阵面面观微矩

发布时间：2025-05-04 17:10:58 来源：行不胜衣网 作者：时尚

microRNA微矩阵面面观

If PubMed is 微矩any guide, the microRNAfield is smokin' hot. Of the nearly 11,100 references that come up in asearch for "microRNA," nearl

If PubMed is any guide, the microRNA field is smokin' hot. Of the nearly 11,100 references that come up in asearch for "microRNA," nearly two-thirds (64%) were published since 2009.

It's no wonder. Although microRNAs (miRNAs) may be small—they average 22nucleotides in length—they carry a big stick, biologically speaking. "We can say with confidence that over 60% of human protein-coding genes are conserved targets of miRNAs," wrote David Bartel and colleagues in 2009. [1]

To date, some 16,772 miRNAs have been discovered and logged in miRBase, including 1,424 in humans.The question for researchers probing these molecules' biology is, whichmiRNAs are active under a given set of experimental conditions, and howdoes that pattern change in the dynamic cellular environment?

Researchers basically have three strategies for assessing microRNA expression: quantitative RT-PCR (qRT-PCR), microarrays and next-gen DNA sequencing.For quantification of a few known miRNAs in a large number of samples, qRT-PCR is the preferred strategy. Sequencing is best for unbiased quantification in a relatively small number of samples, as well as for discovery of novel RNAs. Occupying the middle ground are microRNA microarrays, which can profile the entire universe of known miRNAs in a single shot yet are inexpensive enough to be applied to a large sample set.

Staying current

A handful of companies offer miRNA microarrays, including Affymetrix, Agilent Technologies, Exiqon and Life Technologies. LC Sciences also produces a miRNA microarray but offers it as a service (that is, users submit samples for processing at the company).

One key differentiator among these different platforms is their content."Because it's a fast-evolving field, you want to get the most information from your samples," says Christoph Eicken, head of technicalservices, microarrays, at LC Sciences.

That means considering which version of miRBase is represented on the array. The current iteration of miRBase, version 17, was released on April 26, 2011, and contains, according to the miRBase blog, "more than 400 new human and mouse sequences." LC Sciences, which builds its microfluidic arrays on demand (that is, they are not prefabricated), hasswitched over to the new content already, says Eicken.

Agilent and Exiqon offer microarrays reflecting miRBase, version 16, released in September 2010. Henrik Pfundheller, senior vice president, sales and marketing, at Exiqon, says it typically takes the company "four to eight weeks" to update its content to reflect a new miRBase release. Similarly, Agilent can update its content "within weeks," says Heidi Kijenski, marketing director for microarrays.

Affymetrix's GeneChip miRNA 2.0 Array is based on miRBase, version 15 (April 2010), and Life Technologies' NCode Human miRNA Array V3 is basedon version 10.0 (August 2007). (Only Exiqon's miRCURY LNA microRNA Array, 6th generation, is based on miRBase, version 16; miRCURY LNA microRNA Array, v.11.0 - Other Species takes its content from miRBase, version 11.)

Pre-fabricated arrays also may contain additional small RNA content not necessarily found in miRBase, such as precursor-microRNA sequences (i.e., not mature miRNAs), small nucleolar RNAs and so on.

But don't let lower miRBase versions fool you. John Keefe, senior product manager, strategic marketing-RNA, at Affymetrix, notes that because miRNAs are highly conserved evolutionarily, in some cases a "newly discovered" miRNA from one species may not actually be novel at all, in that it had been discovered previously in another species.

Thus, Keefe says, wide species coverage on an array can sometimes help overcome a dated miRBase release. The GeneChip miRNA 2.0 array contains the entirety of miRBase on the chip, representing 131 different species.(In contrast, other arrays are generally more species-specific; for instance, Agilent's off-the-shelf arrays are available for human, mouse or rat. The Exiqon 6th generation array covers human, mouse andrat, as well as associated viral miRNAs.)

That said, Eicken recommends users stick to species-restricted arrays ifthey can. It's a matter of relevancy, he says. "Especially as a first-time user, you might be impressed by arrays that offer a very large amount of different probes that are not from your species . . . because it sounds so great to have all those probes," he says. "But whatinformation do all those probes give?"

Sensitivity and specificity

Two key variables when selecting a microRNA-microarray platform are sensitivity and specificity.

Sensitivity, says Pfundheller, "is a description of how little total RNAyou can use and still detect a signal." On Exiqon's arrays, he says, 80% of miRNAs, including AT-rich ones, are detected even below 1 attomole, making the array work with as little as 30 ng total RNA. Agilent's arrays can use 100 ng total RNA, detecting as little at 0.1 attomole, says Petula Dandrade, a scientist at the company.. And Affymetrix can detect 85% of all miRNA species on its array with just 1 attomole of input material—about the equivalent of 10 copies per cell in1,000 cells, says Keefe.

