MicroRNA Shows Macro Potential

Courtesy of Bonnie Bartel and David P. Bartel Reprinted from Plant Physiol, 132:1-9, June 2003. CURRENT MODEL: (A) A microRNA (MIR) gene encodes a primary transcript with a stem-loop structure. A hairpin precursor (in brackets) has been found in animals but not in plants. The enzyme Dicer cleaves the transcript to form a microRNA (miRNA) which nestles in a ribonucleoprotein (RNP) complex that is similar, if not identical, to the RNA-induced silencing complex (RISC). (B) When an mRNA has near-perfect complementarity to an miRNA, the mRNA is cleaved. (C) When an mRNA has less complementarity to an miRNA, the mRNA's translation is attenuated. The RNA paradigm is due for another correction. Once, RNA seemed a mere conduit between genes and proteins. Its status expanded with reports of catalytic RNA 20 years ago and of endogenous RNA interference 10 years later. Now microRNAs, the 20- to 24-nucleotide transcripts that underlie much of that endogenous interference, appear to be more biologically significant than first suspected. Until April, studies implicated microRNAs in development of the worm Caenorhabditis elegans and differentiation in the plant Arabidopsis thaliana. Papers published in the past two months, however, report that microRNAs also affect apoptosis, tissue growth, fat metabolism, and nervous system patterning.1-4 These papers are the first fruits of a new focus in this small but expanding field. Having located hundreds of microRNA-encoding genes, investigators are seeking microRNA targets and functions with tools ranging from traditional genetics to computer-based genome scanning. This time, the RNA paradigm should accommodate unusual findings more easily than it once did. Sidney Altman, the Yale University biologist who codiscovered catalytic RNA, recalls encountering "extreme resistance" to getting his work published, though it eventually earned him a 1989 Nobel Prize in chemistry. Contrast that experience with this reminiscence by Victor Ambros, who identified lin-4 as the first microRNA-encoding gene: As he presented a poster about lin-4 at a 1992 Gordon Conference, the editor of Cell "came by and said, 'When it's ready, send it to us.'" LOLLIPOP-LIKE STRUCTURES Ambros, now a genetics professor at Dartmouth Medical School, had cloned lin-4 in C. elegans because it had an intriguing property when mutated: It caused constant repetition of the first larval stage. "The gene turned out not to encode a protein," he says. "So we were forced to figure out what the gene product was, and it ended up being a little RNA." This snippet became bound to a nearly complementary messenger RNA (mRNA) transcript and repressed its translation into protein. Ambros' seminal paper on lin-4 appeared in 1993.5 Seven years later, geneticist Gary Ruvkun, at Harvard Medical School, reported let-7 as the second microRNA-encoding gene in C. elegans.6 Many researchers still regarded such genes as relics of the occasionally unique, weird world of worm molecular biology. But by late 2001, that view had become untenable after several labs detected genes similar to let-7 and lin-4 in many animal and plant species. The number of suspected microRNA genes has ballooned. David P. Bartel, an associate biology professor at the Whitehead Institute for Biomedical Research, estimates there are 90-120 such genes in C. elegans and 200-255 in humans. About 40 genes encoding 19 microRNAs have been located in Arabidopsis, he adds. MicroRNA genes encode primary transcripts containing complementary stretches that curl into lollipop-like stem-loop structures. The enzyme Dicer then cleaves these transcripts into short double-stranded RNAs, one strand of which becomes a microRNA. (MicroRNAs should not be confused with short interfering RNAs, or siRNAs. The two RNA species are generated by Dicer cleavage and can have similar effects, but siRNAs are not transcribed from discrete genes.) ZAPPING PLANT TRANSCRIPTION An incomplete model of how microRNAs operate has emerged, "supported by a lot of bits and pieces" of evidence, says professor Gregory J. Hannon, of Cold Spring Harbor Laboratory. After transcription and processing--most details of which remain unclear--the RNAs nestle in protein complexes known as RISCs (RNA-induced silencing complexes). RISCs associate with ribosomes in cell-free extracts but maybe not in vivo. RISCs in cells, however, definitely bond with mRNAs through base-pairing between microRNAs and mRNAs. An mRNA is cleaved if a stretch of nucleotides is exactly or almost exactly complementary to a microRNA's sequence. The sites of mRNA cleavage are known, but the responsible enzyme is not. Translation is repressed if the two RNA pieces contain more mismatches. What occurs during repression "is a real mystery," says Hannon. No one has claimed success yet in isolating a microRNA-target duplex by immunoprecipitation or any other means. Courtesy of Bruce A. Hay Reprinted from Curr Biol, 13:790-5, April 29, 2003 THE EYES HAVE IT: The smaller Drosophila melanogaster eye (left) expresses the protein Reaper, an apoptosis activator that shrinks the organ's size. The larger eye (right) expresses both Reaper and mir-14, a microRNA that suppresses apoptosis. Transposons with an eye-specific promoter ferried the Reaper and mir-14 genes into flies' genomes. Because plant microRNAs are nearly complementary to their targets, genome scans have found about 60 targets in Arabidopsis. Many of these targets "are transcription factors that control some specific developmental events," notes James C. Carrington, director of Oregon State University's Center for Gene Research and Biotechnology, in Corvallis. "For example, a lot of these