Illustration: Ned Shaw
Nearly 20 years ago, Tom Cech and Sidney Altman discovered that some naturally occurring RNAs could perform enzymatic reactions, earning these researchers the 1989 Nobel Prize in chemistry. Scientists have now identified several examples of RNA enzymes, or ribozymes. Most make or break the phosphodiester bonds in nucleic acid backbones, but some perform other reactions: The ribosome's peptidyl transferase center is also a ribozyme.1
What makes ribozymes particularly interesting is that they are amenable to tinkering. Creation of a protein enzyme de novo, or modification of an existing one to function in a different way, is a monumental, and often futile effort. But nucleic acids lend themselves to this type of research because they uniquely couple genotype and phenotype, says Scott Seiwert, group leader, diagnostic systems, Ribozyme Pharmaceuticals (RPI), Boulder, Colo. In other words, when an active ribozyme is selected, it can be cloned and sequenced directly; the same is not true of active proteins. Also, ribozymes recognize their nucleic acid substrates largely through secondary, not tertiary structure, and they have a modular structure that allows scientists to graft together different functional units. These attributes make manipulating ribozyme behavior relatively straightforward.
"If I've got your average protein, could I graft on an ATP-binding domain to that protein in a way that would make that protein ATP-dependent?" asks Andrew Ellington, professor of chemistry and biochemistry, University of Texas, Austin. "No way. There'd be no way I could do that without an incredibly detailed knowledge of the tertiary structure, and even then I doubt that I would have a lot of success. But, for ribozymes, all I need is a knowledge of the secondary structure and I can immediately carry out a series of experiments, one or more of which is likely to be successful."
Researchers have indeed been very successful. Using a brute force approach coupled with clever experimental strategies, scientists are creating new catalytic ribozymes, coaxing them to polymerize RNA, charge tRNAs, and cleave viral RNAs, for example. These scientists are putting these reagents to work probing our evolutionary heritage, developing clinical and research diagnostic tools, and investigating new therapeutic stategies.
To create these novel ribozymes, researchers use selection-amplification, an iterative process involving multiple rounds of selection, reverse transcription, PCR amplification, and transcription. It is a molecular lottery writ large, seeking one or two winners from a pool of about one thousand trillion (1015) contestants, each containing a random sequence 40 to 200 nucleotides long.
The selection procedure usually relies on modification of the desired ribozyme. Some strategies change the size of the starting molecule, for example by ligating a new sequence onto it. Others attach a chemical tag, like biotin, to facilitate enrichment and purification. Some researchers add an additional level of complexity, controlling the selection process to yield allosteric ribozymes, which are catalysts controlled by the presence or absence of a ligand.
THE RNA WORLD A number of scientists believe that life was based on RNA before the evolution of the modern nucleoprotein-based world. Several researchers are trying to recreate bits of that world, to prove that RNA can perform the enzymatic tasks, such as replication, that must have been required if an all-RNA-based life form ever existed.
"It's generally accepted," says Michael Yarus, professor of molecular, cellular, and developmental biology, University of Colorado at Boulder, "that RNA creatures were our predecessors. If that's true, RNA cells, or ribocytes as we call them, have to be possible. ... and that's a notion that can be tested by setting out to make the bits and eventually put them together."
Modern organisms, says Yarus, harbor evidence of their evolutionary past--the ribosome's peptidyltransferase center, for example, and ribonucleotide-containing enzyme cofactors like NAD. He calls these holdovers "bits of our patrimony ... from these [RNA-based] creatures." Yarus' lab tries to recreate several critical elements of those organisms' biology, including a translation system, metabolism, and transmembrane transport.
Hiroaki Suga, associate professor of chemistry, State University of New York at Buffalo, also studies translation. His lab recently developed a ribozyme analogous to an aminoacyl tRNA synthetase, which contains a recognition element that specifies which tRNA the ribozyme will charge.2 This work has practical applications, as it may enable researchers to insert nonnatural amino acids into proteins synthesized by in vitro translation systems, says Sidney Hecht, professor of chemistry, University of Virginia.
David Bartel, associate professor of biology at the Whitehead Institute for Biomedical Research, and Massachusetts Institute of Technology, concentrates on yet another essential activity: ribozyme replication. Last year his lab created an RNA polymerase ribozyme,3 based on an RNA ligase ribozyme that he found as a graduate student in Jack Szostak's lab at Harvard University.4 This work suggests that RNA can replicate itself, a crucial piece of the "RNA world" hypothesis, but it is only a first step. "I'm sure that in every way, our polymerase is worse than modern protein-based RNA polymerases," Bartel says. The polymerase can extend a primer-template pair by as many as 14 nucleotides, but it does so extremely slowly, because the enzyme has a very low affinity for the template. It also exhibits poor affinity for nucleotides and relatively poor (98.5%) accuracy.
