Arrival of the Fittest: Solving Evolution's Greatest Puzzle (25 page)

So we are back to a familiar story. The regulatory circuit library has the same layout as the metabolic and the protein libraries. Circuits with the same gene expression phenotype are organized in vast and far-reaching genotype networks. And that has consequences for a crowd of readers wandering aimlessly along such a network, figuratively seeking something new to read, actually propelled only by the steady if directionless force of mutation that slowly changes circuits, one regulatory interaction at a time: Even though some steps garble a circuit’s expression code, many others preserve it and thus allow readers to move along the genotype network. While the readers wander, they reach ever-new neighborhoods that contain texts with ever-new meanings, ever-new expression phenotypes, one of which may seed the next big thing in life’s architectural contests. Once again, genotype networks and their diverse neighborhoods create innovability.
⁶⁰

These similarities among different libraries are mysterious. How could innovability in metabolism, in proteins, and in regulation circuits have the same source, a library full of chemical meaning with a common cataloging system? The answer is held by an invisible hand that guided the world long before life’s origin—self-organization, a peculiar kind of it. We will turn to it next.

I
n 1944, the Nobel Prize–winning theoretical physicist Erwin Schrödinger published a series of his lectures under the title
What Is Life?
The brief book was an attempt to reconcile physics with what was then known, in the days before Watson and Crick, about evolution. The book is brimming with ideas, and one of them has spilled into the mainstream of popular science culture: It is the idea that evolution increases order and decreases disorder—what Schrödinger called “negative entropy.” Four years later, the American electrical engineer Claude Shannon connected the thermodynamic concept of entropy to the problem of transferring information through a telegraph line. The concepts of evolution and information have been linked ever since, usually in a fairly primitive way. Disorder: bad. Order: good. Positive entropy: bad. Negative entropy—now also known as information—good.

In the years since Schrödinger’s book, we have become more sophisticated in thinking about entropy. Order and information remain central to evolution, but in recent years we have also learned, thanks to genotype networks, that perfect order is as hostile to innovation as total disorder. Nature doesn’t just
tolerate
disorder. It
needs
some disorder to discover new metabolisms, regulatory circuits, and macromolecules—in short, to innovate.

Let’s put Lego blocks to another metaphorical use, and consider the difference between a disorderly jumble of those familiar plastic tiles and an arrangement in which every tile has been presorted into a “proper” place, and where a child must assemble them in a specific sequence to build a pirate ship following a plan helpfully provided by Lego. The disordered jumble of Lego tiles has greater potential for innovation than the carefully organized one, and not just because it stimulates a child’s natural creativity to find new ways for building pirate ships. A deeper reason is that there are many more ways to build a pirate ship than those contained in Lego’s instruction book.

In biology this simple fact is manifest in the multiple solutions that nature found—courtesy of genotype networks—for problems like protecting organisms against freezing. And it is also deeply connected to a biological phenomenon little appreciated until the end of the twentieth century, but in fact so widespread that it deserves to be called a hallmark of life
: robustness,
the persistence of life’s features in the face of change.

The meaning of robustness is best illustrated with the difference between typographical mistakes in a traditional book and in a computer program. A book containing the letter sequence

N smll stp fr mn, n gnt lp fr mnknd

would raise eyebrows, but the meaning of this sentence remains understandable. However, a single misplaced letter or as little as a missing comma in a thousand pages of computer code can bring a million-dollar software package to a crash. Software bugs like this cause billions of dollars in economic loss every year. Human language is robust. Programming language, not so much.

The suspicion that life is robust arose at least as far back as the 1940s, when the biologist and philosopher C. H. Waddington studied flies with different genotypes and discovered that they had indistinguishable bodies, down to the minutest details of their wings’ venation and the numbers of bristles that cover their backsides. He called the phenomenon through which development can produce “one definite end-result regardless of minor variations in conditions”
canalization
—another word for robustness.
¹
And although his research hinted that the body plans of flies are robust to genetic change—there are many ways to build a fly’s body—research into robustness remained a backwater for another half century.

