Figure and Ground
Translating the Genome

by Tom Sgouros, Jr.

October, 2000

Above the table where I write, there is a still life oil painting. The rendering is brusque, almost telegraphic, and the painting readily dissolves back and forth between an image -- a coffee pot, a bottle and some garlic -- and a meaningless pattern of brushstrokes. The uneasy balance between meaning and mess is delicious, and it's what I enjoy most about the painting, aside from the fact that my father painted it.

This essay is also a pattern, made up of 26 letters and punctuation. It, too, is potentially meaningless. Like the painting, the interpretation of the meaning described by the pattern of these letters is a fragile thing, liable to be impossible except under certain special conditions.

Meaning is granted to this arrangement of letters by the reader's command of the English in which it was written. The author's intentions are not irrelevant, but they are only a part of the story. I do have a meaning in mind when I arrange these letters, but that meaning is only conveyed to readers who share my language.

My language is not just English, but English informed by the time and place in which I live. It is not simply a set of dictionary definitions, but a set of definitions augmented by the formal and informal rules of usage in effect right now, where I sit. If I write of someone that his elevator doesn't go all the way to his top floor, I divide the speakers of English into a group that understands the meaning and a hopefully much smaller group that does not. My use of the word "hopefully" in the previous sentence further divides the world into a group that disapproves of this usage as a corruption of the adverb, and another that considers it an unremarkable part of everyday English.

But what is often overlooked is that the context in which the pattern of this essay is most accurately interpreted is tremendously, absurdly, unimaginably, more complex than the essay itself. As we all have private fancies with which we arm ourselves against the cruel world, I imagine I have thoughts about life that others may find interesting. But in the deepest embrace of this fancy, in the wildest ravings I've ever suffered, I've never once imagined I could alone build an intellectual edifice to rival the richness and complexity of the accumulated store of the American English language in the 21st Century. I may occasionally wander in fields of self-delusion, but I know the way home.

This essay can be translated into another language, allowing people who don't speak English to share its meaning. But some translations are more challenging than others. France, for example, shares with America a great deal of intellectual tradition. The French have elevators, for one thing. There are relatively few utterly untranslatable cultural references, perhaps excluding a few very current ones. Consequently, a French-English dictionary and some basic grammar is all you really need to do an adequate translation. Of course, the more sensitive you are to the nuances of each language, the better the translation, but the point is about the conveyance of meaning, not the elements of style.

In contrast, a translation into one of the many indigenous languages of the interior of New Guinea -- where elevators are scarce -- may be more challenging. To make a real translation into such a language, I must make explicit much of what I can leave implicit to those who share my culture.

The cultural common ground on which the reader and the author stand is often slighted in the same way it is easy to overlook the air we breathe. But, like the air, close examination reveals normally unseen whorls of complexity.

So far, so obvious. So what?

The point is that somehow, this common sense analysis leaves us when we consider the complexity of DNA. Like this essay, a molecule of DNA contains a pattern made up of a small number of "letters" arranged in slightly larger groups which could be considered "words" which are themselves arranged into groups fancy could call sentences or paragraphs. The whole sequence, across all the separate chromosomes, makes an organism's genome. The comparison to a written human language is nearly unavoidable, and few people -- or at least few writers -- avoid it.

But if it is indeed possible to consider the genetic code a language, it is a language so foreign that in order to translate any of it, we need a boatload of extra explanation. A native speaker of an indigenous New Guinea language would need help to understand some of the references in this essay. In the same way, to follow the story in our genes, we need an explanation of the unimaginably complicated sequence of events that make a growing child.

When biologists want to clone a piece of DNA, they put it into some container, and add some enzymes and nucleic acid bases. If it's done right, the result is a test tube full of DNA copies. But sitting there, isolated in solution, DNA is mute. It won't build anything at all unless you use it to make some RNA, and add some amino acids to make some proteins.

But it's not that simple; you have to get the proportions of these additional ingredients correct. Without enough of some amino acid, a protein that depends on it won't get built, no matter what the DNA says to do. The proportions of ingredients are information that must be added to the test tube to get the DNA to do its thing. To translate this essay into a culture where elevators are scarce you have to add information about elevators. To translate the DNA into a human-readable form, you have to add information about the context in which the DNA is to be read. You have to make explicit information that is implicit in the construction of the cell from which the DNA came.

