In 1891, a young embryologist named Hans Driesch carried out an experiment whose results were so startling that they changed his entire outlook on life. Working at the Stazione Zoologica in Naples, Italy, Driesch was studying the development of sea urchins. Following the first cell division of the early embryo, he carefully separated the two cells and watched them develop. To his great surprise, each cell produced a perfectly symmetrical and anatomically complete larva, albeit at half the normal size. The opposite experiment, combining cells from two embryos, produced double-sized but otherwise normal larvae. So astonished was Driesch by these results, that he abandoned experimental biology and turned to philosophy for the remainder of his professional life.
Driesch’s surprise is understandable. It is quite natural to think of embryos as an especially delicate phase of the life cycle. Surely any disruption in an early embryo will be amplified during subsequent phases of development, wreaking greater and greater havoc and eventually leading to death? This intuition lies at the heart of a recent critique of evolutionary biology by Paul Nelson, a Senior Fellow at the Discovery Institute. He argues that mutations that act in early embryos are invariably lethal, thus rendering impossible the evolution of large-scale changes in anatomy.
Good question: can mutations generate new body plans?
In order to understand Nelson’s argument, let’s start by examining the claim that he seeks to refute, namely that all animals are the descendants of a common ancestor. Nelson finds this claim problematic because animals have a wide range of body plans. The term “body plan” lacks a precise meaning, but it generally refers to features such as symmetry, the presence or number of limbs, and the location of sense organs. Consider a fish and a starfish: one has a head and bilateral body symmetry; the other lacks a head and has five-fold symmetry. What kinds of mutations, Nelson asks, could produce such large differences in anatomy?
The reality is that there exists no firm relationship between when a mutation acts in the life cycle and the extent of the damage it can cause. By the same token, the opportunity for a mutation to alter a trait in a way that might be adaptive is also not limited to late development.
The answer is that no mutation can turn a fish into a starfish. That’s not only implausible, it’s inconsistent with descent from a common ancestor. What evolutionary biologists claim is that a series of mutations altered the body plans of descendants of their common ancestor. But, argues Nelson, any such mutation would need to influence the patterning of the embryo and would therefore have a cascading and catastrophic effect on later development, inevitably causing death. Mutations affecting late development are not necessarily fatal, he posits, because of their limited scope of play, yet this also renders them powerless to alter body plans. If correct, this “unsolved problem of macroevolution” undermines the claim that animals with different body plans evolved from a common ancestor.
Embryos are “astonishingly resilient” to disruption
But is the premise correct? We now know that Driesch’s experiment with sea urchins produces exactly the same result in ourselves as it does in sea urchins. Strike up a conversation with an identical twin, and you will find yourself speaking to someone who is in every sense a complete person, even though they developed from just half of an embryo. Nine-banded armadillos go a step further: most litters are identical quadruplets derived from a single fertilized egg. Fertility clinics routinely remove a single cell from human embryos in order to check for genetic defects before implantation. It doesn’t matter which cell is removed—the later fetus suffers no damage.
Embryos are astonishingly resilient to disruption. There are limits, of course. Birth defects can result from genetic mutations or environmental exposures. But perhaps less appreciated—and truly remarkable—is the enormous range of mutations and conditions that leave no discernible trace. This is because robust feedback mechanisms operate in embryos, constantly sensing and reacting to changing conditions. Rates of cell division, cellular migration pathways, the formation of neural connections, and many other vital processes in embryos are not “hard wired” but are instead modulated with exquisite precision.
Thousands of experiments performed in hundreds of laboratories on dozens of different animal species reveal a beautiful and potent property of embryos: they adjust. Returning to sea urchins, more recent experiments have plucked out four cells of early embryos that are the embryonic precursors of the cells that later produce the skeleton. This treatment causes a slight delay, during which a different group of cells reprogram themselves, migrate to the correct location, and construct a perfectly normal skeleton.
As amazing as this adjustment is, however, it is a response to a physical change. What about genes? A cascade of interactions between several genes endows the skeleton-forming cells of the sea urchin with their unique identity. This is, in fact, the earliest patterning step in the sea urchin embryo, the first “decision” regarding cell fate. Experimentally interfering with the function of some of these genes halts the normal production of skeleton-forming cells. But again, the embryo adjusts. A reconfigured set of interactions among these genes reprograms the fate of the distinct population of cells mentioned earlier, and they go on to fabricate a skeleton.