Not all ribozymes fare so poorly in head-to-head competition against their proteinaceous counterparts. According to Ron Breaker, associate professor, molecular, cellular, and developmental biology, Yale University, some ribozymes, under the right circumstances, can more than hold their own against the proteins they mimic. And Bartel notes, "We shouldn't be too hard on this polymerase--it came out of random sequences. It hasn't had the benefit of billions of years of evolution, and it can do a pretty sophisticated reaction."
NEXT-GENERATION BIOSENSORS Several researchers, and a few biotech firms, are bullish on allosteric ribozymes, or aptazymes, envisioning a bright future in both clinical and research diagnostics. One such company is Archemix of Cambridge, Mass. The company is developing aptazymes as "next-generation biosensors," says Breaker, one of Archemix's cofounders. Aptazymes, he explains, "are turned on or shut off when they come in contact with some target molecule. They act as molecular sensors, and when they trigger, they fire a chemical reaction." Generally, these targets are small organic molecules, but they can also be proteins.5
Allosteric ribozymes currently come in two basic flavors: ligases and nucleases. In a typical application, when a ligase binds its activator, it covalently attaches a fluorophore-labeled oligonucleotide to itself, so that the presence of the analyte triggers an increase in fluorescence.6 Nucleases, on the other hand, are generally based on the separation of a fluorophore from a fluorescent quencher by a cleavage reaction. These latter enzymes are often based on hammerhead ribozyme motifs. Hammerhead ribozymes are small, naturally occurring plant virusoid enzymes whose normal function is to process multimeric viral RNAs into monomers.
Two recent reports hint at the power of aptazymes in diagnostic applications.7,8 Michael Famulok at the University of Bonn, and colleagues developed a series of reporter nucleases, which could monitor protein-nucleic acid and protein-protein interactions to identify drug candidates.7 Seiwert and coworkers at RPI generated cleaving ribozymes that could distinguish between the phosphorylated and unphosphorylated forms of ERK2, a protein kinase.8 According to Ellington, another Archemix cofounder, Seiwert's paper "shows the power of these effector- activated ribozymes to be able to look at individual proteins, or to be able to do things in the context of proteomics that maybe even antibodies might have a little bit of difficulty with."
Chief among these problem areas: protein arrays. Researchers have had trouble developing protein biochips because of the difficulty in arraying hundreds or thousands of different, functional proteins.9 But aptazyme arrays, says Ellington, could deliver the power of protein chips with existing DNA array technology. "You can easily envision [DNA array manufacturer] Affymetrix generating aptazyme arrays using the technologies they already have in place." The difference, says Breaker, is that while traditional DNA arrays can detect other nucleic acids, aptazyme arrays can detect nucleic acids, organic molecules, proteins, toxic metals, and other targets. The resulting arrays could advance drug discovery, proteomics, and even metabolomics research.
Breaker and his colleagues at Yale have already developed a prototypical biosensor array, which accurately identified a metal ion, cyclic nucleotides, metabolites, and drug analytes.10 Ellington describes this work as "cool"; Breaker describes it as "a Model-T RNA biochip." But he also says the research was a harbinger of the field's promise: "It really struck us that with some good biochemistry, some good molecular engineering, and some savvy interfacing with solid supports, ... you could actually build a next-generation biosensor using molecules that have been evolved in the test tube."
Larry Gold, CEO of Boulder, Colo.-based SomaLogic, cautions, however, that aptazyme arrays may be difficult to develop at high feature density; the instability of such long RNA molecules in the biological samples within which they will likely be used, he says, is a significant technical barrier. SomaLogic focuses instead on RNA- and DNA-based aptamers--short, analyte-binding nucleic acids devoid of catalytic activity--for diagnostic use.
Ellington points out another potential area of exploration: nanotechnology. "I recognize it's just a buzzword," he concedes, "but when I look at what people are trying to do at the nanoscale level, it seems that one of the problems they have is creating ... components at that level." He envisions aptazyme applications in mechanical, electrical, and optical transduction. "When I look at what kind of molecules could be adapted in sort of biological-nanotechnological hybrids, there aren't that many, and aptazymes, because of the ability to engineer them easily and change their properties, look to me like a good bet," he concludes.
RPI is also interested in diagnostics. In collaboration with Fujirebio of Japan, RPI is integrating its Halfzyme™ nucleic acid detection system to test blood for viral contamination, says Seiwert. "PCR is certainly available to do that," he acknowledges, "but as an assay that's run in high-throughput at a clinic, it's in some ways just too complex." The company is also developing protein and small molecule biosensors.
