Bit by bit

David Lindley

  • The Triumph of the Embryo by Lewis Wolpert
    Oxford, 211 pp, £14.95, September 1991, ISBN 0 19 854243 7

In the old days, when organic matter was supposed to be infused with some vital spirit that distinguished it from the cold clay of the material world, and the variety of human types and the possession of free will were in-controvertibly attributed to the powers and generosity of God, it was not so hard to understand how an embryo grew into a whole human being. It was merely another of His miracles that a tiny organic seed, infused with the requisite life-spirit, could contain all that was necessary for the creation of a new person. And if the mechanics of the process seemed complex beyond human comprehension, that only proved the necessity of divine intervention.

But even from Aristotle’s time there was a semi-scientific debate as to how God accomplished His miracle. Was the embryo a tiny person, a homunculus, that straightforwardly grew into a full-sized human being, or was it rather, as Aristotle guessed, an unformed thing from which a new person developed bit by bit, one organ appearing after another? In the last century Victorian scientists tended to picture the sperm as a homunculus, with large head and tiny body and limbs tucked up behind, and, true to the times, thought that the male contribution contained all the important stuff: the woman’s role was merely to provide a nourishing egg and a cosy place for the embryo to grow.

Modern science refuses to allow such mysterious but lazy supernatural solutions into any explanation of how the embryo grows into a human being, and because we now understand that the embryo is a single cell – more complex than a liver cell or a brain cell or a muscle cell but of the same general form – we would like to understand embryonic development in terms of the fundamental processes of cell growth and division which are behind all biological change. The egg cell, fertilised by the sperm, starts to grow and divide, grow and divide, in something like the way a single bacterium, put in a suitably nourishing medium, will proliferate. But there is an obvious difference: a bacterium generates nothing except identical bacteria, replicas of itself, whereas an egg cell manages to grow into liver and brain and muscle cells, and does so moreover in an organised way, so that the brain and liver and muscles all end up in their appropriate places, properly connected. Aristotle was broadly right: the apparently formless egg cell generates organs and limbs in some sort of sequential manner. The modern biologist would like to understand what secret internal instructions tell the embryo to do one thing after another in the correct sequence, but that is a task of daunting complexity. As soon as one begins to wonder how a single, round blob of an egg cell can generate all that it is required to generate, the development of an embryo begins to look as mysterious and miraculous as it must have done in Aristotle’s day.

As Professor Wolpert hopes to persuade us, however, the problem is beginning to yield, one step at a time, to observation and experiment. It is hardly true to say that we now know how an embryo develops to the same extent that we understand how, for example, a virus invades the body’s healthy cells. The attack of a virus is a single purposeful stroke; the growth of an embryo is an advance on many fronts, using many devices. But something coherent can be glimpsed from the pieces that biologists have assembled.

At the very outset, multiplication of the egg cell does not seem all that dissimilar from the replication of a bacterium. The single egg divides into two cells, then the two into four, and so on. For a while, these cells are sufficiently alike for scientists, for example, to switch around the cells in an early mouse embryo at random, yet still find that a perfectly normal mouse develops. Two collections of cells from different embryos can be merged and a normal mouse – not a giant mouse or Siamese twin mice – will develop. In humans, the occasional division of an embryo into two pieces, which then grow into separate but identical twins, tends to happen when the fertilised egg has already grown and divided into something containing hundreds of cells. Each of these half-embryos is, in fact, a whole embryo, from which an entire person is able to grow.

At this stage, it seems, one cell in the embryo is as good as another; each contains a full prescription for the whole creature. But this versatility does not persist for very long. As soon as the embryo begins to change from an undifferentiated grouping of cells into a thing with shape and structure, cells begin to serve specific purposes, and can no longer be forced into other roles. A fundamental early event in the development of any mammalian embryo is the formation of the spinal column, which will grow to house the nerve cells of the spinal cord. The spinal structure first appears as a rift running half-way around the more or less spherical embryo; the rift deepens, and outer cells close in at the top to form, in a pleasingly tidy way, an enclosed tube. This column then develops ridges which will end up as the vertebrae. At this stage, all cells are most certainly not equal. If a section of the incipient spinal structure is snipped out and put back in upside down, it will develop as an up-side-down piece of spine. Each cell has evidently been told what it is meant to be, and will grow into that form regardless of external influences.

The same sort of process is repeated, on successively smaller scales, in the formation of arms and legs, then fingers and toes. Limb buds, when they first appear, contain cells that know only, so to speak, that they are to grow into digits. Experiments with chicken embryos involving the rearranging of cells in or near the limb bud show that the same cells are capable of turning into different parts of the final chicken wing depending on their position in the limb bud. But, as with the spine, once development has proceeded to the point where detailed structures – individual vertebrae, or different fingers – are becoming apparent, the purpose of the cells has become fixed, and cannot be made to revert to some other function.

Professor Wolpert’s aim is to impress upon the reader the idea that despite the enormous variety of changes that the growing embryo goes through, general principles can be discerned. Nature has been economical, and having found ways to make certain structures does not hit upon new methods for new structures but adapts existing mechanisms to new purposes. In order to make a spinal column, the embryo must have at the outset some sort of ‘polarity’ (a top and a bottom), so that when the external layer of cells begins to fold up, different cells can know to what part of the sequence they belong. Likewise, when limb buds start to sprout fingers and toes, they are able to do so correctly because cells know their place.

