstream application) is whether the resulting amplified RNA is sense or anti-sense. Many kits—especially two-round kits—produce RNA identical in orientation to the starting material.

For ongoing efforts, the number one suggestion Setterquist offers is: “Settle on a standard operating procedure which would include the kit, how you’re going to perform it, who’s going to perform it, the whole deal.”