by
(Issue 2; posted February 20, 1997; archived March 6)
Obviously, these summaries are both extreme parodies of how most scientists work, but I nonetheless believe they contain a nugget of truth. Therefore, I had hoped that an Internet forum might lead to a "meeting of minds" that would shed light rather than encourage controversy (this despite my own observation that newsgroups seem to create endless controversy).
The end product of the Great Debate was, unsurprisingly, both similar to and completely different from what I had imagined. On the one hand, I don't think there have been many in-depth scientific discussions that have approached the level of interaction that we can view in the debate. On the other hand, neither side seemed willing to budge an inch, to give credence to opposing points of view. I suspect the discussion remained static partly because written exchanges without verbal and nonverbal clues are frequently interpreted as hostile. Another part of the deadlock may be because scientific views are often passionately held, and belief may sometimes outweigh objectivity.
But I think a large part of the controversy may arise from the subject rather than the venue. The debate over intron origins has been with us for so many years because it is fascinating and unresolvable.
Unless someone invents a time machine, there is no way to demonstrate that introns are old. None. We could have the sequence of every genome of every organism on the face of the planet and we would still be unable to extrapolate beyond events that occurred in the progenote, the last common ancestor of modern life.
Given that the last common ancestor was already an extremely complex critter, all of the evidence necessary for even a minimal phylogenetic extrapolation back to the time of the origin of protein-encoding genes died several billion years ago. Many, many scenarios can be imagined, either with intron-endowed pre-progenotic organisms or without them.
Therefore, folks who work in the 'introns early' field can pretty much say whatever they want and there will be no one that can prove them wrong. The original power of the 'introns early' argument was that introns fostered recombination, and this power remains in the form of obvious examples of exon shuffling. But explanatory power is far different from truth, especially given that there are many equally good ways to construct and evolve protein coding sequences without introns.
The original intellectual rationale for believing that introns were old quickly spawned a more social one: it gave molecular biologists something to write about in the Discussion sections of papers. If you sequenced a gene, and introns were known to divide up functional domains of proteins, then you could talk about the putative functional domains of your protein, in addition to just circling all the various regulatory boxes.
In the modern era, arguments about intron origins bizarrely piggyback on arguments about an equally unprovable hypothesis, the RNA world (and I say this as an avid RNA world theorist and experimentalist). This association of tenuous hypotheses does nothing to improve the viability of either one.
In this, I disagree strongly with Russ Doolittle (see his synopsis) that the heart of the matter is, "How were the earliest genes derived from scratch? If there was a transition from an RNA world, how could introns not be involved?"
Well, easily. While speculations about the RNA world and introns are interesting, there are a dozen other speculations about the RNA world that are equally interesting, equally compelling, and have nothing to do with introns. Straining to tie the origin of introns to a model for the evolution of metabolism that is not (and cannot) be proven does not advance what we truly know. It merely invites speculation for its own sake. While intellectual curiosity may drive science, it is not science. There are some things that are just not knowable by empirical inquiry, and intron origins may very well be found within that set.
Similarly, there is no way to conclusively demonstrate that introns are young. Despite the best arguments of parsimony, it may be that the blah intron in the blah gene in the blah organism just happens to be 3.75 billion years old and has been spliced continuously during its progression from random sequence spacer to ribozyme spacer to protein spacer.
The arguments for 'introns early' rest on numerous assumptions. For example, all arguments about phase correlations rest on the assumption that phase of insertion is unimportant for protein evolution beyond serving as a marker for the recombination of exons. This may be true, but it is still the sort of assumption about which we have very little data.
Likewise, the arguments for 'introns late' rest on equally unprovable but unfortunately different assumptions, mostly regarding the rates and mechanisms of evolution in different branches of the phylogenetic tree. Just because some lineages have apparently lost introns at a fantastic rate while others have not does not mean that we can speculate freely on what happened in unknown or vanished lineages. At best, the proponents of 'introns late' can paint a self-consistent and probabilistic picture of intron gain that will have to serve as the next best thing to experimental proof.
However, it is not clear how such a picture provides insights into modern biology beyond the argument over the origins of introns. The genomes of deep-branching organisms must still be examined individually for introns, the use of introns in constructing mosaic proteins is obviously a useful evolutionary tool for modern organisms whether or not introns are old, and the relationship or lack thereof of spliceosomal introns to Group II introns is a question whose answer will shed light on RNA catalysis and evolutionary mechanisms regardless of intron antiquity.
In short, if Walter Gilbert had never postulated either the antiquity of introns or their relationship to a RNA world, would there be any other rationale to support the phylogenetic study of intron distribution? Is 'introns late' merely a creature of contrary disposition? Is this much ado about nothing?
In the end, the problem with the "Great Debate" is that one side is painting a beautiful picture that bears little resemblance to reality, while the other side has gone about collecting data that may contradict the opposing perpsective, but does not appear to have a fundamental import of its own.
So, now that I have likely offended everyone, I have a suggestion: what would happen if each side tried to make the best possible case for the opposing side? That is, can we, as scientists, play Devil's advocate and attempt to hone our own understanding of how we view the data and the theories that underlie the data?
What I would like to see is the 'Great Debate' played out in reverse, where the sole goal is not to have your arguments hold sway, but to hone the best possible arguments for all the data and to come up with the best possible experiments to test those arguments. For example, I really want to see someone introduce a robust self-splicing intron containing a reverse transcriptase gene into an intron-sparse species and experimentally observe what happens, rather than attempting to figure out and pontificate about what may have happened millions or billions of years after the fact.
Andrew D. Ellington is Assistant Professor in the Biochemistry Department at Indiana University, Bloomington. His current projects address the molecular evolution of new nucleic acids, new proteins and new organisms.
This article is posted as the introduction to the BioMedNet discussion group the established for readers of the intron debate.
Debate participants have responded:
"...Making the living past tangible with images of the present is evolutionary biology's burden. Introns? If we never will agree whence introns arose, we can shrug it off disgruntled. But biology needs an answer when children or taxpayers ask, 'where do genes come from in the first place?'" - from William Martin's response to Andy Ellington"....at the risk of offending everyone that Andy has not already offended, those who claim that nothing can be inferred about the past, that it is all "pontificating", are like novice fisherman who claim that the fish are too rare or too smart to be caught. They should hone the necessary skills, or step aside and let the experts take over. The first requirement to become an expert who can contribute directly to an understanding of evolutionary history is to abandon the 'model system' approach used by molecular biologists and biochemists in favor of systematic methods used by evolutionary biologists." - from Arlin Stoltzfus' response to Andy Ellington
Please join the discussion! Or, tell us your thoughts on Dr. Ellington's opinions via Feedback.
