Each month, more and more evidence emerges to link microRNAs to a range of diseases. Despite all their progress in making miRNA-gene expression connections, researchers on the front lines of RNA research say they have only recently begun to understand how these tiny molecules — typically one one-thousandth the size of an mRNA molecule — actually function.
"The basic biology of microRNAs is still poorly developed, and although we have really seen an explosion in the study of microRNAs in the last two or three years, it takes longer than that to really have a good grasp of what the general functions of them are — and even longer to figure out the specific functions. So I think it's very early in the field," says MIT's Phillip Sharp, who was the co-recipient of the 1993 Nobel Prize in physiology or medicine. "Clearly, you can deliver single-strand nucleic acids or inhibitors of single-strand nucleic acids [to] cells in a clinical setting. I think there are open avenues and it will just take awhile to work it all out."
Moving miRNAs into the clinic has gained momentum since 2005, when researchers at Rockefeller University used modified single-stranded RNA analogs, or antagomirs, to target specific miRNAs. During the last few years, a small but growing number of miRNAs have been incorporated into pharmaceutical companies' drug-development pipelines. Some miRNA researchers say they envision diagnostic roles for these molecules more than therapeutic ones. Late last year, Rosetta Genomics, which has brought three miRNA tests to the market, announced an improved version of its miRview Mets diagnostic tool, which the company says can identify 42 tumor types of unknown origin.
There are currently a handful of miRNAs in the sights of several pharmaceutical companies. One miRNA target that is ripe for clinical development is miR-21, which is thought to be up-regulated in cardiac and lung fibroblasts as well as in several different types of cancers. Regulus Therapeutics and Sanofi-Aventis have partnered up to establish methods for inhibiting miR-21 with antagomirs. With support from several studies demonstrating that miR-34 played a key role in the p53 tumor-suppressor network, Mirna Therapeutics began a program last year to develop a -miR-34 mimic that could be used to treat various cancers. Miragen Therapeutics has licensed the 2008 work of the University of Texas Southwestern Medical Center's Eric Olson — in which he linked miR-208 to heart disease — with the hopes of developing an miRNA-targeting drug for heart failure.
But the miRNA that has attracted the most attention for the clinic is miR-122. Rosetta Genomics, Santaris, Regulus Therapeutics, and Mirrx Therapeutics are all interested in this particular miRNA, which has been shown to play a role in replication of the hepatitis C virus. The only miR-122-targeting drug that has been tested in humans to date is Santaris Pharma's phase II hepatitis C therapy, miravirsen. It might not be the only one for long, though, as Regulus Therapeutics and GlaxoSmithKline teamed up last year to develop their own hepatitis C drug candidate that targets miR-122.
In the liver
As miR-122 has been shown to be expressed highly in the liver, diseases that affect this organ are attractive — and potentially the most tractable — prospects for miRNA-based therapeutics.
Sharp says that while many studies have shown that miRNAs play pivotal roles in tumorigenesis, targeting liver cells is easier than pinpointing cancer cells, so it's likely that diseases like hepatitis C will be the primary focus of early clinical studies. "There are clear examples of microRNAs being involved in the regulation of hepatitis C, and we're targeting a particular microRNA in model systems that has been shown to ... have the effect of down-regulating or suppressing the replication of hepatitis C," Sharp says. "I think there are a number of targets. We have ways of inhibiting microRNAs by using locked nucleic acids and antisense, and there's ways of increasing the level of microRNA by using what are called microRNA mimics in cells. There are animal studies that are ongoing, and I anticipate that there will be more human trials in the future."
With all of these pharma chomping at the bit and throwing miRNA drug targets at the wall to see what sticks, one might think that researchers know the ins and outs of miRNAs biology.
"For the last 10 years we've been identifying microRNAs and developing more efficient ways to find ones we missed earlier. That phase is now drawing to a close, but the realization that there were many microRNAs has brought up important questions, some of which will keep the field occupied for many more years," says the Whitehead Institute's David Bartel. "What are these little RNAs doing, and which messages are they targeting? What happens when they target a message? To what extent do they cause the mRNA or the protein [to]change?"
Bartel's lab was one of several that helped to uncover the abundance of miRNAs during the early days of the field by developing cloning and computational strategies for miRNA gene discovery in animals and plants. He and his colleagues also found hundreds of miRNA genes in the human genome. Over the years, they also learned more about how miRNAs regulate their targets and which mRNAs are targeted by miRNAs. While it is clear that most mammalian genes are targets of miRNAs, scientists still know little about the particular impact of miRNA-target interactions. "We've developed ways of predicting which messages are going to be targeted by a particular microRNA — those methods are useful but not perfect. Some messages that we think should be strongly regulated by the micro-RNA are indeed regulated, but there are others that are not regulated so well, and part of that is because the UTRs are not annotated so well," Bartel says. "We're developing methods to more accurately find out what the UTRs are, so we can find out which mRNAs are being regulated by the microRNAs. Then we look into particular microRNA-target interactions to see why a particular interaction has been conserved for so long in evolution — what is the biological reason that the microRNA is targeting that gene. We do this primarily in plants and mammals, but also have been working with worms, flies, and zebrafish."
