Ob-La-Dee, Ob-La-Daa February 25, 2000

By the time I was in high school, I had decided that I wanted to study biology. My baseball career had stalled, I had yet to discover the joys of soccer, my attempts at writing were pretty silly, physics seemed indecipherable, and, even though the Beatles had broken up, my musical skills were only pleasing to lower forms of life. It was a time when m any people worried that the schools were doing a poor job teaching children (which is a lament that has been raised for as long as there has been public education in the US).

The government sponsored several programs to help. We used brand new yellow textbooks designed by the Biological Science Curriculum Study (there is a blue and green version also) to give us a "cutting-edge" exposure to biology. Government grants funded programs such as the Student Science Training Program. A fancy name for a summer I spent at the University of Texas before my junior year in high school, listening to general biology seminars in the morning, doing research in a real lab in the afternoon, and trying to figure out how to sneak into A Clockwork Orange in the evening. All of this confirmed my desire to pursue biology as a career.

And in all of this great training, there was never ANY mention of evolution, the Theory of Natural Selection or the common descent of animals. The central, underlying principle of biology was ignored. Although it was mentioned in the texts, we never seemed to get around to it. There was a lot of biology to cover that the school district had decided was more important to learn than evolution. This lack of exposure has resulted in a lifelong interest in many aspects of evolutionary theory and its applications.

Evolution is behind everything we see in biology. So much of our work implicitly depends on the basic underpinnings of evolution. Every time we use a sequence of DNA from mice to identify a human sequence we are relying on the principle of common descent; the fact that mice and humans have similar DNA sequences because they share a common ancestor.

I am going to spend the next few weeks discussing some aspects of evolution that are becoming clearer, due to our increasing understanding of its mechanics. Not only are we using modern technology to learn how animals change, we are gaining a better knowledge of how life may have first began. One common theme is the appropriation of a structure for a new use after it was originally selected for another purpose.

I have mentioned some aspects of an RNA world before. The conjecture that the initial forms of life were based on RNA and not protein was an epiphany for many scientists. I worked on a Science Fair project in high school based on the prevalent theories of the time, that life started first from proteins. Lightning discharges in a reducing atmosphere produced amino acid precursors in the ocean. When these amino acids were splashed onto hot rocks, primitive proteins (proteinoids) with enzymatic properties could be created. In my project I mixed purified amino acids, melted them to simulate the hot rocks, and then dissolved the conglomeration in water. Depending on which amino acids were present, some very interesting objects called microspheres would be created. Under a microscope, they looked like primitive cells, were able to transport ions, etc. So here was a possible first step towards life. The problem was incorporating DNA, lipids, RNA and everything else we know as living.

RNA, while not being able to organize into a cell-like structure, has many other attractive properties useful for simple forms of life. It can provide the information storage needed for creating progeny. It can possess catalytic activities needed to replicate and sustain primitive life. Drawbacks include the relative inefficiency of these reactions and the instability of RNA. Proteins can help overcome both of these drawbacks. Perhaps the first functional RNA molecules were captured inside a proteinoid microsphere. The protein would help protect the RNA and the RNA could help organize the amino acids into protein. This sort of mutual cooperation sounds familiar. Kind of sounds like a ribosome.

The ribosome may be our best prototype for examining an RNA world. It appears very likely that the single catalytic activity of the ribosome, peptidyl transferase, is intrinsic to the rRNA present, not to any protein. The internal proteins of the ribosome appear to only provide structural stability. All the major aspects of translation (initiation, elongation and termination) can be performed in the absence of any external proteins. Initiation, elongation and termination factors involved in translation may only be present to provide stability and to increase efficiency.

In fact, some people have postulated that the first primitive ribosomes were actually RNA replicases, making new RNA from a template strand (see Poole et al. Bioessays 1999 21:880). tRNAs were used to bring together new ribonucleotides instead of amino acids. Now, there is still a large handwaving argument to allow the substitution of amino acids for ribonucleotides in this model, but there are some intriguing aspects to this theory. One is that the first primitive proteins may have been selected not for any novel metabolic activities, but because they mimicked some of the activities of the early RNAs. They were maintained because they did a better job than the RNA molecules they replaced. They could then have been selected for other uses, such as enzymatic activity. If true, we might expect to find proteins that look and act like RNAs. Well, guess what. We can find just such proteins and, not too surprisingly, they are intimately involved in translation.