Parallel experiments with embryos of other species often reveal similar results. The conclusion is clear: genes with critical roles in early development can in some cases be mutationally or experimentally disrupted without causing permanent harm, much less death. To be clear, this is not a universal outcome, but it is also far from being rare. Any argument that rests on negative evidence is vulnerable to contradictory evidence. Much as the positive evidence of fossils undermines a “God of the gaps” claim based on negative evidence, so too mutations that alter embryonic patterning but are not lethal undermine Nelson’s claim based on negative evidence.
Contrary to Paul Nelson’s claim, there are many examples of non-lethal, adaptive mutations.
Nelson’s argument seems logical, but the premise is demonstrably incorrect. Mutations affecting early development do not necessarily wreak havoc. (His claim that mutations affecting late development are necessarily limited in scope is also incorrect but less germane.) The reality is that there exists no firm relationship between when a mutation acts in the life cycle and the extent of the damage it can cause. By the same token, the opportunity for a mutation to alter a trait in a way that might be adaptive is also not limited to late development.
Snakes provide a striking example. The limbless body plan of snakes opened a wide range of ecological opportunities such as burrowing, constricting, and moving efficiently through landscapes with narrow passageways. Snakes have radiated spectacularly into diverse habitats, lifestyles, and anatomical specializations as a direct consequence of their modified body plan. Molecular geneticist Marty Cohn and colleagues demonstrated that the mutations that produced limblessness in snakes do so by altering the expression of a gene called Shh. This gene operates in embryos, including humans, where, among other tasks, it establishes the position of the four limb buds (or fin buds in the case of fishes). A change in the location of where Shh functions in snake embryos prevents the establishment of limb buds, neatly excising a specific sub-program of development. Meanwhile, other functions of Shh, including its critical role in patterning the brain, remain intact.
So much for deleting body plan features. What about rearranging and reshaping them? The genes most closely intimately involved in patterning the body plan are known as Hox genes. Changes in Hox gene function in embryos are responsible for evolutionary changes in appendage number and anatomy in crustaceans, the transformation of fins into limbs in vertebrates, a reduction in the number of body segments associated with miniaturization in tardigrades, a vast expansion in the number of rib-bearing vertebrae in snakes, and the difference between eight legs in spiders and six in insects. In each case, the effects on anatomy are restricted to particular organs and body regions. Furthermore, mutations or experimental manipulations of Hox gene function alters the body part in question without wreaking wider havoc, much less causing death. For instance, disrupting the function of the Hox gene Antennapedia-1 results in a spider with 10 rather than the usual eight legs which, notwithstanding, walks, feeds, and mates as normal.
High degree of modularity of gene function facilitates evolution of anatomy
This high degree of modularity in gene function—a situation where a gene’s effects are tightly circumscribed to a specific region or cell type in the embryo—is overwhelmingly the rule rather than the exception. Indeed, this is one of the most important discoveries from research in developmental genetics over the past several decades. Modularity of gene function is one of the reasons embryos are so resilient to mutation and to environmental influences. It is also a feature of every group of animals whose development has been studied.
Far from being a barrier to body plan evolution, genes that operate in embryos actually facilitate adaptive changes in anatomy. The highly modular action of these genes is not only vital to the way they work in embryos; modularity increases the likelihood that a change in the function of a developmental regulatory gene due to mutation will alter one feature of anatomy without affecting others. This property is so well documented that evolutionary biologists now routinely turn to developmental regulatory genes as the “usual suspects” when studying the genetic basis for the evolution of anatomy. And, in many cases, one of these genetic “suspects” turns out to be the culprit. The modified beaks of Darwin’s finches, the elaborate mimicry systems of butterflies, the expansion of the human brain, and numerous other examples are the product of mutations that affect the way regulatory genes operate in embryos.
Had Driesch known about Shh and the results of modern developmental genetics, he might have stuck with embryology. Instead, Driesch dropped his promising career as a developmental biologist because he couldn’t imagine a material basis for the astounding ability of embryos to adjust and compensate for disruptions. We now understand some of the processes that endow embryos with their remarkable resilience. Interactions between genes, each with limited scope, can adjust development when things go awry. Not only do these adjustments resolve the paradox that Paul Nelson proposes—they reveal a fundamental mechanism through which morphology evolves.
So What Is BioLogos?
Well it all began with a scientist and a book. Francis Collins, the physician and geneticist who led the Human Genome Project, wrote the book, The Language of God. In it he describes his own journey from atheism to Christian faith, and the harmony between Christianity and science.
Today, BioLogos continues to carry out the vision of Collins, showing that you don’t have to choose between modern science and biblical faith.
Join the conversation on the BioLogos Forum!
At BioLogos, “gracious dialogue” means demonstrating the grace of Christ as we dialogue together about the tough issues of science and faith.