On the Counterintuitive evolutionary truths thread, I expressed amazement at the sheer number of distinct kinds of intragenomic conflict that have been discovered by science. In response, Allan Miller recommended the 2006 book Genes in Conflict, by Austin Burt and Robert Trivers. Burt’s name is unfamiliar to me, but Trivers is famous for proposing the theory of reciprocal altruism.
I ordered the book (
28 for the paperback), and so did Gralgrathor, so I thought it would be nice to have a discussion thread for the book as we read it. Anyone is welcome to join in, of course, whether or not you are reading the book.
The opening passage of the book will be an eye-opener for anyone who still thinks that natural selection invariably works to maximize the fitness of the organism:
The genes in an organism sometimes “disagree” over what should happen. That is, they appear to have opposing effects. In animals, for example, some genes may want (or act as if they want) a male to produce lots of healthy sperm, but other genes in the same male want half the sperm to be defective. Some genes in a female want her to nourish all her embryos; others want her to abort half of them. Some genes in a fetus want it to grow quickly, others slowly, and yet others at an intermediate level. Some genes want it to become a male, others a female — and the reason they want it to be a female is so that a quarter of her fertilized eggs will be defective! In plants too there can be internal conflicts. Some genes want a plant to make both pollen and seeds, others only seeds. Some genes want the plant to allow a particular pollen grain to fertilize an ovule, and others want to kill that pollen grain.
And how should an organism manage its DNA? Some genes want to protect chromosomes from damage, while others want to break them. Some genes want the organism to snip out bits of DNA and insert them elsewhere in the genome; other genes want to stop this from happening. Some genes want to activate a particular gene, and others want to silence it. Indeed, in the extreme case, some genes want to inactivate half of the genome, while the targeted half prefers to remain active.
Some of these genes are known from only a few species, others from virtually all. In addition there are genes whose existence is predicted, but no yet confirmed, such as genes in fathers to favor their sons versus genes to favor their daughters. Or genes that want a female to judge her mate as more attractive than he really is versus genes that want the opposite effect. And so on.
These conflicts arise because genes are able to spread in a population despite being harmful to the larger organism. Such genes give themselves a benefit but typically cause negative effects on other nonlinked genes in the same creature. In that sense, they are selfish. Indeed, we can define selfish genetic elements as stretches of DNA (genes, fragments of genes, noncoding DNA, portions of chromosomes, whole chromosomes, or sets of chromosomes) that act narrowly to advance their own interests — in other words, replication — at the expense of the larger organism. They, in turn, select for nonlinked genes that suppress their activity, and thereby mitigate the harm. That is, the evolution of selfish genetic elements inevitably leads to within-individual — or intragenomic — conflict. This occurs over evolutionary time, as genes at different locations within the genome are selected to have contradictory effects. It also occurs over developmental time as organisms experience these conflicting effects. In this sense, we speak of “genes in conflict”, that is, genes within a single body that are in conflict over the appropriate development or action to be taken.
D’oh! Why do I keep calling it Genomes in Conflict? 😀
This is, incidentally, substantially a story of sex. Many of these elements operate because they have a chance to distort the Mendelian 50/50 ratio in meiosis. Others operate due to the gender-related asymmetry of mitochondrial transmission, still others because of gender asymmetry between parents in optimal outcomes. We diploids have three genomes in our bodies, not one, and they are always scrapping!
My favourite element, the homing HEG, simply nicks its homologue at its own relative position. The DSB repair machinery kicks in to repair the gap, must use the HEG sequence and – voila! – the HEG gains a copy. But this has striking similarities to the mechanism of DSB initiation during crossover in meiosis. That is, crossover could have been initially non-adaptive, simply a result of a selfish gene tricking the repair machinery.
Allan Miller,
Makes your head hurt to think about it, doesn’t it? As when I first read The Selfish Gene, with this book too I keep wanting to grasp for some teleological handle on what’s happening. My brain keeps insisting that causation and intent are somehow inextricably connected. Bad brain! But I can understand why it’s so difficult for our ID-otic brethren to grasp philosophical naturalism…
Gralgrathor,
I think Dawkins pitched it about right, and I have found the ‘gene-centred’ view a useful way of looking at the matter ever since. It can be over-applied, but as exponential increase (achieved through nucleic acid’s complementary nature) is very much at the heart of the higher-level dynamics, I think it’s the ‘right way round’.
But it amazes me how often this viewpoint is misunderstood, and not just by the ‘opposition’. Even our gracious hostess has, I feel, too jaundiced a view of the connotations of the concept. And Gould? Huh!
For the real head-hurting stuff, look to the complexities of fungus gnats, scale insects and fig wasps. Obviously designed by committee.
Allan,
You like drama. A full-scale war is more exciting than a mere skirmish. 🙂
Allan,
Adding to the confusion is the fact that the ‘selfish’ in ‘selfish gene’ is broader than the ‘selfish’ in ‘selfish genetic element’. Dawkins regarded all genes as selfish in the sense of ‘trying’ to get themselves replicated. The term ‘selfish genetic element’ seems to be restricted to those elements that get themselves replicated without providing any benefit to the host organism.
Allan,
Responds the Designer: I meant to do that.
That’s crazy. How long until they find the Ask Toolbar in the human genome?
Allan,
I skipped ahead to the chapter where they talk about HEGs and found this:
That sounds a little scary to me. Wouldn’t horizontal transmission be a possibility for an artificial HEG?
Yikes! On the face of it, that may be the worst idea I have ever heard. And I have, in my time, read some kf.
keiths, DNA_Jock,
Yep, I tend to think ‘Sorceror’s Apprentice’ whenever I read about fancy ideas for control. That’s the trouble with releasing Design into the wild, things are likely to come up you hadn’t thought of. Unless you’re super-duper Intelligent, of course.
B & T say that horizontal transmission is necessary for HEG persistence, since successful ones will rapidly saturate their species and degenerate for lack of targets. One factor that reduces worry in the example they give, though, is that their engineered HEG is deliberately associated with a gene. Transferred between species, the gene association would disappear, since I don’t think an HEG can cross a species boundary while at the same time retaining its precise chromosomal position. It only stays in place within a gene pool, infecting homologues during synapsis in meiosis or in diploid cells, which doesn’t occur between reproductively isolated groups. Travel between them would be by a non-HEG mechanism (eg virus), and would likely not be position-dependent. You can only ‘home’ when you get a close shot at your target – ie, when you are in the same cell.
I read further, and they at least acknowledge the danger later in the chapter:
Before I was banned from UD I was arguing that design is impossible except through evolution. An exaggeration, but not, I thing\k, a great exaggeration.
What this threat highlights is the high dimensionality of design. there are thousands, perhaps millions, of dimensions to fitness.