Part 2 on The Plausibility of Life

How does evolution shape living things? The fact that evolutionary forces, such as natural selection, can shape living creatures is well-established, but how malleable those creatures are, and what the increments of change are is less well established. We have a fairly good idea of how genes can change, but how does that genetic change translate into physical changes in the shape and functioning of the organism itself - that is, how does genetic change translate into changes in the organism's phenotype?

The authors of The Plausibility of Life, Marc Kirschner and John Gerhart, argue that this issue has been ignored in evolutionary theory (although they go on to say that it was justifiably ignored for a long time - before modern molecular and cell biology, there was no way to effectively address this question):

What if evolutionary biologists were wrong to think of phenotypic variation as random and unconstrained? How much would it matter if we really understood how genetic variation leads to phenotypic variation, and in particular, how facile or difficult is it to achieve a specific phenotype?

These questions get to the heart of the evolution of complexity. For example, is the mollusc lineage infinitely malleable? If webbed feet were to provide molluscs with an adaptive advantage, would natural selection be able to, after many generations, produce them? (Keep in mind that when talking about natural selection, the semantics get tricky - I don't mean to imply that natural selection works towards a focused goal like webbed feet. I'm not assuming that it does in these arguments, but we occasionally bump up against the limits of language, or at least my language, when trying to talk about these issues efficiently.)

Here is another way that Kirschner and Gerhart state the question. Take a line from Shakespeare, "To be or not to be", and imagine a computer randomly generating letters on a screen - how effectively would a computer generate Shakespeare's line? (Again, the analogy to evolution is imperfect - there is no pre-set goal in evolution analogous to a pre-determined sentence.)

If the computer was varying all letters randomly, and all correct 18 (including spaces) had to show up simultaneously at once to get the line, it would take some time before you would expect to see Shakespeare's line show up by sheer chance. Now if you added a little selection, your line would likely show up much sooner - if every time you got a correct letter, you kept it, only varying the other letters, you'd generate the line much faster.

But what if you didn't vary letters, what if you varied entire words? Instead of having the computer choose at random 18 different letters, what if it randomly chose six 2- or 3-letter words? If you used selection together with random word variation, you'd get Shakespeare's line very quickly. In the case of individual letters, you can get many "non-functional," gibberish combinations - that is, the first hurdle is to get a functional word, and then put those words into a sentence.

Kirschner and Gerhart argue that the variation we observe in real organisms is like variation of complete words, not individual letters. In organisms, there are functional pieces ready to go, pieces which can be varied as coherent plug-and-play modules to build new variants of complex systems.

Much of this probably sounds obvious, but its significance for evolution is often unappreciated. It is much easier for an insect or vertebrate lineage to evolve new types of limbs, for example, than it is for a mollusc to evolve webbed feet. Why? Because insects and vertebrates already have the molecular toolkit for making limbs; to make a new type of limb (such as wings on a bat or dragonfly) requires only some tweaks to that tool kit - bats and insects did not have to evolve wings from scratch.

Like I said, this all probably sounds obvious, but what is not obvious is how pervasive these ready-to-go molecular toolkits are. Almost all of the most important cellular processes evolved very early in the history of life, and much evolution since then has consisted of applying the fundamental organism-building toolboxes in new ways. A good deal of our metabolism is very similar to what exists in bacteria. Our cells divide using the same machinery employed in yeast cells. Biologists study embryological development in flies because the machinery that operates there is extremely similar to what operates in human embryos.

This similarity of basic cellular processes was a huge surprise to many prominent evolutionary biologists. Ernst Mayr, one of the most successful and influential mid-20th century evolutionary biologists (and pretty damn smart, writing sharp books at age 100) wrote that:

Much that has been learned about gene physiology makes it evident that the search for homologous genes is quite futile except in very close relatives. If there is only one efficient solution for a certain functional demand, very different gene complexes will come up with the same solution, no matter how different the pathway by which it is achieved. The saying "May roads lead to Rome" is as true in evolution as in daily affairs.

- Animal Species and Evolution (1963), p. 609, quotes in Sean Carroll's Endless Forms Most Beautiful, p. 71-72

This was how Mayr thought that new complexity evolved: evolution built new genetic tool boxes from scratch, over and over again. Today we know that this idea is wrong in most cases. It is actually much easier to evolve new complex functions than was believed at mid-century.

Thus, for a long time evolutionary theory went along without any serious account of how genetic variation produces complex phenotypic variation - how new combinations of mutations produce new complex functions.

For a long time evolutionary biologists could easily neglect molecular biology, but the phenomenon has run both ways - most molecular cell biologists have very little training or interest in hard-core evolutionary theory. They reap frequently reap its benefits, but don't spend much time thinking about how their own work fits into the larger picture of evolutionary theory. Kirschner and Gerhart argue that the way forward in evolutionary biology involves overcoming this neglect. The core cellular processes, conserved for immense stretches of evolutionary time, are the enablers of much of the complex function we see in nature today.

This is the second installment of a series of posts on an interesting recent book by the accomplished biologists Marc Kirschner and John Gerhart. In this book, the authors lay out what they see as the most important research agenda for molecular biologists in the 21st century. Part 1 is here. Part 3 is now up