To understand what genes “do,” we have to consider what happens during development. The first and most striking evidence that the local environment matters for the outcome of development was provided by the experiments of embryologists Wilhelm Roux and Hans Driesch in the late nineteenth and early twentieth centuries. Roux had hypothesized that during the cell divisions of the embryo, hereditary particles were unevenly distributed in its cells, thus driving their differentiation. This view entailed that even the first blastomeres (the cells emerging from the first few divisions of the zygote – that is, the fertilized ovum) would each have different hereditary material and that the embryo would thus become a kind of mosaic. Roux decided to test this hypothesis. He assumed that if it were true, destroying a blastomere in the two-cell or the four-cell stage would produce a partially deformed embryo. If it were not true, then the destruction of a blastomere would have no effect. With a hot sterilized needle, Roux punctured one of the blastomeres in a two-cell frog embryo that was thus killed. The other blastomere was left to develop. The outcome was a half-developed embryo; the part occupied by the punctured blastomere was highly disorganized and undifferentiated, whereas those cells resulting from the other blastomere were well-developed and partially differentiated. This result stood as confirmation for Roux’s hypothesis.

During the 1970s, more puzzling observations were made. The first was that the genome of animals contained large amounts of DNA with unique sequences that should correspond to a larger number of genes than anticipated. It was also observed that the RNA molecules in the nuclei of cells were much longer than those found outside the nucleus, in the cytoplasm. These observations started making sense in 1977, when sequences of mRNA were compared to the corresponding DNA sequences. It was shown that certain sequences that existed in the DNA did not exist in the mRNA, and that therefore they must have been somehow removed. It was thus concluded that the genes encoding various proteins in eukaryotes included both coding sequences and ones that were not included in the mRNA that would reach the ribosomes for translation. These “removed” sequences were called introns, to contrast them with the ones that were expressed in translation, which were called exons. The procedure that removed the intron sequences from the initial mRNA and that left only the exon sequences in the mature mRNA was named “RNA splicing.”

One important, and for some the most surprising, conclusion of genome-wide association studies (GWAS) has been that in most cases numerous single nucleotide polymorphism (SNPs) in several genes were found to be associated with the development of a characteristic or the risk of developing a disease. As already mentioned, the main conclusion has been that the relationship between genes and characteristics or diseases is usually a many-to-many one, as many genes may be implicated in the same condition, and the same gene may be implicated in several different conditions. In fact, the same allele may be protective for one disease but increase the risk for another. For example, a variation in the PTPN22 (protein tyrosine phosphatase, nonreceptor type 22) gene on chromosome 1 seems to protect against Crohn’s disease but to predispose to autoimmune diseases. In other cases, certain variants are associated with more than one disease, such as the JAZF1 (JAZF1 zinc finger 1) gene on chromosome 7 that is implicated in prostate cancer and in type 2 diabetes. Therefore, we should forget the simple scheme of gene 1 → condition 1/gene 2 → condition 2, and adopt a richer – and certainly more complicated – representation of the relationship between genes and disease. Additional GWAS on more variants in larger populations might provide a better picture in the future. But insofar as we do not understand all biological processes in detail, all we are left with are probabilistic associations between genes and characteristics (or diseases). The “associated gene” may be informative, but its explanatory potential and clinical value are limited – at least for now.

This chapter is about the public image of genes. But what exactly do we mean by “public”? Here, I use the word as a noun or an adjective vaguely, in order to refer to all ordinary people who are not experts in genetics. I thus contrast them with scientists who are experts in genetics – that is, who have mastered genetics-related knowledge and skills, who practice these as their main occupation, and who have valid genetics-related credentials, confirmed experience, and affirmation by their peers. I must note that both “experts” and “the public” are complex categories that depend on the context and that change over time. There is no single group of nonexperts that we can define as “the” public, as people around the world differ in their perceptions of science, depending on their cultural contexts. We had therefore better refer to “publics.” The differences among experts nowadays might be less significant than those among nonexperts, given today’s global scientific communities, but they do exist. Finally, both the categories of experts and publics have changed across time, depending, on the one hand, on the level of experts’ knowledge and understanding of the natural world, and, on the other hand, on publics’ attitudes toward that knowledge and understanding.