But almost overnight in the 1990s, robustness entered center stage with a flourish when molecular biologists were baffled by a discovery superficially unrelated to Waddington’s: Many genes apparently serve no purpose.
²

What’s baffling about such genes is why they would exist at all. Not only does a superfluous gene waste scarce resources, but mutations that incessantly rain down on DNA would eventually erode it, transforming it over time into something like an abandoned building that crumbles to dust over the years.
³

Many of these “purposeless genes” were unearthed after the genome of an organism that we already met in chapter 5 had become fully sequenced. It is a microbe, the brewer’s yeast
Saccharomyces cerevisiae
that helps us make beer and wine, and it is as useful for understanding cell biology as fruit flies are for embryology.
⁴
The yeast genome in hand, biologists realized that thousands of its genes had an unknown role in the microbe’s life. To reveal this role, they began to engineer “knockout mutations” into the genome, so called because they delete a single gene, an entire meaningful paragraph from a genomic text.
⁵
The logic of the experiment is essentially like analyzing the workings of a car by eliminating one part at a time: Remove the disk rotor, and if stepping on the brake pedal no longer slows the car, you have learned that the rotor is needed for braking. In the same way, if you knock out a particular yeast gene and find that its cells can no longer divide, the gene was involved in cell division. Knock out a fruit fly gene and if the mutant no longer forms wings, you know that the gene helps build wings.

The results of gene knockout experiments had trickled into the scientific literature gene by gene, until gene-knockout technology became powerful enough to delete thousands of genes. That’s what researchers at Stanford did in impressive experiments starting in the late 1990s, when they used the list of yeast genes revealed by the yeast genome and set out to delete every single yeast gene. They created some six thousand different yeast mutants, each of them missing a gene, placed these mutants into chemical environments where their unmutated ancestor could have thrived, and examined each mutant for specific defects, clues about the missing gene’s function.
⁶

What they found was completely unexpected. Thousands of these mutants do just as well as their ancestor and show no obvious defects. The genes that had been deleted to create these mutant genomes served no obvious purpose. Since then, scientists have blocked countless genes in many other organisms. And these genes tell the same story as a vowel-free English sentence: Like natural language, life is robust—in this case to gene deletions.
⁷

Discoveries like this do mostly one thing: They create new questions. One of them was
how?
What mechanism creates robustness?

For some genes the mechanism was straightforward. These genes were duplicates, stretches of DNA that occur more than once in a genome, like pages in a book that someone has photocopied twice by mistake. Gene duplications happen when an organism copies or repairs DNA, and are by no means rare: About half of the genes in our own genome have duplicates.
⁸
Since identical duplicates can do the same job, one of them can take over if you knock out the other.
⁹
Like the redundant power supplies that hospitals use to safeguard against power failure, like redundant computer memory to prevent data loss, like redundant circuitry in commercial aircraft to prevent crashes, some genes are only “useless” until they’re needed.
¹⁰

But many of the dispensable genes have no duplicate—they are single-copy genes—and for them the causes of robustness are not as simple.

We understand those causes best for genes that encode the enzymes of metabolism. A metabolism’s chemical reaction network resembles the dense road network of a city’s core, like that formed by the right-angled streets of midtown Manhattan. A driver who wants to get from 42nd Street and Second Avenue to 48th Street and Seventh Avenue has any number of choices for following the street grid six blocks north and seven blocks west. The major arteries have multiple lanes—think of them as redundant, because even if one is blocked, the driver can continue in another. But even a complete roadblock is not a problem, because the driver can use a different part of the grid, and a really intrepid driver might even cut through parking lots with entrances on two parallel streets. Such detours slow down but don’t halt the journey.