Specificity is a reflection of the assay's ability to distinguish two highly similar, but not identical, sequences. A classic example, says Keefe, is the let-7family of RNAs. "A lot of family members differ only by one base, often on the 5' end or the 3' end," he says, making it exceptionally difficult to tell the different members from oneanother.

These variables become especially important when working with precious samples (such as samples obtained via laser capture microdissection) or degraded samples, such as those obtained from formalin-fixed paraffin embedded (FFPE) tissue sections.

A related variable is the melting-temperature profile of the array's probes; variable melting temperature, or Tm, can lead to false measurements. "Because miRNAs are very short sequences, it's important that these probes are Tm-normalized," says Eicken.

LC Sciences Tm-normalizes its arrays, Eicken says. So does Exiqon. According to Pfundheller, the melting temperature for standard DNA/RNA hybrid capture probes across all of miRBase varies "maybe over 30 to 40oC." But Exiqon doesn't use DNA/RNA hybrids; the company builds its capture probes using Locked Nucleic Acids (LNAs), enabling it to tweak the physical properties of its oligos to produce a Tm-normalized array (without affecting the probes' lengths). "Those [oligos] that are AT-rich will have higher LNA content than those that are GC-rich," Pfundheller explains.

Custom content

Overall, miRBase represents over 150 different species. Not all of them have their own microarrays.

For researchers who study such species, custom microarrays are sometimesavailable. For instance, researchers can build custom Agilent microarrays through the company's eArray web tool (both 8x60K and 8x15K formats are available)—but they can only select probes that are already in miRBase (that is, users cannot add novel miRNAs discovered in their own labs, except through a collaboration with the company).

LC Sciences offers its users a service called SeqArray, in which the company first runs transcriptome sequencing on submitted samples to identify miRNAs and then designs custom microarrays to screen for those transcripts in larger sample sets.

Alternatively, for those interested in just a small number of miRNAs, Pfundheller suggests considering qRT-PCR. "Either use one of the pre-validated assays, or if you have found 20 miRNAs, why not custom design 20 primer sets and simply analyze your samples using those 20 sets of qPCR primers?" he says.

qPCR arrays

That's not to say qPCR can be used only for small numbers of transcripts. Both Life Technologies and SABiosciences, a Qiagen company, offer qPCR-based arrays for broader miRNA analyses.

"Since [there are] only about 1,000 targets, that plays well with other platforms like high-content RT-PCR," says Iain Russell, senior product manager, next-gen qPCR business unit, Life Technologies. "We've seen tremendous interest in our customer base for using TaqMan®-based solutions for microRNA profiling."

With that in mind, Life Technologies designed TaqMan Assays for all of miRBase: some 15,000 assays. The company then collected some of the mostpopular for human, mouse and rat, and packaged them into pools for researchers interested in using the technology to screen several hundredmiRNAs simultaneously. Alternatively, researchers can submit their own sequences to Life Technologies for TaqMan assay design. "We have taken all the learning of the past five years, wrapped that into an automated design pipeline, and now customers have access to that," Russell says. Either way, the company's TaqMan Array MicroRNA Card and TaqMan OpenArray® MicroRNA Panel support low- and high-throughput analyses (respectively) of up to 754 targets per sample, Russell says.

Using the TaqMan OpenArray platform, researchers can run three plates (3,072 33-nl wells per plate) in parallel, each profiling three samples,in about two hours. That means in a typical work day a researcher couldprocess 36 samples. (Life Technologies released its TaqMan OpenArray Human miRNA panel in April 2011; a rodent panel is anticipated shortly, says Russell.) As a result, says Russell, scientists can burn through a couple hundred samples in a week or so, whereas using the TaqMan Array Cards, "it would take several weeks to do that."

Final thoughts

After researchers determine that a given miRNA is alternatively expressed under, say, diseased conditions, they next want to discover what transcripts that miRNA might be regulating. "That's probably the biggest challenge in the field," says Eicken. At least some miRNA-microarray software packages offer this capability. For instance, Agilent's GeneSpring GX software includes a TargetScan module for this kind of pathway analysis. Alternatively, there's LC Sciences’ free web site, miRFocus.

First, though, they'll need to select a gene expression platform. And when doing that, perhaps the biggest variable to consider is existing infrastructure. If a lab (or core facility) already has, say, an Affymetrix GeneChip Scanner, its users will be restricted to using Affymetrix GeneChip microarrays (unless they choose to outsource their work to a colleague or service provider).

Whatever platform a lab ultimately chooses, says Eicken, one thing should never change: the RNA extraction protocol. "While pretty much all[extraction] kits are good, the profile you extract from your sample will differ a little bit [from kit to kit]. So to minimize the variation, stick to one extraction method."

Reference:

[1] R.C. Friedman et al., "Most mammalian mRNAs are conserved targets of microRNAs," Genome Res, 19:92-105, 2009.

The image at the top of this article is from Agilent Technologies' microRNA overview.

相关文章