⇐ Back to Articles, Essays, and Other Writing
A Brief and Opinionated Overview of What's Happening in the Life Sciences Today
By Greg Bear
The revelation that the human genome consists of about thirty thousand genes, and that more than two thirds of these are alternate splicing genes, which code for more than one protein, and sometimes many more, puts the final nail in the coffin of reductionist views of evolution and genetics. Stephen Jay Gould says as much in his opinion piece in the February 19 New York TIMES:
"The collapse of the doctrine of one gene for one protein, and one direction of causal flow from basic codes to elaborate totality, marks the failure of reductionism for the complex system that we call biology..."
Dogma after dogma has fallen in biology in the last thirty years. The central dogma-that one gene produces only one protein-died in the last decade with the discovery of alternate splicing. (Genes also produce other, non-protein products, such as ribosomal RNAs.) The sidebar to this dogma, which claims that DNA is read-only-implying that the genetic material changes only through random mutations, not through insertion or rearrangement of genetic material-collapsed some time ago with the discovery of mobile genetic elements such as transposons and retroviruses. Nevertheless, the Central Dogma is still mentioned, nostalgically, sometimes almost reverently, in new textbooks.
RNA editing after transcription, and before translation, in mitochondria and elsewhere shatters the assertion that DNA is the final blueprint of proteins. Interchange of genes between mitochondrial and nuclear DNA poses real problems for those who advocate the total dominance of DNA templates in translation to proteins. Genes, it appears, migrate in many different ways, and for many different reasons. To ascribe all this activity to random accident or undirected molecular impulse is simply obtuse.
The evolutionary relationship between some retrotransposons, the so-called "jumping genes" first discovered in corn by Barbara McClintock, and many retroviruses has been firmly established.
One-generation gross changes in morphology, reacting to environmental stimulus, has been conclusively demonstrated in water fleas, arabidopsis, and other organisms(i). Evidence of massive gene exchange in bacteria and archaea blurs any chance of establishing a useful evolutionary tree for these microbes.
It's been estimated that thousands of human endogenous retroviruses lie semi-dormant in our genome. They emerge in swarms in the placentas of pregnant women, but have not yet been shown to be transmitted laterally, that is, by infection, from individual to individual. I suspect that barrier will soon be broken, as well, leading to the possibility that some viruses operate as species-level retroviral infections, over tens of thousands or millions of years.
In one monograph on viruses(ii) published by the Santa Fe Institute, the editor goes so far as to suggest that we start thinking of viruses not as invariably pathogenic alien intruders, but as an extension of "self." Viruses can act as emissaries shuttling information between cells and between organisms. This is demonstrably true for phages in bacteria, and is now apparent in metazoans, even humans. Non-pathogenic viruses flood our tissues, and the ocean is filled with phages. Evocative? You bet. Are these viruses diseases, or commensals, or both?
In 2000 NATURE published an article(iii) suggesting that an ancient viral gene facilitates implantation of human embryos in the womb. Similar viral genes perform similar functions in other mammals. This suggests that genes from any source can become an indispensable part of an organism's genetic tool-kit.
The addition of a third color in the vision of great apes and humans bears the suspicious marker of retrogene duplication of an existing gene, which can be described as a random event only with great difficulty(iv) .
There are many traits being discovered that are passed to offspring not through the nuclear genes, but through cellular components, including maternal mitochondria, surface proteins, and prions.
Prions in yeast may serve a regulatory or innovation function in creating new proteins by allowing read-throughs across open reading frames.
The unfortunate aspect of the rancorous debate on evolution in the last seventy or more years has been the fossilization of hypotheses. One side says "God and God only," the other says, "Random mutations and natural selection and nothing more." Both are likely wrong. A third variety of "intelligent design" has long been awaiting our attention.
In my novel, DARWIN'S RADIO, published in spring of 1999, the vector of evolutionary change is an infectious human endogenous retrovirus-an idea derived from the curiously self-directing and self-regulating evolution of bacteria through gene exchange by phages and plasmids.