A knocked-out metabolic gene is a bit like a blocked road that halts the flow of molecules through a network of metabolic reactions. The detours around the roadblock are alternate metabolic pathways, sequences of chemical reactions that can absorb the backed-up molecular traffic, synthesize needed molecules in different ways, and ensure that life in metabolism city goes on.
¹¹
This isn’t just an abstract metaphor. Biotechnologists can create metabolic roadblocks by knocking out metabolic genes, and when they do, organisms like brewer’s yeast often survive by rerouting the flow of essential molecules. In metabolism, this kind of robustness is even more important than redundancy.
¹²

Robustness isn’t limited to metabolism and whole genomes. It is just as pervasive in individual proteins like
lysozyme
. This protein kills bacteria by destroying their wall of protective molecules. It appears not only in human saliva, tears, and even mother’s milk, but in a large number of other animals, and even in viruses that attack bacteria.
¹³
When scientists want to find out how a protein like this works, they do something akin to knocking out genes in a genome, but on a smaller scale—they change individual letters in the protein’s amino acid string and observe the effects of each change. When they engineered more than two thousand lysozyme variants, each one with a single altered amino acid, they found that some sixteen hundred variants—more than 80 percent—could still kill bacteria. Proteins like lysozyme, and there are many, are as robust as metabolisms. And the same holds for regulation circuits—we already heard about a circuit in the bacterium
Escherichia coli
that can be rewired in the laboratory without ill effects (chapter 5).

The most obvious benefit of such robustness is that it keeps organisms alive. Its importance goes back all the way to the first self-replicating RNA molecules and the fatal error catastrophe, in which small errors compound over time until replication becomes impossible (chapter 2). That was a true catch-22: RNA molecules have to self-replicate with few errors to acquire the ability to self-replicate with few errors. But only a bit of the robustness in today’s RNA could lower the bar for this problem to a manageable height: Because a few replication errors in a robust molecule do not erode its ability to self-replicate, robustness provides a stay of execution by error catastrophe, perhaps long enough to stumble upon better replicators.
¹⁴

But the importance of robustness goes far beyond that. It explains the mystery of genotype networks and of innovability.

To see why, we need to revisit nature’s libraries, where each metabolism (or protein, or regulation circuit) is represented by a single text, and where each of this text’s neighbors differs in a single letter, a single reaction, or a single enzyme and its gene. We know from gene deletion experiments that many of these neighbors, for example metabolisms where a single reaction has been eliminated through a gene knockout, suffer no ill effects. This means that even when the
genotype
has changed, there need not be any change in the
phenotype,
in the organism itself and its observable features. An organism like this is robust. The extent of its robustness is reflected in the number of its neighbors—variants a single small change away—whose phenotype remains unaffected by the change: The more neighbors with the same phenotype, the more robust the organism is.
¹⁵
Think of this phenomenon at its theoretical limits: If a metabolism, or a protein, or a regulatory circuit had no viable neighbors it would be maximally fragile. Change one of its parts, and death follows. At the other extreme, if every possible change were viable, if every neighboring metabolism had the same phenotype, the metabolism would be maximally robust: No single change could kill it.
¹⁶

These extremes do not exist in the real world. No real organism completely lacks robustness, and no organism is perfectly robust. But all organisms, their structures and activities, are
to some extent
robust, and it is precisely this robustness that allows populations to explore nature’s enormous libraries. The number of texts with any one meaning in these libraries is vast, but these texts fill a tiny fraction of the library, like a droplet of molecules in an ocean. In the complete absence of robustness, many texts might tell the same story, but none of their neighbors would. No explorer could browse one text and find a neighboring one with a single page—or word, or letter—changed but its meaning nonetheless intact. Genotypes with the same phenotype would be like stars in the sky—a billion twinkles isolated by light-years of empty space.

Luckily, the biological world is different. Starting at any one robust text, we can step to one of its many neighbors with the same meaning, and we can step to one of
its
robust neighbors, and so on, never changing this meaning, and thus exploring ever-new regions of nature’s libraries that harbor untold innovations. Robustness allows
some
disorder in genotypes, and permits nature to explore new configurations of its Lego blocks through the genotype networks it helps create.