transcription factors are required to maintain stem cells as undifferentiated cells in the right abundance in plant meristems. Other genes known to be targets are involved in controlling radial pattern and controlling what determines the types of cells that develop on the top of a leaf versus the bottom of a leaf." To account for this reliance on microRNAs, Bartel hypothesizes that as a plant cell differentiates, it must switch from one set of transcription factors to another. If it had to rely on promoter-binding proteins to inhibit DNA transcription of the earlier set of transcription factors, those factors would not be turned off "until their mRNA is gone," Bartel explains. "So the idea is that a microRNA comes in and targets those messages for degradation. And so you can move from one program of transcription factors to another one very quickly without having to depend on constitutively unstable messages." The next challenge for plant researchers is to find out how relevant these targets and functions are in vivo. Carrington cites several strategies: One is to insert an enhancer near a microRNA gene to boost its activity; another, ironically, is to lower its activity by exogenous RNA interference. His lab is using Affymetrix-type microarrays to monitor which mRNAs are cleaved when production of a microRNA rises or falls. Some investigators are also mutating the third-codon positions of a microRNA's putative target. "Now you have a target gene that should be fully functional but is no longer a substrate for cleavage by the microRNA," says Carrington. "Then you ask, exactly what happens when you put that back into a plant?" Since the mutant genes are dominant, he adds, phenotypes can emerge even when the wild-type target gene is not eliminated. FLIES, FLUORESCENCE, AND FAT Animal microRNAs are not as easy to study as plant microRNAs. One reason: The animal transcripts are not nearly as complementary to their targets, so mRNA targets are hard to detect by genome scans. In addition, multiple copies of microRNA genes in some animals stymie a knockout approach to determining function. Mice, for example, have at least a dozen copies of the let-7 gene. Nevertheless, Ambros, Bartel, and H. Robert Horvitz, an MIT biology professor and a 2002 Nobel laureate, are collaborating to knock out microRNA genes in C. elegans. Ruvkun observes that most such knockouts are expected to have subtle functional changes "because so few have emerged from traditional genetics up to this point." Another explanation for geneticists' earlier inability to detect microRNAs is that the genes are too small to be hit easily by mutagens. Until this spring, only a handful of animal microRNA-mRNA reactions were reported: lin-4 targeted lin-14 and lin-28, and let-7 targeted lin-41. Three other genes' mRNAs, however, now appear as likely targets. One gene is the worm's lin-57 (also known as hbl-1), a homolog of the Drosophila melanogaster gene hunchback. Two teams of researchers recently announced that the 3' untranslated region (UTR) of lin-57/hbl-1 contains multiple possible binding sites for let-7 microRNA.3-4 Various data indicated that lin-57/hbl-1 is probably a downstream target of let-7 microRNA, which thereby influences development and nervous- system patterning. Two Drosophila microRNAs and their targets also just made their debuts. Using traditional mutagenic techniques, Stephen M. Cohen, of the European Molecular Biology Laboratory in Heidelberg, Germany, discovered the gene bantam in a nonprotein-coding region of the fly genome. He found that bantam encodes a microRNA that stimulates cell proliferation and prevents apoptosis.1 Then he used an unpublished computational method to identify the apoptosis-inducing gene hid as a target whose translation is repressed by bantam microRNA. Experimental evidence backed this hypothesis: when bantam was coexpressed with hid, hid protein levels fell, but its mRNA levels stayed the same. Courtesy of Bruce A. Hay Reprinted from Curr Biol, 13:790-5, April 29, 2003 FAT FLIES: In normal Drosophila melanogaster, small lipid droplets are scattered throughout adipocytes' cytoplasm (left). Deleting the mir-14 gene boosts the number and size of lipid droplets (right). Cohen's study also pioneered use of an in vivo sensor, which Ambros calls a "beautiful approach for detecting where microRNAs are expressed." The sensor can substitute for in situ hybridization, which has yet to succeed in tracking microRNAs because they are apparently too short to generate a sufficient signal. Cohen made a transgene that expressed green fluorescence protein (GFP) and, in the gene's 3' UTR, he placed two copies of a target sequence exactly complementary to bantam microRNA. Wherever fly cells expressed bantam, the transgene was cleaved and fluo-rescence was dimmed. Cohen suggests a similar approach for helping establish that a UTR is a microRNA's target. It involves placing a gene's 3' UTR behind a GFP transgene and testing how the microRNA affects fluorescence. "That's dead easy," he says. "It just takes time." Bruce A. Hay, a fly biologist at the California Institute of Technology, generated flies whose overexpression of Reaper, an apoptosis-promoting protein, transformed Drosophila's normally big red eyes into small, pale ones. He then mated these flies with others whose genomes contained randomly inserted transposons. These insertions had a promoter that increased eye-specific expression of nearby genes. When progeny had normal eyes, that outcome meant that an inserted transposon was turning on a nearby gene that suppressed cell death. Ten insertion sites resulted in normal eyes, but only six of these sites were near protein-coding