In fact, there are important features of any organism that are simply not encoded in its DNA. For example, DNA does not encode the location of a fruit fly's head or tail. The location of the embryonic cells destined to become one end of a fly or another is determined by an asymmetry to the egg cell engineered by its mother. In flies and other invertebrates, there are "nurse cells" to do this, sitting next to the egg developing within the mother. The nurse cells inject some RNA into the egg cell, and the RNA makes proteins and the concentrations of those proteins determine how the DNA is read, and therefore how regions of the egg differentiate into various body parts. The construction of the nurse cells is, of course, determined by DNA, but these nurse cells will help create the sons and daughters of the animal that contains them, not the animal itself. Fruit fly eggs require 15 different nurse cells apiece to add information not already present in the egg's own DNA.

In vertebrates, there are similar mechanisms to do similar things, though as you'd expect, the processes seem more subtle. "Follicle cells" in humans surround the developing egg, adding protein layers and injecting information to be used by the developing embryo. Whatever the process, the result is the egg: a complicated little device, on whose structure an animal's development depends. The "structure" in this case is the variety of building blocks an egg contains -- RNA, proteins, amino acids, nutrients, and the remaining undiscovered what-have-you -- and the relative concentrations of these chemicals in different microscopic regions of that egg. We can think of cell interiors as a soup of cytoplasm, but it's a soup where the noodles and vegetables are all arranged in some pattern and it matters which noodle is where.

And since it's the real world we're talking about, the complications don't end there. The part of the fertilized and already developing fruit fly egg cell that will become the underbelly of the fly is the part that happens to come into contact with the lining of the mother's egg ducts. An egg doesn't just get nutrition from its mother; it also gets important directions that control its development. There is no "switch" turned on at the moment of fertilization, after which development proceeds autonomously. Fertilization is simply one step among many in the interaction between genes and their environment that eventually produce a new organism. The unsettling part -- the aspect that mocks attempts to analyze separate steps of the process -- is that in the early stages of development, half of the genes controlling the environment are the same as the ones developing in the egg. The genes control development, but they do it both from within the egg, and from within the parent. The egg needs its genes to develop, but it also needs its mother.

In a laboratory, people can get DNA to make a few proteins on command in a test tube, because they now know a little of the information implicit in the cell. At present, however, nobody can get DNA in a test tube to build anything more complex than that, certainly not another organism. To accomplish that magic, you rely on a native speaker; you have to take a little bit of the DNA and put it back into the nucleus of the kind of cell from which it originally came. That is the only way now, and it is likely to be the only way for a long time to come. The word "clone" conjures images of petri dishes and test tubes, but Dolly the sheep, the first large mammal clone, was created by teasing out the intact nucleus of a cell in her original's body and inserting it into an egg cell whose nucleus had been similarly removed. Then the result was put into the uterus of a sheep who bore her. The devil is in the details, and the specific techniques for doing these things are not simple, but the principle is straightforward.

Science has learned how to tease out some of the meanings in the language of the genes. We can examine some of the proteins certain genes build. We can correlate some genetic changes with changes in an organism. But is this the same as decoding the language? Suppose you spoke no English, and could not read this essay. You show the essay to a friend who can read English, and ask her opinion of it. Now you start changing single words here and there and asking her to reread the whole thing and tell you about the changes. Some changes she might not notice, while others would be more obvious. Some words are important to the sense of the essay, but some aren't. (Like this one: "purple.")

When you change a word and it has an effect on your friend, you do know something, but maybe not much. For one thing, you have no way really to know whether the word you've removed was the important one, or perhaps it was the one you've added. You could change words in parentheses to anything you want, and not affect the overall meaning. What could you conclude from that? I could even write sentences that only make sense with nonsense words: "`Bfstplk' is unpronouncable." What will you learn when you experiment with substitutes for that word?

Faced with the task of translating an essay in a foreign language, finding the letters and the words is the easy part. The hard part is deciphering the meaning behind the words. And it's even harder when the only method at hand is to substitute other words and see what happens. But that's roughly the way scientists are proceeding to "decode" the meaning of our genes. You might find the protein produced by some stretch of DNA, but proteins don't come with labels for their use. They are big ugly complicated molecules, and the only way to figure out their purpose is somehow to catch them in action. Even then, you don't always know if the function you've observed is the only thing they're good for. People spend entire careers studying a small handful of proteins, and there are millions of them to figure out.