One of the first ports of call to getting a firm handle on the biology of miRNAs — which will in turn broaden the horizon for miRNA-targeted therapeutics — is nailing down their functions. Despite several exploratory efforts using model systems, the paltry amount of functional data currently available is something many investigators in the field cite as a major hurdle. Norbert Perrimon, a professor of genetics at Harvard Medical School, is generating reagents to create generic flies that express "miRNA sponges" — transcripts expressed from promoters containing multiple tandem binding sites to a specific miRNA — for each miRNA of interest, so that they can perform systematic loss-of-function studies.
"We have some pieces of information about the biology of some of those microRNA, but it's still very few. If you look in fly, for example, you probably have more than 150 to less than 200 microRNAs — and I would say that we really only know the function of 10 or more maybe — so there's still a lot of biology to be extracted from those," Perrimon says. "A few labs are generating reagents to do loss-of-function and get a functional analysis for microRNAs and I think it's, again, not easy. In most cases, there are not obvious phenotypes — you don't really know what to look for."
Part of what makes the task of obtaining clinically relevant functional data for miRNAs difficult is that traditional gene knockout approaches aren't effective for this purpose. For example, the let-7 family of miRNAs has 12 human homologues, so when researchers have attempted to infer their functions by doing double- or triple-knockouts in mice, they would sometimes observe no phenotypic effects — the animals would remain alive and well.
"We're knocking out members in let-7 as well as the lin-4 families in combination, and we're finding that there are a lot of redundancies across miRNA families, so there [are] these huge, complex regulatory networks that are occurring within microRNA families to control development processes as well as [other] processes in the adults," says Aurora Esquela-Kerscher, professor at the Eastern Virginia Medical School.
Esquela-Kerscher is currently using C. elegans to understand how the let-7 family works in conjunction with the lin-4 family to affect the cascade of events associated with gonad formation and cell-fate determination, both of which have implications for human cancers. She and her colleagues are looking at how all human miRNAs — including lin-4 and let-7 family members — contribute to urothelial cancers, like prostate cancer, and how those miRNAs could be used as biomarkers. "When you think therapeutics, people are quick to jump into the next primate model or go into the safety trials, but we don't really have a handle on the huge range of targets that a micro-RNA can actually regulate," she says. "What's lagging in this field is the functional data."
So far, the literature has shown that several mi-RNAs are dysregulated in certain cancers; they've been labeled both tumor suppressors or oncogenes, depending on their context. And while both roles have been supported by in vitro and bioinformatics data, Esquela-Kerscher says there is a notable dearth of in vivo data that clearly depict miRNA functions and, therefore, the therapeutic potential these molecules may hold.
Chief among the challenges for target discovery and validation is the need to develop a sufficiently high-throughput experimental approach to find and verify miRNA-mediated mRNA targets — a daunting task as some miRNAs are estimated to regulate anywhere from 10 or 20 to 100 different targets. At present, researchers are limited to only looking at a few targets at a time, instead of comprehensively revealing the entire picture of targets. "Experimentally, we need better tools for detecting certain miRNAs," says Jingfang Ju, co-director of translational research at Stony Brook University.
Ju, who studies miRNAs in the context of chemoresistance in colon cancer and solid tumors, recently developed a novel approach to discover both degraded and non-degraded miRNA-mediated targets in order to investigate the therapeutic potential and prognosis significance of several miRNAs — miR-140, miR-192, and miR-215 — that play a crucial role in colorectal cancer through cell proliferation, cell cycle control, and chemoresistance.
"People are trying to develop in situ microRNA detection approaches — the current immunohistochemistry approaches for detecting miRNA are not sensitive enough, they're only good for high-abundant microRNAs," Ju says. "The field needs to develop novel approaches to force high--sensitive in situ detection because we need to know not only the level of microRNA, but also where they are expressed in the three-dimensional disease tissue, for example."
But some researchers say that when it comes to harnessing small RNAs in the clinic as prognostic and diagnostic tools, miRNAs are already out of date, and that it's time to broaden the scope to include additional classes of small RNAs. Bino John, an assistant professor at the University of Pittsburgh School of Medicine, is focused on viral RNA products that are even smaller than miRNAs, aptly called "unusually small" RNA — or usRNA — which are only 17 nucleotides in length. This new class of RNAs is the smallest for which there is functional evidence. "It is only a matter of time [before] we will be using a combination of multiple molecular fingerprints — such as that of mRNA, microRNAs, piRNAs, usRNAs — in the clinic for both diagnostics and prognostics," John says. "The idea of small RNAs being useful for diagnostics is demonstrated by the Rosetta Genomics miRview tests, for instance, and this field of diagnostics and prognostics will benefit from all sorts of small RNAs, including usRNAs."
Last year, John and his colleagues found usRNAs in human and viral genomes and published a paper in Nucleic Acids Research detailing a new northern blot-based protocol for detecting RNAs 15 to 40 bases in length using digoxigenin-labeled oligonucleotide probes containing locked nucleic acids and 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide for cross-linking the RNA to the membrane. He was also part of a team that helped uncover antisense termini-associated short RNAs — or aTASRs — which indicate the presence of an as-yet uncharacterized endogenous pathway that can copy RNA. Efforts in John's lab are also underway to further develop highly sensitive methods to identify usRNAs, as they can be difficult to detect.
"The critical advantage of these small RNAs, like usRNAs, is that they are very stable in the cell and are perfect candidates for cell-to-cell transfer. And [the] stability of the molecule, and the ability to profile [it] in body fluids, is of paramount importance in diagnosis and prognosis," he says. "Using microRNAs as a gene expression profile of the cell to let us know what type of diseases you're going to get is something I believe is only going to continue to grow in the field."