All life on Earth is cellular and uses DNA to store genetic information. However, evidence suggests that, on ancient Earth, much complex chemical activity preceded cellular development, and it was probably not DNA-based at the start. What was the nature of this activity, and how did it lead to life? "Molecular evolution" is a term used to describe the stages that preceded the origin of life on Earth. The term implies that information-containing molecules were subject to the process of natural selection, whereby genetic structures were capable of both replication (the copying of specific nucleic acid sequences) and mutation. In addition, it is theorized that certain chemical reactions may have taken place on early Earth, before true evolution began, and some of these reactions may have helped to form these informational molecules that enabled replication and mutation.
The Antiquity of Life
The oldest known fossils date from about 3.5 billion years ago, and have been found in Western Australia. Called stromatolites, these domelike rocks are formed today by photosynthetic organisms (cyanobacteria) that live in shallow waters. Large communities of cyanobacteria form microbial mats that trap sediment, providing a sturdy foundation on which another layer of bacteria can grow.
The search for older fossils has been frustrated by the scarcity of pristine rocks that have not been altered by high temperatures and pressures. However, evidence from geological deposits in Greenland dated at 3.87 billion years ago may point to the presence of life at an even earlier time. Since our planet itself is approximately 4.6 billion years old, and several hundred million years were needed for the surface to cool to "hospitable" temperatures, the origin of life must have been quite rapid compared to the vast span of terrestrial history.
Building the Building Blocks: Rna Nucleotides
A productive hypothesis, which has stimulated the design of laboratory experiments that might mimic the formation of biochemical compounds on early Earth, has been the notion that ribonucleic acid (RNA) was the primordial genetic material. Many scientists believe that RNA arose before DNA, and that DNA later took over the information-storage role from RNA. Among the reasons for this view are the modern routes by which pieces of DNA are made. For example, the characteristic sugar that forms the backbone is generated by an enzyme that removes an oxygen (the "deoxy" in deoxyribonucleic acid) from the corresponding RNA sugar, prior to assembly of the chain. This enzyme probably evolved after the appearance of RNA components on early Earth.
RNA chains are composed of repeating units called nucleotides. Each nucleotide consists of a phosphate and a ribose sugar, to which one of the four "bases" (uracil, adenine, cytosine, and guanine) is appended. While phosphate minerals were common on Earth early in its history, each of the other components has been the target of laboratory simulations to determine how they might have been formed.
The bases appear to have been relatively easy to create. Hydrogen cyanide, a possible ingredient in the early oceans and lakes, reacts in the presence of ultraviolet light to give off adenine. Other cyanide derivatives (along with urea) can produce the other bases. A reasonable natural setting for this chemistry would have been a shallow lake or lagoon, as envisioned by Stanley Miller in his preparation of cytosine.
More challenging to those studying the origins of life has been determining how ribose may have been formed and to explore how it may have reacted with the bases (especially cytosine and uracil) and acquired the phosphate to form RNA. Formaldehyde (the same chemical used to preserve specimens) is a likely prebiotic molecule that reacts with itself to give a very complex mixture of products that includes traces of ribose. A more effective route, however, starts with formaldehyde and a derivative known as glycolaldehyde phosphate: these react under alkaline conditions to give mainly a ribose compound with two phosphates attached, and the reaction is promoted by certain minerals.
Heating a mixture of ribose with either adenine or guanine to dryness results in some bond formation between the base and the sugar, but this strategy has failed when cytosine or uracil was used in place of adenine and guanine. More research is still needed to establish plausible routes to these RNA precursors on early Earth, and some skeptics have even proposed that a simpler type of backbone may have preceded that found in nucleic acids today.
Linking Subunits Into Chains
Mineral catalysts have also been useful in forming chains of RNA-like molecules from the activated precursors. James Ferris has focused on the ability of montmorillonite (a type of clay) to promote the assembly of the ribose-phosphate backbone, starting with a special, uracil-containing compound known as a phosphorimidazolide. Although phosphorimidazolides do not occur in nature today, they are highly reactive species and closely related to the modern building blocks of RNA. Most scientists do not maintain that these compounds were present in the primordial soup, but they are convenient substitutes for the natural precursors of RNA.
When the adenine derivative of this compound binds to the mineral surface, the products include chains of up to ten units long, primarily with the "biologically correct" bonds between adjacent riboses. By repeated additions of the starting material, the process can extend the structure up to fifty units. The uracil derivative reacts in a similar fashion, although the chains are somewhat shorter. These data suggest that binding to mineral surfaces may have been important in controlling the proximity and orientation of molecules that could give rise to the first RNA-like fragments, and set the stage for subsequent replication.
Rna Replication Without Enzymes
Once formed, how would the first RNA chains cause copies of themselves to be created? Replication is the process by which DNA or RNA makes copies of specific base sequences. However, the "copy" is not identical to the original, called a template. Rather, it is analogous to a photographic negative, with predicable differences from the original. In the case of RNA, the new strand has a different pattern of bases, determined by the specific interactions (hydrogen bonding) between the old strand and the new one. Thus, adenine bonds to uracil (causing it to become part of the copy), and guanine "directs" the incorporation of cytosine in the same way. As an example, a template sequence abbreviated as AACCAA would be replicated as UUGGUU (the letters stand for each of the four bases). Of course, the faithful transfer of genetic information is a much more complex process, involving an array of complicated protein enzymes and requiring special "activated" precursors for each of the bases.
