Should genes be seen as controllers or nudgers of biological development?
For this piece, I was going to revisit a subject I have long been interested in: the evolutionary origins of the human face. I published a book on it eight years ago. Its central question was: how did the human face become so different from the basic mammalian face and even that of our nearest animal cousins, the chimpanezees, and why are those differences important?
I decided, however, that it might be best to preface that discussion with a more general one on what genes do and how they do it. This is a basic issue in genetics but there has been much confusion about it, particularly amongst non-biologists. Here, I will address it in a short piece.
As a subfield of biology, genetics has a distinctive history. Most disciplines in biology have rather hazy beginnings, involving a few relevant discoveries whose connections are often not clear at first. Then, a few people realize that a new scientific field is coming into existence and they start shaping it. Usually, its key problems come into focus rather gradually. In contrast, genetics can be assigned a definite year of birth and it had two key questions at its start, which were fairly apparent.
The first question was “how does biological heredity work in plants and animals?” and the second was “how do genes achieve their effects?” The first was basically solved within a few years, though much work has elaborated and refined the answer. The second had to wait for its first provisional solution for about 60 years while a further 60 years later we are still struggling with it.
The birth year of modern genetics was 1900, when three researchers published their discovery of the pioneering work on heredity by the Moravian friar Gregor Mendel, an amateur scientist. He published his work in 1866 but it lay largely ignored for decades. Then, in 1900, the three scientists independently discovered his paper, while doing their own work on heredity. All three confirmed that he had been on the right track. It was clear that there was now beginning to be an actual science of heredity. The world had to wait a few years, however, before this new field acquired a name, “genetics”.
Here, we will begin with a brief look at the persisting old question “how do genes work?” but then focus on its companion question, “How do changes in genes – termed ‘mutations’ or ‘genetic variants’ – alter their operation?”. Simplifying, we can say that genes, which are sections of long linear molecules of deoxyribionucleic acid (DNA) work by specifying long stretches of protein molecules termed polypeptide chains 1. Each polypeptide chain is made in one or more cell types or tissues (usually more than one, sometimes hundreds depending upon what it does and how widely it is needed). It also has to be made in a certain appropriate quantity, within a certain range, in each of the cell types or tissues in which it is made. The typical form of the gene is termed the “wild-type allele” and the normal amounts in which it is made in the different sites of the body in which it is used are said to be its typical “gene activity”. What genetic changes or variants of that gene do is to change its level of activity.
Between the 1930s and the 1960s, it came to be appreciated that broadly there were three kinds of consequences that a genetic change (a mutation) could produce. First, it could reduce the amount of that gene activity, either partially or completely. A partial loss is said to produce a “leaky” mutation while a mutation that eliminates a gene’s activity is said to be a “null mutation”. The biological effects of partial vs total loss are often very different but both can be lumped together as “loss-of-function mutations”.
On the other hand, certain mutations can increase the gene’s activity or alter it to a slightly different one, without necessarily changing how it carries out its original function. (For instance, think of a gene that specifies a polypeptide chain that contributes to a particular enzymatic activity but which now has a new one in addition.) Such enhanced activities are said to result from “gain-of-function” mutations. As with loss-of-function mutations, the quantitative effects are important. There can be either dramatic or modest increases in activity but the biological effects can vary greatly, depending upon what the gene’s product (its polypeptide chain) does and how many sites in the body there are in which it is active.
In the 1950s, it was assumed that all mutations created one or the other of these effects, a loss-of-function or a gain-of-function. In the mid-1960s, however, it was discovered that many mutations actually have no measurable biological effect. Many single-base pair changes in the DNA sequence of a gene have no detectable effect on gene activity. Such alterations are said to be “neutral mutations”. Indeed, we believe today that most single base-pair mutations are neutral mutations. This is because much of the DNA in our genomes is of suspected of having no function and even where it does have one, that function can be insensitive to many individual base pair changes 2.
To sum up: mutations, though incredibly diverse in their individual effects can be broadly classified in character as belonging to one of three categories: loss-of-function mutations, gain-of-function mutations, and neutral mutations.
