Illustration by Alison Schroeer
In 1910 Thomas Hunt Morgan found a single white-eyed fly among the teeming thousands of normal red-eyed flies he raised in his lab. He nursed it to adulthood—legend has it that he kept the delicate new fly by his bedside until it was strong enough to breed—bred generations of new flies from it, and found that the white-eyed trait could be propagated in fly after fly. This is the canonical model for how new traits appear: A spontaneous mutation causes a new phenotype, natural (or in this case, artificial) selection can then do its work, and the trait spreads through the population. The genotype precedes the phenotype.
Can it work any other way? Can the phenotype appear, and be selected for, before a specific gene for the trait is present? It sounds heretical, and the idea is mildly controversial, but there is a process called genetic accommodation that provides a mechanism for novel phenotypes to precede a specific genotype.
First, let's consider a specific recent example, some lovely work by Suzuki and Nijhout, as a concrete instance of the phenomenon. Then I'll explain the mechanism behind this curious process.
Advertisement
Gardeners and entomologists are familiar with the hornworm, the impressively large caterpillar of the moth, Manduca. There are several species, and one, Manduca quinquemaculata, exhibits an interesting polyphenism: It has two possible forms, with the larva either turning entirely black if it is raised at cool temperatures (the better to absorb heat from the sun), or all green if raised at warmer temperatures (the better to hide in foliage). Another species, Manduca sexta, exhibits no such flexibility. The wild type is always green, no matter what the temperature, and a mutant form is always black, again without regard for temperature. Suzuki and Nijhout wondered if they could give M. sexta the polyphenic abilities of M. quinquemaculata and carried out a clever experiment in evolution in the laboratory.
Development is a plastic process in which organisms respond not just to a genetic program, but also interact with the environment. In particular, stress can trigger strong molecular responses to better enable cells and tissues to carry on; one common experimental stressor is heat shock. Raise the temperature of the developing organism enough to force it to struggle to cope, but not enough to seriously injure it, and sometimes surprising and unpredictable changes occur. Note that these changes are not to the genetic material, but to the developmental processes—the phenotype alone.
Warming the developing larvae of the always-green form of M. sexta to 42°C for a few hours has no effect on the animal's color. The always-green form seems to be strongly stabilized and unresponsive to change. The same treatment given to the always-black form, though, triggers a wide range of responses: Some stay black, others turn green, others exhibit varying degrees of mottling and color. Variation, of course, is the raw material of evolution, which suggests the obvious experiment—let's try selecting from this pool of variant forms.
So here's the experiment: Take 300 black caterpillars and hit them with a toasty, uncomfortable heat shock of 42°C for six hours (this is not an outrageous temperature—it's akin to a particularly hot subtropical summer day). Afterward, pull out the caterpillars that turn the brightest green, and raise them to adulthood. This is selection for those forms that most readily switch colors and are going to be the progenitors of a polyphenic line. Similarly, pull out the larvae that resist changing color and stay black—this will be the monophenic line. Repeat for generation after generation, always selecting the polyphenic line for the individuals most likely to change color, and the monophenic line for individuals most able to resist color changes. Of course, you must also set up another line, a control, in which the individuals selected for the next generation are picked randomly without consideration of their color.