Lively, hardy, and inexpensive, the zebrafish is a popular species for beginning tropical-fish hobbyists. An attractive small fish native to the Ganges River and other freshwater sites in South Asia, it is also a darling of developmental biologists because, in addition to being easy to maintain in large numbers, zebrafish develop rapidly—their major organ systems undergo substantial development in just 24 hours—a phenomenon made all the more remarkable by the transparency of the fish’s embryos. Through a microscope, says zebrafish specialist Antonio J. Giraldez, Ph.D., scientists have an intimate, clear view as the fish’s adult form swiftly unfolds.

Some basic principles of molecular biology have lately seemed to change just as dramatically. The wholly unexpected discovery 20 years ago that short stretches of genetic material called microRNAs (miRNAs) exert profound and pervasive control of gene expression is a case in point. The study of miRNAs has since become one of the fastest-growing areas in biology, and now, by taking advantage of the zebrafish’s unique qualities, research in the laboratory of Giraldez, assistant professor of genetics and the Lois and Franklin H. Top, Jr. Yale Scholar (see box below), is prompting scientists to rethink strongly held ideas on how miRNAs are formed.

Until the early 1990s, it was thought that, for the most part, genes are activated or suppressed by transcription factors, which bind to dna to promote or inhibit the transcription of genes into messenger RNA (mRNA), and hence determine which genes are ultimately translated into proteins. But while studying the development of the microscopic roundwormCaenorhabditis elegans (another handily transparent organism), Victor Ambros, Ph.D., then at Harvard University, made the startling discovery that a tiny stretch of RNA, only 22 genetic letters long, switched off a crucial gene that orchestrates the timing of developmental events in C. elegans.

Over the next 10 years, scientists determined that small RNAs like that discovered by Ambros are a ubiquitous, fundamental regulator of gene expression throughout the plant and animal kingdoms. More than 700 miRNAs have been identified in humans, each of which may regulate hundreds (or even thousands) of genes; with such wide reach, miRNAs may interact with more than 60 percent of our genome. These findings have revealed that miRNAs “have deep implications not only in how humans and animals are made, but in the development of human diseases,” says Giraldez, also a member of the Yale Stem Cell Center. In 2006, the Nobel Prize in Physiology or Medicine was awarded to two American scientists for elucidating one of the main mechanisms by which miRNAs silence genes, an extraordinary turn of events considering that just 15 years earlier miRNAs were not even known to exist.

Scientists have revised many of their ideas over the past 20 years regarding how miRNAs work and how they are formed, but one character in the miRNA story has remained unchanged: Dicer, an enzyme that, as its name implies, snips complex precursors into the 21- to 23-nucleotide length that characterizes functional miRNAs. In addition to the zebrafish’s other attributes, the species is also amenable to precise genetic manipulations. In recent work with zebrafish mutants, Giraldez has confirmed Dicer’s central role, showing that fish develop abnormally if the enzyme’s function is compromised. Though diagrams in journals and textbooks have varied in their depictions of the early and intermediate steps of miRNA maturation, up to now Dicer has always stood toward the end of the line, the enzymatic gateway through which all precursors must pass to become miRNAs.

But miRNAs continue to surprise. Giraldez and Yale colleagues, collaborating with scientists from Massachusetts to Japan, have now shown that some miRNAs can be processed by an alternative pathway that does not require Dicer. Moreover, the researchers provide evidence that miR-451, a Dicer-independent miRNA they analyzed in depth, is necessary for the normal production of red blood cells. In the new research, published in the June 25 issue of the journal Science, Giraldez and colleagues expanded on his earlier work with zebrafish mutants. After introducing a mutation to suppress the activity of Dicer, the researchers looked to see whether any functional miRNAs were present in these mutant fish. Using high-throughput genomic sequencing tools, the group found that several miRNAs had indeed been successfully processed. Giraldez recalls that, because the belief that Dicer is essential to miRNA processing was so firmly established, he and his team found this result “extremely weird, and in fact it took us almost two years to believe it,” during which time the group conducted every imaginable experiment until they were satisfied that the finding held up.

Eventually, they concentrated on one Dicer-independent miRNA called miR-451, because it is present in many species (in March, for example, researchers at Ohio State University proposed that miR-451 may regulate the growth of brain tumors in humans), and because its distinctive configuration is a bad structural match for Dicer processing, which suggested that some other pathway was at work.

In typical miRNA processing, once Dicer has cut miRNA precursor strands to a proper-length miRNA, the strands are loaded into a molecule known as the RNA-Induced Silencing Complex (RISC), where the business of gene-silencing actually takes place. Inside RISC, a second enzyme called Argonaute 2 (Ago2) slices up any mRNA strand containing a sequence that exactly matches the loaded miRNA, destroying the genes in that mRNA. Since Ago2 has its own slicing activity, the Giraldez team hypothesized that Ago2, rather than Dicer, might process miR-451, and this supposition proved to be correct: in a zebrafish mutant with faulty Ago2, levels of miR-451 were sharply reduced. Previous studies had shown that miR-451 is essential for normal red blood cell production, and the Ago2-mutant zebrafish were found to be anemic (see photo).

This observation nicely complements another new study by a Cold Spring Harbor Laboratory research team (two members of which also worked with Giraldez and colleagues on the zebrafish study), who found that mice with compromised Ago2 also had reduced levels of miR-451 and anemia. Because the Ago2 processing pathway for miR-451 and other miRNAs has been conserved in vertebrate animals over evolutionary time to regulate processes as basic to life as red blood cell production, Giraldez believes that these seeming exceptions to the Dicer-based rule may prove to have much wider implications.

“There is an immense, vast sea of small RNAs out there, and it is difficult to sort out what is junk from what is functional,” says Giraldez. “This discovery really opens the door to finding new families of these RNAs that influence many forms of biological activity. With this new ‘molecular scissors,’ we have another tool to find small RNAs that are important to life, that activate genes in disease, and may be important in developing new therapeutics.”