Leslie Orgel and his associates have carried out extensive studies of replication since 1980. Their goals were, first, to establish whether this process could occur without the help of enzymes (which would not have been present in the environment of early Earth); second, to analyze the accuracy of the copies; third, to explore the limits on the types of sequences that could be copied; and fourth, to determine if the copies could then serve as templates for self-perpetuating replication. These aspects have met with different degrees of success, as discussed below.
Orgel's first breakthrough came with the reaction of activated guanine derivatives (again using phosphorimidazolides) on a template consisting of repeating cytosines, which was thus very similar to a small piece of RNA, except that it had only one type of base. In the presence of a zinc catalyst, the guanine derivatives bound to the template and formed chains with more than thirty guanines linked to one another. (Without the template, the only products were those with two or three bases.) Equally striking was that the guanine chain contained mainly the same type of phosphate-ribose backbone as in native RNA, and the template preferentially bound the "correct" base more than 99 percent of the time. Even if the activated derivatives of uracil, adenine, and cytosine were present, they become incorporated into the product with less than 1 percent efficiency. These early experiments demonstrated that templates could accurately form long chains with the appropriate bond between the sugar and phosphate.
The copying of other sequences using this approach has been more difficult, however, partly because of the unreactive structures that the templates often form. For example, RNA chains containing only adenine mixed with other chains of uracil tend to form aggregates that hinder replication. Guanine is even more unusual, in that it organizes into arrays with four chains locked together, which also prevents replication, although this problem may be overcome by employing very dilute reaction conditions (more relevant to early Earth). The greatest success has been achieved with mixed, cytosine-rich templates that contain adenine, guanine, or uracil as isolated bases separated by at least three cytosines.
A further difficulty is that none of the systems studied by Orgel is capable of replication beyond the first stage: the copies can never serve as templates themselves, because they remain tightly bound to the original template. However, Gunter von Kiedrowski made an important advance in 1994, when he showed that sets of three bases containing guanine and cytosine on a DNA backbone could assemble on a template six units long and then separate. For example, two fragments of GGC could link together on a CCGCCG framework, and then break apart to form a GGCGGC template, available for further replication. The reaction conditions were quite different from what might have existed on early Earth, especially the chemical used to form the bond between the GGC units, but it supports the concept that true replication of this type might be possible.
Rna Can Act As an Enzyme
In cells, replication is controlled by protein catalysts called enzymes. However, since proteins are thought not to have been present on early Earth, how could replication and its related reactions been catalyzed? A key discovery, which has affected how molecular biologists and biochemists view RNA, is that it can also act as a catalyst. In the early 1980s Thomas Cech and Sidney Altman independently discovered RNA (that is, non-protein-based) catalysts in a variety of cell types, whereas scientists had previously thought that only proteins have this power. The concept that RNA could both store information and accelerate biochemical processes made it a much more likely candidate for catalyzing reactions in early life.
The catalytic RNAs (known as ribozymes) found in nature today mainly promote reactions that involve removing certain sequences from an RNA chain before it can be used in protein formation. However, a variety of exotic tools in the molecular biologists' arsenal have now allowed the creation of new ribozymes by in vitro evolution (evolution in a test tube), a method developed by Jack Szostak. He and his associate, David Bartel, showed in 1993 that a novel ribozyme with the ability to link together two RNA chains could be evolved in the laboratory through successive rounds of selection and amplification (creating many copies of the most effective sequence from the previous round). After eight rounds, the best catalyst was faster by a factor of three million, compared to the uncatalyzed reaction.
Bartel has applied this strategy toward the development of the first RNA that catalyzes replication: it binds a smaller RNA template, which then creates a copy that is up to 14 units long. When a set of these replicated products was analyzed, the average accuracy for incorporation of the correct base was 96.7 percent (lowest accuracy was achieved for adenine, greatest for guanine). Although the ribozyme itself was 189 units long and thus represents a highly complex structure that would probably not have formed spontaneously in the early ocean, the experiment demonstrates that RNA can promote the critical step of "peeling off" the copy to allow another cycle of product formation. Such evolutionary exercises thus provide a powerful method for exploring the relationship between ribozyme sequence and the catalytic activity.
Goals for Future Research
Theories of molecular evolution have been based on the paradigm that RNA (or a similar structure) served as the first genetic molecule. Many gaps remain in our understanding, from the assembly of the individual units, to the replication of RNA sequences. Further advances in the study of catalysis, especially with minerals and ribozymes, may provide new avenues for future researchers to explore.
Bibliography
Ferris, James P. "Chemical Replication." Nature 369 (1994): 184-185.
Fry, Iris. The Emergence of Life on Earth: A Historical and Scientific Overview. NewBrunswick, NJ: Rutgers University Press, 2000.
Hayes, John M. "The Earliest Memories of Life on Earth." Nature 384 (1996): 21-22.
Joyce, Gerald F. "RNA Evolution and the Origins of Life." Nature 338 (1989):217-224.
———. "Directed Molecular Evolution." Scientific American 267 (Dec. 1992): 90-97. Orgel, Leslie E. "Molecular Replication." Nature 358 (1992): 203-209.
———. "The Origin of Life on the Earth." Scientific American 271 (Oct. 1994): 77-83.
Von Kiedrowski, Gunter. "Origins of Life: Primordial Soup or Crepes?" Nature 381 (1996): 20-21.
Wills, Christopher, and Jeffrey Bada. The Spark of Life: Darwin and the Primeval Soup.Cambridge, MA: Perseus Publishing, 2000.
—William J. Hagan