This is not to belittle the enormous work already done in reading the genetic code, nor the tremendous medical advances made possible by that work. But the fact remains that in the attempt to understand how organisms make themselves, "reading" the genome -- figuring out what parts code for what proteins -- is but a very small first step. When we know what protein is made by every single part of your DNA, we still won't have the slightest idea how to make another one of you.

We think of an organism as a thing of its own. A thing has boundaries to separate it from the rest of the world, and it has a beginning to its existence, and (generally) an end. We want a developing animal to start from a clean slate, read its own instructions and build itself, but that's just not the way it works. Eggs inherit information from their genes (from the nucleus and the mitochondria), but also directly from their mothers, in the egg's cytoplasmic soup, and in guidance to the process of development.1

The genes guide the embryo's development, and they also moderate the actions of the parents, so they do ultimately control most of the process, but they do it through this strange symbiosis between generations. Looking for a "beginning to life" -- trying to pinpoint the moment when a developing organism becomes an independent thing -- is an exercise in futility. The beginning of life was billions of years ago. For now, life just is. I am a continuous extension of my mother's life, as she was of her mother's and so on and on right back to the blue-green algae mats floating in shallow Pre-Cambrian seas.2

Spatial boundaries have a similar problem. We want a thing to exist, contained in its environment, as a brick might sit in an empty room. But the environment is not a simple container, and no living thing is as docile as a brick. Living organisms interact with their surroundings, often changing them in profound ways by the very act of living. A rabbit eats the grass in its fields; a beaver cuts down the trees in its woods. Some changes are minor, but some are dramatic: a virus might kill its host; Earth is habitable for us only because billions of years of plant photosynthesis have produced enough oxygen for us to breathe.

Life is not a Platonic Form, existing in some airy theoretical vacuum. Life is a process carried out in the constant dialogue between an organism and its habitat (which includes other organisms). Change one and you change the other.

A gene can't be read unless it's in the right sort of nucleus, in the right sort of egg, in the right sort of organism, which itself needs to be in the right sort of environment to thrive. Not only is there a continuity in time, as we trace our beginnings right back to the beginning of life, but there is a continuity in space, as we trace the chain of dependence out from the gene, like spreading ripples on a pond.

Genes did not evolve to satisfy the dictates of Platonic categories, so we shouldn't be surprised to find unexpected relationships between genes and their surroundings. Genes evolved by chance and natural selection. That is, they evolved to work, and nothing more. They work fabulously well, and have changed the planet by so doing, but the structure they have created is riddled with subtle and delicate interdependences, between different parts of the gene, between different cells, and between different organisms.

The nucleus incorporates the gene, and depends on it. Our environment incorporates us, and depends on us as well. Just as it makes little sense to attempt to search for the moment when life begins, it makes little sense to try to draw lines to separate ourselves from our environment; the threads of interdependence are far too tangled.

Unfortunately, the philosophical trend since Plato is to see ourselves apart from the rest of the world, as a separate (and exalted) thing. But the evidence for that view has always been thin, obtained largely by philosophical navel-gazing rather than accumulation of data. The past century's close scientific observation of how life works has shown this separation is nothing more than a vain illusion.

In a painting, one speaks of a relationship between the "figure" and the "ground," between the subject of the painting, and the space around it. The figure defines the ground, and the ground defines the figure. It makes no sense to talk of one without the context defined by the other, and a successful painting is one where the relationship between the two elements adds to the overall effect.

The study of genes has us on the beginning of an exciting road. But each new step in understanding how life carries on will reinforce this point: it is as impossible to separate any life from its context as it is to analyze a painting by only looking at the figure. It is pointless to talk of a developing embryo separate from the mother surrounding it, or to talk of that mother as separate from the water she drinks and the air she breathes.

In twenty years, we will have made some progress in disentangling the threads of interdependence. We will understand much better how life works. But, possibly aside from screening and preventing simple genetic dysfunctions, we still won't be able to engineer our children to have desirable traits, which are complicated combinations of genes and environment. But we will understand, more deeply than ever before, how important the world around us is to our own lives, and how delicate the balance that maintains us.

A generation from now, one hopes that the popular understanding of science will have shifted from wonderment at the advances in genetics to a real appreciation of the the vastly greater complexity -- and the precariously teetering structure -- of the context in which those genes thrive: nucleus, cell, organism, and environment.

  • Footnotes