If you were taught Mendelian genetics at school (see Figures 2.1 and 2.2) you should be aware that it is an oversimplified model that does not work for most cases of inherited characteristics. Human eye color is a textbook example of a monogenic characteristic. It refers to the color of the iris – the colored circle in the middle of the eye. The iris comprises two tissue layers, an inner one called the iris pigment epithelium and an outer one called the anterior iridial stroma. It is the density and cellular composition of the latter that mostly affects the color of the iris. The melanocyte cells of the anterior iridial stroma store melanin in organelles called melanosomes. White light entering the iris can absorb or reflect a spectrum of wavelengths, giving rise to the three common iris colors (blue, green–hazel, and brown) and their variations. Blue eyes contain minimal pigment levels and melanosome numbers; green–hazel eyes have moderate pigment levels and melanosome numbers; and brown eyes are the result of high melanin levels and melanosome numbers. Textbook accounts often explain that a dominant allele B is responsible for brown color, whereas a recessive allele b is responsible for blue color (Figure 4.1). According to such accounts, parents with brown eyes can have children with blue eyes, but it is not possible for parents with blue eyes to have children with brown eyes. This pattern of inheritance was first described at the beginning of the twentieth century and it is still taught in schools, although it became almost immediately evident that there were exceptions, such as that two parents with blue eyes could have offspring with brown or dark hazel eyes.

Perhaps you were taught at school that genetics began with Gregor Mendel. Because of his experiments with peas, Mendel is considered to be a pioneer of genetics and the person who discovered the laws of heredity. According to the model of “Mendelian inheritance,” things are rather simple and straightforward with inherited characteristics. Some alleles are dominant – that is, they impose their effects on other alleles that are recessive. An individual who carries two recessive alleles exhibits the respective “recessive” characteristic, whereas a single dominant allele is sufficient for the “dominant” version of the characteristic to appear. In this sense, particular genes determine particular characteristics (e.g., seed color in peas), and particular alleles of those genes determine particular versions of the respective characteristics. Mendel, the story goes, discovered that characteristics are controlled by hereditary factors, the inheritance of which follows two laws: the law of segregation and the law of independent assortment.

Rapid evolution can be observed happening in nature when selection is unusually strong. We are all familiar, these days, with the evolution of antibiotic resistance in bacteria and the evolution of pesticide resistance in insects. Less familiar, but also very rapid, is the evolution of resistance to heavy metals in populations of plants that have adapted to growing on the spoil-heaps surrounding zinc and lead mines. These cases of unusually strong selection and consequently rapid evolution are all associated with human modification of the environment. The classic case study of evolution happening – industrial melanism in moths – also fits into this category.

Evo-devo has come a long way since its origins a mere four decades ago. Many exciting things have been discovered, and there will be many more discoveries to come in the years ahead. I have tried, in this book, to give you a flavour of this new branch of science. Here, I summarize what I think are its most important conclusions so far and the most important challenges that lie ahead.

In the previous chapter we looked at several different kinds of developmental bias. One of our conclusions was that there are both specific biases, such as the numbers of centipede trunk segments and mammalian neck vertebrae, and general biases, such as the tendency for variant developmental trajectories – and in particular viable ones – to be clustered close to the ancestral trajectory. For example, in the case of snails we noted that the forms of developmental repatterning that were generally available for natural selection to act on were slight quantitative modifications of the pattern of development of the snail that was the ancestor of the clade concerned – an example being developmental trajectories leading to differences in adult shell size. Acknowledging this form of bias entails accepting that evolution of body form does not usually take place via radical-effect macromutations. This is interesting because we saw in Chapter 2 that from the late nineteenth century to the mid-twentieth century, prominent biologists who had a specific interest in the evolution of development, such as William Bateson, D’Arcy Thompson, and Richard Goldschmidt, took a macromutational approach.

Although today we call the scientific study of the relationship between evolution and development ‘evo-devo’, neither that term, nor its longer counterpart ‘evolutionary developmental biology’, existed before about 1980. Yet the study of the relationship between the two great creative processes of the living world has a much longer history – effectively starting in the nineteenth century, the first century in which there was a well-articulated theory of evolution (first Lamarck’s, then Darwin’s). We generally refer to evo-devo’s nineteenth-century antecedent as ‘comparative embryology’. Although in the subsequent period from about 1900 to 1980 there were further studies of the relationship between evolution and development, there is no collective term for this endeavour, because mainstream developmental biology and evolutionary biology were largely separate undertakings during that stretch of time. The few biologists who tried to deal with the two together over this 80-year period might be described as mavericks. Each of them produced interesting bodies of work, but these did not really link up to form a scientific discipline.