That simple categorization however hides a wealth of complexity in the biological effects. In particular, many genes function to inhibit, to one degree or another and for differing lengths of time, other gene activities. Hence, a loss-of-function mutation in such an inhibitory gene would result in some increase in activity (hence gain-of-function) in other genes. Conversely, a gain-of-function mutation in a gene that inhibits other gene activities might increase that inhibitory role and further damp down the activities of the responding genes.
Finally, a neutral mutation might later come to have a positive or a negative effect, depending on whether other things in the environment, internal or external, have changed. The key point here that genes never act in isolation or with full independence. Every gene is part of some larger set of genes that its activities influence or are influenced by. Such networks of gene interaction are termed “genetic regulatory networks” or GRNs. They are multi-level sequential hierarchies of genes that turn on or turn off other genes 3. Hence, when a gene’s activity changes through some change in its DNA sequence, it will often change something in the activities of its partner genes within the GRN(s) it is part of. (No simple diagram of a GRN can do it justice while full diagrams overwhelm with their complexity, while failing to capture their dynamism.)
The idea that individual genes act via interactions with other genes first developed strongly in the 1990s, as the evidence for their existence accumulated. In doing so, it put paid to the notion, frequently expressed in the 1960s and 1970s that individual genes “controlled” or “determined” complex traits. It is true that there are certain powerful regulatory genes that directly activate or inhibit the activities of other genes but it is assigning them too much individual influence to say that they have this omnipotent power. One can still hear echoes of this idea in comments about the genetic material determining the properties of the organism or from people who want to clone their pets. (For the other factors, see Time to move on from Nature vs. Nurture) but few scientists who work in this area use this terminology of genes “determining” traits.
With these fairly basic thoughts about how mutations affect genes’ activities, and the importance of GRNs, we can begin to look at how the human face develops in the embryo and foetus and how it came to evolve from the face of some ape-like ancestor. The reality is that the human face is different in several key respects from the generic, typical mammalian face and less so but still significantly from that of our nearest living primate relatives – and we will soon examine those differences.
Yet, in some sense, we can see our face as produced by a process that merely tweaks the growth processes that give rise to these other mammalian faces. These tweaks can be seen as nudges that either accelerate or slow down or prolong growth of different parts of the face and head. The nature and strength of these nudges are provided by mutations that either cause small, partial loss-of-function or enhanced or only moderate gain-of-function of particular gene activities. Those small changes are often then amplified as they propagate through particular GRNs., with the result that small changes in the activity of particular genes can end up producing very visible, obvious changes in the features of the face. In this way, we can understand both the basic similarities in mammalian faces and their interesting differences.
We will return to discuss the evolution of the human face in more detail in the next BioBuzz article.
For a clear account of the basic facts of genes and gene actions, there is still no better source than James Watson’s The Molecular Biology of the Gene (Cold Spring Harbor Laboratory Press), first published in 1970. There have been seven subsequent editions, all larger than the original book and most with more participating authors but the first two editions are still the best for getting the big picture.
The human genome consists of about 3 billion base pairs of DNA sequence, split up amongst and located on 23 chromosomes. Of this large amount of DNA, roughly 2% consists of protein-coding sequences, for about 20,000 genes in total. A further yet undetermined fraction of the total DNA consists of DNA sequences, perhaps 10-20%, that regulate the turning on and off of genes. The rest has no known function and may consist of DNA copies of retroviruses that have lost their viral functions but have been maintained in the genome. A long-standing challenge in molecular biology remains sorting out what all this seemingly useless DNA is doing, if anything. Even among the coding sequences, there are many base pairs that could be replaced by others with little or no effect on the encoded proteins. All changes in the base pairs in DNA that do not change the coding or regulation of genes would presumably be neutral mutations.
As far as I know, there is no current easy to read general review article on GRNs but there is a large scientific literature on them. A key figure in this area was the late Eric H. Davidson (1940-2018), a molecular developmental biologist at the California Institute of Technology. In two landmark papers, in 1969 and 1971, he and his colleague Roy Britten, initiated thinking about GRNs. He also wrote a number of books on the subject. I recommend “The Regulatory Genome: Gene regulatory networks in development and evolution” (2006). Elsevier Publishers: Amsterdam.