There are basically two ways that cells can know what they are supposed to be doing. They must either know where they are in relation to other cells, or they must remember where they came from. In the formation of a limb, both methods are used. There appears to be a chemical gradient across the tip of the limb bud, so that cells sensing a high concentration of chemical messenger turn into a thumb and those sensing a low concentration turn into a little finger. But into the outgrowth of the arm, from shoulder to elbow to wrist to fingers, cells seem to rely on timing. The limb grows because cells multiply behind the leading edge, which thus moves outwards: the first cells to be left behind form the shoulder, the next lot turn into the elbow, and so on.

For any of this to work, cells have to be able to tell each other things. This is a complicated subject, just beginning to yield to understanding, but the sixpenny version is fairly simple. Every cell in the body contains the same genetic information, in the form of a DNA molecule coiled up in its nucleus. This fact led, some years ago, to the story that a particularly egotistical millionaire had had himself ‘cloned’, or duplicated, by using the genetic information in one of his cells to make a completely new but identical millionaire. The problem with this is that although every cell contains the same genetic material, most of it is inactive. The DNA molecule can be thought of as a sort of instruction set, which contains the information to make every protein needed to build the body. But to actually make any individual protein, the DNA must receive a chemical signal activating the appropriate stretch of the long, coiled molecule. The reason we can’t clone people, millionaires or otherwise, is that we don’t know how to switch on the DNA molecule to make all the right proteins in the right order, the way the embryo does.

But we can at least see in principle how chemical signalling is able to make a human being out of an egg cell by switching on and off the appropriate instructions at the appropriate time. The picture to think of is not that, within the egg cell, there are already chemical arrangements of sufficient complexity to say: ‘make a liver cell here; make a brain cell here; make a muscle cell here’. This would be the modern equivalent of the homunculus theory. What has been realised instead is that simple instructions carried out successively can generate complex structures. Scientists today are in the position of trying to follow a chess game, but with only a vague understanding of the rules. Many chess games start out identically, with the same four or five opening moves: this is the first undifferentiated multiplication of identical egg cells. But on the sixth move, different games start to move along different paths. The difference at this stage may be tiny – a bishop goes one square further – but the game may by that action be tilted towards a very different finale. The position of the pieces on the board after fifty moves is, in some sense, a consequence of what happened on that sixth move, but it would be hard to explain how exactly that variation in the bishop’s move back at move number six became translated into the queen being on the far side of the board in one game and being lost altogether in the other.

Such an explanation, however, or at least a sketch of it, is what Professor Wolpert is attempting here. He argues persuasively that bits and pieces of embryonic development (the generation of the spine, the formation of limbs and digits) are understandable, if not exactly understood, in terms of variations upon a surprisingly small repertoire of ‘moves’, and he explains that chemical signalling acting in concert with the generative powers of DNA is able to accomplish the required cell growth and transformations. But, try as he might, he can never quite turn this assembly of processes into a grand design.

Beyond the obvious difficulty that the entire process is far from understood, there is also the fact that, all probability, no such grand design exists. The route that an embryo takes in order to change from a single cell into a whole creature is the result of billions of years of evolutionary trial and error, and for that reason it is not the most efficient route one could invent. At no point does Nature have the opportunity, having found a way of making arms develop with fingers on their ends, to review her method and rewrite the genetic in-structions in a more purposeful and compact way. From the evolutionary point of view, creatures first developed in the sea, and had gills, then went upon the land, breathed air, and acquired mouths and jaws. Embryonic development reflects this history: the growing mammalian embryo temporarily sprouts gill-like structures which then turn into jaws. This happenstance, noticed during the last century by the German biologist Ernst Haeckel, led to the principle that ‘ontogeny recapitulates phylogeny,’ or, prosaically, that a growing embryo retraces the evolution of the creatures from which mammals such as ourselves evolved. This principle is half-right. In fact, the embryo retraces in part the evolution of earlier embryos, not of adult creatures. A certain sequence of instructions, followed in the fish embryo, leads to the generation of gills; evolution builds upon that sequence, so that a later creature has jaws instead of gills; but because the creature with jaws is a refinement of the one with gills, the developing embryo at first follows the same successful series of moves, and only later branches out onto a different path.

This is the chess game again. A complex position, late in the game, may be reachable through a countless number of sequences of moves, of which some are direct and some perverse. Evolution, in working out a sequence of moves to control the growth of a human embryo, does not start from the finished product and work backwards, but zigzags forward to a goal that is not, at the outset, predictable. The criterion for a certain set of developmental moves to be retained is that it should be reliable, and once a reliable sequence has been incorporated into the genetic machinery there is no pressure to change it. Early on, a way was found of making fins, and because this method of producing a limb-like structure was robust, the rules to make chicken wings and mouse legs and human arms grew out of the fin-making procedure. There is therefore no deep reason why embryos sprout gill-like or fin-like structures that turn into jaws or legs or arms: that just happens to be the way things were done. The economy of Nature’s method lies not in finding the most direct way for an embryo to develop, but in making conservative use of tried and tested methods to generate complex structures from simpler ones.

Professor Wolpert’s book reflects the state of science itself. The casual reader, rushing through it, will be faced with a dense blizzard of facts in which the outlines of grander schemes are occasionally glimpsed. Although the author begins by stating his faith in the existence of overarching ideas, his writing shows the scientist’s fascination with detail: the fun is in figuring what minute differences persuade one limb bud to turn into a leg and one into an arm, or in understanding what small initial variations turn one limb into a human arm and another, by varying the proportions, into a chimpanzee’s arm. Wolpert’s writing is demanding. He lists facts and ideas in short, staccato sentences which are individually easy to digest but which taken together ask the reader to pay attention; the alert reader, like the inquisitive scientist, will find these details begin to connect up, though never completely. Professor Wolpert does not offer any easy solutions, because there are none, but for this reason he shows why there is plenty of mystery left in this old subject, and how new knowledge has not diminished its fascination.