Body-plan features that have been discussed so far include symmetry, segmentation, skeletons, and limbs. When these are encountered in different phyla, are they homologous or convergent? There are examples of both of these, plus examples where the answer is not yet clear. Bilateral symmetry of the overall body plan seems to have originated just once. So the fact that vertebrates and arthropods are both bilaterally symmetrical is due to their having inherited that body layout from their last common ancestor; in other words, their bilaterality is homologous. However, although vertebrates and arthropods both have skeletons (whereas animals belonging to many other phyla do not) these represent convergent rather than homologous skeletons – this is clear from the fact that one is ‘endo’, the other ‘exo’. Turning to segments and limbs, the fact that both vertebrates and arthropods have these component parts is hard to interpret with certainty one way or the other. The reason for this is our lack of knowledge of that ancient animal that we call the urbilaterian, or ‘Urby’ for short. Direct evidence of this creature will probably never be forthcoming, since it was almost certainly small and soft-bodied, and has left us with no fossils from which to infer its living form. Instead, we can only make rather indirect inferences based on the point in the animal evolutionary tree at which we think bilaterality arose. However, indirect inference is better than nothing, so here goes.

Here, I list ten important issues where I think that there is a significant risk of misunderstandings among those who are new to the field. After each potential misunderstanding, there is a statement of the correct situation, as I perceive it. Many of these issues are related to the rationale underlying the emergence of evo-devo as a (relatively) new discipline.

Our starting point for discussion of evolutionary pattern is the word ‘clade’. This was introduced by the British biologist Julian Huxley (grandson of Darwin’s bulldog T. H. Huxley) in the 1940s. It means a taxonomic group of a particular kind: one that includes all the descendants of a particular ancestral species, and no others. This kind of group can also be called monophyletic. When the German taxonomist Willi Hennig founded the new approach to taxonomy that we now call cladistics, in the 1950s and 1960s, the idea of a clade was central. For those not familiar with cladistics, Hennig’s main concern was that the evolutionary trees that were used through much of the literature of evolutionary biology confounded two things: closeness of ancestry and similarity in body form.

The two great creative processes of biology are evolution and development. You and I, as adult human beings, are products of both. Evolution took about four billion years to make the first human from a unicellular organism that emerged from the primordial soup. Development, in the form of embryogenesis together with its post-embryonic counterpart, takes less than 20 years to produce an adult human from a different unicellular organism – a fertilized egg or zygote. By this measure, development operates more than 200 million times faster than evolution. However, despite their very different timescales, the two great creative processes of biology are intrinsically interwoven. Evo-devo is the scientific study of this interweaving. Its full name is evolutionary developmental biology, but because this is an unwieldy phrase it is almost universally referred to by its nickname.

It is constructive to approach this issue from a historical perspective. Some aspects of animal relatedness have been known for a long time – centuries – while some have only been established in the last few decades. And others remain to be worked out or confirmed. A useful starting point for this historical approach is the 1817 four-volume work Le Règne Animal (The Animal Kingdom) by the French comparative anatomist Georges Cuvier, who divided the kingdom into four embranchements (branches): vertebrates, molluscs, articulates (outwardly segmented animals), and radiates (radially symmetrical animals). We should note here that Cuvier was an anti-evolutionist; he was opposed to the evolutionary theories of his fellow Frenchmen Jean-Baptiste Lamarck and Étienne Geoffroy Saint-Hilaire, and he did not live to see the publication of Darwin’s Origin of Species. However, many non-evolutionists prior to Darwin (from Aristotle onwards) made good attempts at the classification of animals, even though the fruits of their labours would not be given an evolutionary interpretation until later. Here, I will discuss Cuvier’s suggested groups as being evolutionary ones, even though that is not how he saw them.

Although I received my doctoral training within the neo-Darwinian tradition, in a university department (at Nottingham) that was largely devoted to population genetics, there is a view of evolution adopted by some neo-Darwinians that I have always rebelled against. This is the view that evolutionary processes can be understood in terms of only two levels of biological organization – the gene and the population. At its worst, this view is associated with actually defining evolution in those terms alone. For example, in their 1971 book A Primer of Population Biology, the American biologists Edward O. Wilson and William H. Bossert defined evolution as ‘a change in the gene frequency of a population’. Evo-devo can be seen as a rebellion against this overly reductionist approach.

Taking an evo-devo approach to the living world changes our views of both of biology’s great creative processes – evolution and development. However, this is not a symmetrical change. Evolutionary biology is a more theory-driven discipline than developmental biology. Why this is so is itself an interesting question, but perhaps more one for philosophers and historians of biology than for biologists themselves. Anyhow, this difference between the disciplines means that our altered view of development is easier to adopt than our altered view of evolution – in the sense that it causes less tension in a pre-existing theoretical framework. Because of this fact, I will start with the simpler case (development, this section) but will spend longer on the more complex one (evolution, the other sections).