Creation (24 page)

To get a sense of how this system has been breached by scientists in the last few years, we first need to understand how the machinery of the cell works. Protein construction is remarkably like a mechanical production line. The gene being translated is first transcribed from DNA into a free-floating copy in RNA, maybe a thousand bases long, untethered to the rest of the DNA of the host cell. This RNA message drifts to one of many ribosomes, the center where the protein will be constructed. In the cellular milieu around the ribosome float the amino acids that will be strung together according to the order set out in the messenger (or m-) RNA, each one specified by a three-base codon—the specific triplet of letters that encodes an amino acid. Several key molecular players shift these molecules around in the manner of a factory production line, and the process works as follows. The gene to be translated (in the form of mRNA) is fed into the middle of the ribosome, like feeding paper ribbon into a ticker-tape machine. The RNA bases are read codon by codon, and the corresponding amino acids are added one at a time. The protein is ejected like the ticker tape being spat out. This output is the basic protein, to be folded and transported to its site of use. Yet the agents in this process are effectively manufactory handlers. The ribosome itself is of fundamental importance: its shape and construction, largely out of RNA itself, is the forge into which all the ingredients trundle along this biological conveyor belt. There is also a complex of three molecules that very specifically deliver the amino acids to the ribosome. This combination is basic to all life, but not easy to explain. One part is the amino acid itself, which is collected by the second part, a folded piece of RNA that specifically collects and transfers it to the ribosome. For this reason, these are generically called transfer (or t-) RNA, and work by bearing the anticodon, that is, bases complementary to the codon itself—a
T
to pick up an
A,
a
C
to pick up a
G,
and so on (though there are only twenty amino acids, there are dozens if not hundreds of tRNAs). The final handler is a protein with the unwieldy name of aminoacyl tRNA synthetase, which I shall refer to as synthetase henceforth. This protein clamps the amino acid to its corresponding tRNA, and together the component parts of the protein are delivered for assembly. The code input goes in at one end, the ingredients are delivered from another as a package, and the protein rumbles out, one amino acid at a time.

This process occurs in all life-forms and its universality is one of the cornerstones of the evidence for a unique origin of life. The components are clearly ancient—no matter how distantly related two species are they share the basic workings of this assembly plant. To change it has required an example of evolutionary brute force, an intelligent design pioneered by Peter Schultz at Scripps in California, and by Jason Chin and his team at the LMB in Cambridge.

The existing code has plenty of redundancy built into it: there are sixty-four possible combinations of the four bases, and sixty-one of them are used to encode the twenty amino acids of life.
⁴
That leaves three remaining possible combinations, and in life-forms all of them bear the message
STOP,
that is, the instruction to terminate protein production. These punctuations are essential to indicate the end of a gene. Indeed, the significance of the stop codon is demonstrated in a handful of genetic diseases: Sandhoff disease, lethal in infancy, is caused by a gene bearing a premature stop codon, like a sentence cut incomprehensibly short by a rogue period.

One of them, the triplet
UAG,
is called the amber stop.
⁵
Jason Chin has appropriated this sequence as a codon for building amino acids otherwise unrecognizable to living cells. There is no tRNA-synthetase pairing that will bind to an amber stop, and so Chin's idea is to evolve one. Chin and his team start with this pairing from a species distantly related to the experimental cell. This ensures the pair won't start interacting with the cell's normal workings: the mechanics of protein production are universal, but that doesn't mean that the kit is exactly the same. In order to get the alien synthetase to singularly recognize something that none of the other tools of life can—the amber stop—they use the basic premise of evolution. If for some reason you wanted a fish that only ate something artificial such as jellybeans, you would start with a densely populated pool and use a jellybean as bait. If there was a fish in that pool that, unusually, had a natural taste for jellybeans, you would catch it with that bait. You could then breed a new pool from that fish and create a population that liked jellybeans. By repeating the process iteratively, you would end up with fish that only fed on jellybeans.

The fishing exercise is similar to the process of evolving a tRNA-synthetase pair that will recognize unnatural amino acids. By starting with a pool of versions of synthetase, which have randomly mutated parts in the region that locks onto the amino acid, you can select ones that bind to your new unnatural amino acid, and repeat the process. Eventually, just like the jellybean-eating-fish, you have created a tRNA-synthetase that will only read a stop and an unnatural amino acid.