My hypothesis: through communication by pheromones, viruses, and sexuality, and through incorporation, selection, and editing of complexes of genes by a linguistically based and computational DNA, the genomes of individuals become part of an extensive, species-scale neural network that solves problems on a much vaster scale than science has ever anticipated.
Proposed changes in morphology are communicated through sexual activity and retroviruses and stored up in populations in a genetic "set-aside" area within each individual. A library of records of past adaptations is used to "judge" new phenotypic proposals within the genome, individually and across the species. Possible variations are selected and edited extensively based on evidence culled from the environment by the immune system. (The sophistication of the immune system has inspired some scientists to refer to it as another brain.)
The decision within a species to produce a new type of organism, or subtly modify aspects of an old one, is made using genomic rules we have yet to understand, but which are likely similar to the rules that also allow clusters of neurons, including brains, to solve problems that confront organisms in the environment.
Potential and real pathways for all these interactions have been shown to exist.
When environmental challenges arise, morphological changes are enacted in "suites" of mutually advantageous mutations. To avoid contaminating older populations, or to avoid being contaminated by them, speciation occurs, and the new organisms are allowed to compete as a more isolated breeding population in the arena of nature.
Thus, speciation can occur in bursts rather than over geological time, leading to "punctuated equilibrium." In many instances, gentler modifications occur, within geologically separated populations that, for a time at least, can still interbreed. Evolution is hardly a one-trick pony-or dog!
A species-level network is also closely tied to larger networked systems, ecologies. Ecologies "recruit" and even alter species over remarkably short periods of time, pointing toward evolutionary collaborations between co-dependent species that would have been thought ridiculous in the recent past. Such recruiting and adaptation has little to do with natural selection per se, though of course the environment, and survival, are the final arbiters of the resulting designs.
A conference of molecular biologists held in 1998 reached conclusions similar to mine. Their papers were published in 1999 in a thick report edited by conference chair Lynn Caporale, Molecular Strategies in Biological Evolution, Annals of the New York Academy of Sciences, V. 870.
I have been invited to discuss these possibilities with biologists, most recently at the After the Genome Conference in Tucson, Arizona, chaired by Roger Brent of the Molecular Sciences Institute. There I met with biologists, computer programmers and engineers from MIT and a number of private companies who are being recruited to examine the nature of computational DNA, and to provide clues as to where to look next. While I do not claim that these scientists share all of my views, the discussion was open and very stimulating. The revolution is well under way.
Some refer to this burgeoning new view as "systems biology." For many conservatives in biology, the changes are heartbreaking, even infuriating. But the evidence has been mounting for decades, and clear signs of the necessity for radical change has been evident for over fifty years. Arthur Koestler fought reductionism in psychology and biology from the 1950s to his death.
We're facing a true paradigm shift. Is that surprising? Did anyone actually believe we had all the answers to something as marvelous and complex as life and evolution?
That organisms exchange genes through other than sexual means is now irrefutable for metazoans as well as microbes. Retroviruses, and now perhaps bacteria, may well serve as vectors for such exchanges. Commensal bacteria in our intestines commonly interact with our tissues. Surprisingly, there appear to be mechanisms in most organisms for evaluating and either destroying or utilizing RNA from outside sources, including retroviral sources. This evaluation process is extremely important, and understanding it may be key to understanding how the genome works in both individuals and in populations.
The "selfish gene" is certainly a valid concept in some instances, but not in the vast majority. Rather, because genes rely on interaction with many other genes-hundreds in some cases-to be effective, they are less like competitive rogues than tame office-workers. The "social gene" becomes a better model. And in fact the social aspects of the genome have been championed for decades by brave molecular biologists and geneticists, including Lynn Margulis.
Altruism in societies is well demonstrated, and rationally quite defensible. That genes operate in their own societies, and that species both compete and collaborate in those larger societies called ecosystems, functioning as nodes in an extended neural net, makes the problem of cooperation and altruism far more tractable.