Image: Courtesy of Michael Famulok
AN RNA A DAY...? One intriguing area of research is therapeutics development. One approach uses ribozymes to inactivate viral or pathogenic RNAs in vivo; another uses ribozymes to correct mutant gene transcripts via trans-splicing. According to Nassim Usman, RPI's CSO and vice president of research and development, the company has two trans-cleaving ribozymes in human clinical trials: Angiozyme®, which attacks the vascular endothelial growth factor receptor 1 mRNA, and Herzyme™, which cleaves Her-2/neu mRNA. A third ribozyme, HepBzyme™, currently in preclinical development, is designed to attack hepatitis B viral RNA.
One of the key technological hurdles that RPI had to overcome was the development of modifications that would stabilize the RNAs after injection. RPI primarily uses 2'-O-methyl nucleotides, says Usman, though other, "more esoteric" modifications are also employed. Currently, these drugs are administered "naked." Using a formulation instead would reduce the required dosage, Usman admits, but "formulations also add another layer of complexity." The company will likely begin using formulations with their next generation of therapeutics.
Immusol of San Diego is also developing ribozyme therapeutics. The company has three drugs in clinical development, all of which target the proliferating cell nuclear antigen (PCNA) mRNA. PCNA, says Joan Robbins, VP, product development, regulates cell proliferation. "When you heal," she explains, "you want to proliferate to a certain extent to close the wound, but ... you don't want to overheal."
RNA isn't the only nucleic acid with catalytic aspirations; some researchers are working on DNA-based enzymes. These deoxyribozymes are more stable than ribozymes, says Breaker, who discovered them as a postdoc with Gerald Joyce.11 Bartel agrees, but adds that researchers don't yet know whether DNA is capable of generating the same diversity of enzymatic molecules as is RNA. Nevertheless, some scientists are pursuing DNA-based therapies. Many promising candidates are derived from the "10-23" DNA enzyme, developed in Joyce's lab.12 Levon Khachigian at the University of New South Wales, Australia, for example, has created enzymes based on 10-23 that may help prevent restenosis following coronary angioplasty.13
For all that scientists have achieved with nucleic acid-based enzymes, they have only scratched the surface of these reagents' potential. In vitro selection studies generally begin with 1015 molecules, but that pool represents a miniscule fraction of the total sequence space, says Breaker. A pool of every possible 70-mer sequence, for example, would actually contain 1.4 x 1042 different molecules. If it were possible to screen one molecule per second, it would take about 5 x 1034 years to work through such a library. The universe, by comparison, has been around for only about 1010 years.
What all this means, Breaker says, is that "there is, for all practical purposes, an infinite number of sequences that could be tested." He adds, "When we're doing in vitro selection, we're betting that some of those molecules, no matter how rare they are, will have the function that I'm looking for. And if I'm smart enough to devise a selection scheme that will find that molecular needle-in-a-haystack, then I'm going to be successful."
Jeffrey M. Perkel can be contacted at firstname.lastname@example.org.
1. P. Nissen et al., "The structural basis of ribosome activity in peptide bond synthesis," Science, 289:920-30, 2000.
2. Y. Bessho et al., "A tRNA aminoacylation system for non-natural amino acids based on a programmable ribozyme," Nature Biotechnology, 20:723-8, July 2002.
3. W.K. Johnston et al., "RNA-catalyzed RNA polymerization: Accurate and general RNA-templated primer extension," Science, 292:1319-25, 2001.
4. D.P. Bartel, J.W. Szostak, "Isolation of new ribozymes from a large pool of random sequences," Science, 261:1411-8, 1993.
5. M. Koizumi et al., "Allosteric selection of ribozymes that respond to the second messengers cGMP and cAMP," Nature Structural Biology, 6:1062-71, 1999.
6. M.P. Robertson, A.D. Ellington, "In vitro selection of an allosteric ribozyme that transduces analytes to amplicons," Nature Biotechnology, 17:62-6, 1999.
7. J.S. Hartig et al., "Protein-dependent ribozymes report molecular interactions in real time," Nature Biotechnology, 20:717-22, July 2002.
8. N.K. Vaish et al., "Monitoring post-translational modification of proteins with allosteric ribozymes," Nature Biotechnology, 20:810-5, August 2002.
9. A. Constans, "The chipping news," The Scientist, 16:28-30, April 29, 2002.
10. S. Seetharaman et al., "Immobilized RNA switches for the analysis of complex chemical and biological mixtures," Nature Biotechnology, 19:336-41, 2001.
11. R.R. Breaker, G.F. Joyce, "A DNA enzyme that cleaves RNA," Chemistry & Biology, 1:223-9, 1994.
12. S.W. Santoro, G.F. Joyce, "A general-purpose RNA-cleaving DNA enzyme," Proceedings of the National Academy of Sciences, 94:4262-6,1997.
13. F.S. Santiago et al., "New DNA enzyme targeting Egr-1 mRNA inhibits vascular smooth muscle proliferation and regrowth after injury," Nature Medicine, 5:1264-9, 1999.