An excellent review of this work was in a recent EMBO Journal. Aminoacylated-tRNAs (i.e. tRNAs already bearing the apprpriate amino acid at their end) bind to the A-site of a ribosome. The amino acid is transferred to the nascent polypeptide chain at the P-site; A for amino acid and P for polypeptide. Two elongation factors are used to efficiently move the mRNA molecule through the ribosome, by translocating aminoacylated-tRNAs from the A-site to the P-site. The first, EF-Tu, binds to GTP and an aminoacylated tRNA. It protects the ester linkage to the amino acid and helps position the tRNA properly at the A site of a ribosome. This reaction uses one of thephosphates from GTP. EF-Tu:GDP is then released and another factor, EF-G bound to GTP, acts to translocate the tRNA from the A site to the P site, allowing the next codon of the mRNA to be exposed at the A site, ready for another round.

The crystal structures for EF-G and an EF-Tu:aminoacylated-tRNA complex have been determined. EF-Tu and EF-G both contain a large number of structures that are very similar. These appear to be involved in GTP degradation since they show similarities with other GTPases. What is extremely surprising is the high degree of similarity between 3 of the domains of EF-G and the aminoacylated-tRNA. Looking at a space-filling representation is even more compelling. A major portion of this protein looks just like a tRNA. The main activity of EF-G is to displace a tRNA and translocate it from the A-site to the P-site. It seems logical to assume EF-G can efficiently displace a tRNA because its polypeptide nature provides more opportunities for enhanced interactions with the ribosome. It is a better "tRNA" than the tRNA.

EF-G is not the only tRNA mimic. Following termination of translation (due to the effects of releasing factors in the presence of an unoccupied stop codon), a ribosome is dissociated into its components. Ribosome recycling factor (RRF) is involved in the release of the mRNA and the tRNAs. If no RRF is present, the mRNA remains complexed with the ribosome and translation may proceed, producing aberrant proteins. The crystal structure for RRF has been published and, guess what?...it looks like a tRNA. Look at the overlap of structure in part (C). The RRF and the tRNA have almost exactly the same dimensions, with only the CCA-acceptor end of tRNA having no homologous structure in RRF.

A hypothesis: RRF binds to an empty A site, left after the releasing factors recognize the stop codon and terminate translation. EF-G then comes in and "translocates" the RRF to the P-site. The lack of any tRNA at either the A- or P-sites then destabilizes the 70 S ribosomal particle and it dissociates into its component pieces. Perhaps RRF looks so much like a tRNA because it needs to be efficiently compete with tRNAs for the A-site. But why?

It turns out that a significant fraction of all translation products in a cell are not full-length proteins but aborted intermediates comprising a peptide fragment fused to a tRNA. Upwards of 10% of the translation products can be these truncated chains, which have been prematurely released from ribosomes with the tRNA still attached to the nascent polypeptide chain. The cell has efficient mechanisms to degrade these but it seems like a waste of energy to produce so many useless molecules. The most likely possibility is that these aborted products accumulate as a consequence of enhanced accuracy of translation.

RRF is required for these products to be produced. Competition between RRF and tRNAs for the A-site could result in the release of translation products that are not accurate representations of the underlying mRNA. If RRF binds to the A-site first, perhaps because there is not any properly charged tRNA for the appropriate codon, the nascent polypeptide will be prematurely released and be degraded. However, it will not have a chance to misincorporate an improper amino acid. It is better to have a fraction of polypeptides that are prematurely terminated than to create a protein with improper amino acids incorporated. Without RRF, there would be more mis-translated proteins present. So the accurate mimicry of a tRNA by a protein could very well be a requirement for accurate translation.

In both of these cases, it is the structural similarity of the protein to an RNA that is the important underlying property, not an enzymatic activity. However, an activity could be useful at a later time and, while a protein might be originally selected for its shape, any subsequent activity could become more important. Both elongation factors have a GTPase activity that greatly increases their ability to affect the ribosome. In fact, this activity may be important for determining whether these proteins are in an active form that mimics a tRNA or in an inactive form that does not.

Although RNA molecules can exhibit many of the hallmarks of life, the instability of RNA and the limited repertoire of catalytic activities would make this inefficient. Proteins that mimic RNA, though, are much more stable that RNA and could include a much larger spectrum of activities that would be much more efficient. So, even though the initial selection of the proteins might be based on a similarity to RNA, inherent activities of the proteins could be co-opted for other purposes. This makes an interesting hypothesis which further investigations will explore. Many proteins may serve multiple masters.

This is one of the major mechanisms by which life evolves. A protein used for one purpose is shanghaied for another. In fact, if proteins or larger structures were not capable of multiple uses, life most likely would not have developed at all. Initial uses do not equate to future utility. Simple words for a pretty complicated idea. But just remember, simple or not, "Life goes on, braa."