Random processes are also at work in evolution, quite clearly, leading to either uncorrected errors or serendipitous discovery-but I do not think that we can any longer support random mutation as the sole cause or even the major cause of variation. Darwin himself deliberately avoided subscribing to chance as the sole cause of variation, thus leaving the actual cause to be discovered in the future. Later generations leaped in well before the facts were available, and cemented the hypothesis, slowing the pace of biological discovery by actively discouraging alternatives.
A similar reductionist slow-down happened in psychology with Behaviorism, whose central tenets are now largely discredited.
It is likely that medical research could have made more progress in combating retroviruses such as HIV if we had been less attached to near-religious dogmas, more willing to play with new hypotheses in the face of experimental evidence. The paucity of hypotheses in biological science may be something of an intellectual crime, perpetrated by academics protecting their own fiefdoms against assault by barbarian unbelievers-hardly an atmosphere in which to raise and tutor new generations of biologists.
These bills are now coming due. Careers and reputations are about to be massively re-evaluated, and new careers are about to be established, correcting the failures and misconceptions of the past, and perhaps correcting some unfortunate aspects of the culture of medicine and biology.
This is the way science should work, rather than stubbornly defending demonstrably incorrect dogmas, or trying to correct them with endless epicycles based on vague concepts. "Emergent properties," "complexity," and other buzzwords arouse my suspicions because they could serve to hide our ignorance of the basic processes at work in cells, organisms, and throughout nature-processes very similar to the processes in our own brains. These processes involve the encoding of information in morphologically based neural networks, where the "neural nodes" can range from genes to organisms in a species to species in ecosystems, and the use of that information to alter patterns structure and behavior.
To be sure, the theories behind neural networks remain unfinished and difficult at best, and there's much work left to be done. A key question is whether genetic processes are formally describable, as some systems biologists believe, or more closely related to natural languages, as some computer programmers believe.
Information theory will be a rich lode for the exploration of all these topics, but must be used in the right way-and that way, frankly, is not at all clear at the moment. We understand far less about the users of information, and how information is converted into anatomy in living systems, than we do about how information is transported.
The debate on this is going to be enjoyably fierce, and we are going to encounter many more roadblocks in understanding how life works until these matters are resolved, or new and better questions are asked.
But we know the brain works. We know that ants and bees cooperate to solve problems beyond the individual. We know that bacteria form complex communities, social groupings, as in biofilms, to take advantage of resources or face environmental challenges. Community interaction and problem-solving seems to be a key behavior in nature, as important as competition.
What most researchers are seeing, around the world, is this: Too few genes, too much interaction, too much intergenomic activity, and rapid adaptation, much of it, perhaps most of it, highly regulated and organized.
As Lynn Caporale has pointed out, even evolution itself has evolved! And now it is much more efficient and directed than it was, say, in Archaean times. This may explain the immense lag between the first cells and complex organisms: learning how to evolve effectively!
There may indeed be teleological and intelligently directed evolution, but we do not need to blame it on God. DNA itself may be creative, and in its own way, goal-seeking and problem-solving.
This allows evolution to proceed in a more rational fashion, and explains much of what has been observed in the fossil record, and nearly all of what has been observed in living organisms.
Time and good science will tell. As always.
(i) Nature 401, 60 - 63 (1999), Transgenerational induction of defences in animals and plants, Anurag A. Agrawal, Christian LaForsch, Ralph Tollrian Observations of morphological variations in Daphnia go back to Woltereck in 1908. See also S. Dodson, Predator-induced reaction norms, BioScience 39, 447-452.
(ii) VIRAL REGULATORY STRUCTURES AND THEIR DEGENERACY, edited by Gerald Myers, 1998, Addison/Wesley-Santa Fe Institute Studies in the Sciences of Complexity, vol XXVIII, page 2.
(iii) Nature 403, Number 6771 785 - 789 (2000) Syncytin is a captive retroviral envelope protein involved in human placental morphogenesis , SHA MI, XINHUA LEE, et al.
(iv) PROTEIN EVOLUTION, by Laszlo Patthy, Blackwell Science, 1999. Patthay discusses work by Yokoyama and Yokoyama, Molecular evolution of visual pigment genes etc., in Population Biology of Genes and Molecules, eds. Takahata and Crow.