genes. For four years, Hay remained clueless about the other four sites. He was aware of Ambros' papers on lin-4, but, he now laughs, "We put them in the file cabinet that said 'interesting but probably not relevant to our work.'" After his interest was piqued by studies showing microRNA genes in many organisms, Hay confirmed that one of the puzzling transposon insertion sites was near such a gene.2 When he deleted this gene, mir-14, flies became sensitive to stress, short-lived, and obese. To identify sites to which mir-14 microRNA binds, Hay examined the genes of several known apoptosis regulators. He concluded that this microRNA regulates the caspase gene Drice because levels of its protein product changed when he manipulated mir-14 expression. In the absence of RNA measurements, however, Hay could not establish that Drice mRNA is a direct target of mir-14 microRNA. CONSERVATION, KEY TO COMPUTATION Hay did not perform computational genome-scanning, but he and other investigators are hoping that this strategy will detect many potential microRNA targets, especially those outside of known biological pathways. Yet one big problem must be overcome. Because microRNAs are short and often not very complementary to their targets, unsophisticated genome scans yield too many potential targets. The vast majority are undoubtedly false-positive hits. How confounding a factor is loose complementarity? "In a computational study," says Bartel, "You can find that a third of the known elegans microRNAs will hybridize to the lin-14 UTR about as well as lin-4 [microRNA] does." Ruvkun, however, does not view the computational challenge as extremely difficult. "It's only 20 bases you have to think about," he contends. "It's a lot simpler than anything longer, where [RNA] can really fold up into something complicated." The key to a viable scanning approach, researchers maintain, is genetic conservation. The idea is that if some microRNAs did not evolve too much, their targets could not have evolved too much either--and vice-versa. For Thomas Tuschl, head of Rockefeller University's Laboratory of RNA Molecular Biology, a promising strategy means starting with the "20 or so absolutely conserved microRNAs, which you find in C. elegans, fish, humans, Drosophila"; then running a bioinformatics program that seeks approximate matches to those microRNAs; and finally searching among the thousands or millions of hits for conserved mRNA targets (including 3' UTRs) and conserved proteins encoded by or upstream of those targets. "We've learned the hard way that there are complications in how to detect these [targets]," says Chris Sander, head of Memorial Sloan-Kettering Cancer Center's computational biology center, who is collaborating on this work with Tuschl and with Debora Marks at Columbia University. Trained as a theoretical physicist, Sander notes that a genome-scanning program must account for "interesting loop-outs and bulges" in microRNA-target hybridization, the influence of nearby proteins (for example, the RISC complex), and the energetics of RNA-RNA mismatches. A 90-processor supercomputer works best for one calculation. "What we don't understand is basically how a microRNA appears to the target RNA," says Tuschl. Once he and other researchers identify further targets by genome-scanning, he adds, those results will teach them the rules "for recognition of microRNAs by target RNAs. With that knowledge, you can move on to trying to predict targets of the less-conserved microRNAs." Additional papers on microRNAs' targets and functions are expected in coming months, and excitement about the field's prospects is mounting. "Here you have sort of a unique situation, where all of a sudden, 200 or more new kinds of genes--not just new genes, but new kinds of genes--have been identified, and we don't know anything about how they work," observes Cal Tech's Hay. "For graduate students and postdocs, in particular, this is a place where you can really make a difference because everything you learn at this point is new." Douglas Steinberg (dougste@attglobal.net) is a freelance writer in New York City. References 1. J. Brennecke et al., "Bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the proapoptotic gene hid in Drosophila," Cell, 113:25-36, April 4, 2003. 2. P. Xu et al., "The Drosophila microRNA mir-14 suppresses cell death and is required for normal fat metabolism," Curr Biol, 13:790-5, April 29, 2003. 3. J.E. Abrahante et al., "The Caenorhabditis elegans hunchback-like gene lin-57/hbl-1 controls developmental time and is regulated by microRNAs," Dev Cell, 4:625-37, May 2003. 4. S.-Y. Lin et al., "The C. elegans hunchback homolog, hbl-1, controls temporal patterning and is a probable microRNA target," Dev Cell, 4:639-50, May 2003. 5. R.C. Lee et al., "The C. elegans heterochromic gene lin-4 encodes small RNAs with antisense complementarity to lin-14," Cell, 75:843-54, 1993. 6. B.J. Reinhart et al., "The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans," Nature, 403:901-6, 2000. function sendData() { document.frm.pathName.value = location.pathname; result = false if (document.frm.score[0].checked) result = true; if (document.frm.score[1].checked) result = true; if (document.frm.score[2].checked) result = true; if (document.frm.score[3].checked) result = true; if (document.frm.score[4].checked) result = true; if (!result) alert("Please select") return result; } .myradio{background-color : #94bee5} Please indicate on a 1 - 5 scale how strongly you would recommend this article to your colleagues? Not recommended 1 2 3 4 5 Highly recommended Please register your vote

MicroRNA Shows Macro Potential

Interested in reading more?

Become a Member of