For most of us, curiosity about the beginnings of life focuses on the creatures that emerged from the primordial soup: what did they look like? how did they behave? how did they evolve into the life forms we know today?
But some molecular biologists have a different perspective. They want to know what happened before the first living cells appeared on earth: what were the progenitor molecules that gave rise to life?
In a July issue of Science magazine, researchers at the Whitehead Institute for Biomedical Research and Massachusetts General Hospital (MGH) reported what could be a major step forward in understanding this prebiotic world.
Their data suggest that complex "ribozymes" (catalytic molecules composed of RNA) could have arisen in one step from long random sequences of RNA. Using techniques of in vitro evolution, they have produced a large ribozyme whose speed in promoting chemical reactions is greater than that of any known natural ribozyme and comparable to that of some protein enzymes.
This finding has important implications not only for evolutionary biologists but also for the pharmaceutical industry, where ribozymes offer a potential new source of catalysts for medical science.
Interest in ribozymes began with the Nobel-prize winning discovery by Drs. Thomas Cech and Sidney Altman that a small number of natural RNA molecules could act as catalysts to increase the speed of chemical reactions. Most conventional studies of RNA have focused on its role as a template for protein synthesis-an intermediary in the transfer of information from DNA into protein molecules. Protein molecules are the workhorses of the modern cell. Almost all biochemical reactions, from energy production to the breakdown of wastes, require protein enzymes to achieve a suitable reaction rate.
Drs. Cech and Altman's discovery of enzyme-like RNA molecules, or ribozymes, created enormous excitement among evolutionary biologists. It suggested the possibility of an RNA world-a time before protein-based catalysts in which all life forms depended on RNA. Ribozymes appeared to fulfill the two basic requirements of life's "progenitor molecules." Unlike protein molecules, they could act as templates for their own reproduction (in a way that might lead to descendant molecules with new and different characteristics-the basis of evolution) and they potentially had the capacity to catalyze these reproduction reactions.
These early studies galvanized support for the existence of an RNA world, but they provided little information about the dimensions of that world. Despite extensive research, scientists have found only seven classes of ribozymes in nature.
"The challenge of attempting to describe an RNA world based on those seven classes of ribozymes is enormous," Dr. David Bartel, a Whitehead Fellow, said. "It would be like asking a protein biochemist to describe all of protein catalysis based on seven different types of protein molecules; it couldn't be done. We needed more examples of ribozymes to place these natural molecules in context."
In 1993, Dr. Bartel and his thesis advisor Dr. Jack Szostak, professor of genetics at Harvard Medical School and MGH, set out to explore the range of possible RNA enzymes. They used modern technology to construct more than 1,000 trillion random RNA sequences (each composed of 220 RNA building blocks) and then searched for the very rare sequences with catalytic activity. Repeated selection resulted in the identification of 65 synthetic ribozymes, or about one in 20 trillion of the original RNA molecules.
The researchers then used a novel form of evolution to create more efficient catalysts. They copied the 65 ribozymes under conditions that introduced mistakes into the copying process and selected for the most efficient "descendant" ribozymes within the new population of molecules. This strategy led to ribozymes with efficiency levels hundreds of times greater than those of the original 65 molecules.
In the current study, Dr. Bartel, Dr. Szostak and Erik Ekland of the Whitehead Institute went a step further, with some surprising results.
They examined representatives of seven families of RNA molecules that had descended from the original 65. In each case, they identified the portion of the RNA molecule responsible for the catalytic reaction (the catalytic domain) and then determined its size and structure. One family of ribozymes, labeled "class 1," proved to be particularly interesting.
"The class I ribozyme has a minimal catalytic domain of 93 nucleotides (RNA building blocks), a remarkably complex structure for something that emerged from random sequences," Dr. Bartel said. "We calculated the probability of finding such a ribozyme in our pool of 1,015 random RNA sequences-it was less than one in 10,000. We had to ask ourselves why it appeared when our sample size was so small."
Scientists have theorized that the earliest ribozymes in the RNA world were small; that larger ribozymes would have had to evolve over time. The discovery of the class 1 ribozymes challenges this view.
"The most likely interpretation of our data is that for every small ribozyme motif that can catalyze a reaction there are thousands, if not millions, of distinct large ribozyme motifs," Dr. Bartel said. "Each of the large motifs is quite rare, but together they make up a sizable fraction of the ribozymes in the pool."
Mr. Ekland added, "These results suggest that even the most complex natural ribozymes, such as RNAse P [first isolated by Dr. Altman and his colleagues] and the self-splicing introns [first isolated by the Cech group] could have arisen in one step from long random sequences, and that complex ribozymes may have played an important role early in the RNA world."
The Whitehead and MGH scientists also were extraordinarily successful at improving the class 1 catalysts with their in vitro evolution strategy. One improved or "optimized" class 1 ribozyme is almost three times faster than the fastest known derivative of a natural ribozyme. It also ranks near the top in overall catalytic efficiency.
The ability to generate such an efficient ribozyme from random sequences has important implications for industry, where the task will be to find ribozymes that fulfill a specific medical need (for example, as a diagnostic or therapeutic agent) and then remodel the ribozyme to make it as efficient as possible.
This research was supported in part by a medical research grant from the W.M. Keck Foundation to the Whitehead Fellows Program (Dr. Bartel and Mr. Ekland), and by grants from NASA and Hoechst AG (Dr. Szostak and Dr. Bartel).
A version of this article appeared in MIT Tech Talk on September 13, 1995.