That's one part of it. The tRNA is bred in that way to act as the anticodon for the amber stop. Adding a
UAG
into the sequence of a gene so it can be picked up by the tRNA is an awkward process, but essentially uses the tools of DNA manipulation not unlike those used for any gene tinkering—the creation of Freckles the spider-goat, for instance. You can synthesize a whole gene sequence, rewritten from scratch, complete with an amber stop at a crucial point. Or you can use a copying technique called mutagenic PCR that will copy imperfectly and introduce the new triplet in error.
UAG
normally only exists in a gene to indicate its end, so in order for this system to work, all the elements have to be changed: the code, the synthetase, and the tRNA, all to recognize an unnatural amino acid.

This is delicate biological engineering to the most fundamental living process—the “central dogma,” as Francis Crick described it. But it is not mere experimentation for the sake of it. Jason Chin and others are using it to discover how proteins interact with one another. Living cells are a network of protein interactions: some attach themselves to DNA, some to the molecules of metabolism, and frequently they connect with other proteins to perform their vital functions. The unnatural amino acid incorporated into Chin's experiments carries a side chain that does two things. First, when instructed to do so, it will form a locked bond with the protein it is interacting with, which means that we can fish the pair of them out and identify the unknown partner, with the engineered protein acting as bait. Second, the unnatural amino acid is light activated. To get it to lock onto the target, you merely shine ultraviolet light onto the cell, and the bond is sealed.

Chin is not the only person who can perform this act, but his team is the first to get it into multiple species, including animals. Until 2012, these unnatural actions were restricted to cells in culture, in dishes where the environment is controlled and neat. Chin has incorporated this process into fruit flies, one of the standard animals of experimental biology, and shown that no aspect of biology is off-limits when it comes to reengineering: alphabet, code, proteins—the whole system of life is now ripe for rewriting.

These rewriters show that while our system, the only one we know of, is powerful enough to form the basis of all life on Earth, it doesn't have to be that way. Evolution by natural selection was discovered and described a long time before we understood its mechanics. The code that enacts that process of error and trial is unique, but is quite capable of being adjusted, rewritten, or even invented afresh. It's difficult to conceive of a system of biological evolution (that doesn't rely on some form of supernatural creator) other than natural selection, but we only have one system of coded, reproducible information that has worked. XNA, the bases
Z
and
P,
and amino acids that are unnatural to the cell all show that there can be others. This is the first baby step toward the creation of new life-forms that use a Darwinian genetic code invented not by nature, but fully by us.

Digital DNA

With these examples we can see how the encroachment of synthetic biology on nature has given us the power to radically alter the genetic code of DNA, to invent new versions of genetic material, and even to expand the lexicon of genetics so that it can include unnatural proteins.

The language of DNA is not restricted to the creation of life. Our genomes are data storage devices, imperfect by natural design to encourage adaptation to the changing environment. Yet DNA is remarkably stable; it has transmitted information in the same form for billions of years, and the information held in DNA remains intact long after the death of the cell or organism that bears it. More and more frequently these days stories emerge from the new scientific world of ancient DNA. The genome of the wooly mammoth has been published—extracted and decoded from hairs sixty thousand years old, bought on eBay by the geneticists who sequenced it. Our likely evolutionary cousins the Neanderthals joined the genome club in 2010 when their complete DNA was read from 44,000-year-old bone specimens. Yet the current record stands with DNA extracted from frozen mud cores in Greenland, identified as being from the family of perennial plants saxifrages, with an age range of between a whopping 450,000 and 800,000 years old. True, being buried beneath six and a half thousand feet of ice is about as close to a laboratory freezer as nature can manage, but the fact that we can find long-dead tissue and still decode the messages written within it shows what a stable data storage device DNA can be.

This is one of the reasons that scientists in the 1990s began thinking about DNA as a means of storing not just the biological information required for a cell to function but also digital data. Although not the first to take advantage of DNA's data storage talents, Craig Venter's synthetic bacteria, aka Synthia, is the most famous living digital device to date. As mentioned in the introduction, within the computer- and machine-synthesized genome of the goat pathogen
Mycoplasma mycoides,
Venter hid messages in the DNA that would be cryptic even to the machinery of the cell. Computer coders often hide secret treasures or messages—Easter eggs—in their programs to give credit, or simply as fun puzzles for users to discover.

There were four parts to these DNA Easter eggs. The first was a code table. The cryptic messages were all conceived in English; a cipher had to be included so that the four letters of genetic code could spell out the twenty-six letters of the English alphabet plus punctuation. The twenty amino acids are sometimes referred to by a single letter,
A
to
W,
so a conceivable way of concealing an English language message in DNA would be to simply use the existing genetic code in the triplets that spell out amino acids, three bases for each Roman letter. Instead, Venter constructed a new cipher for the English alphabet, so this needed to be deciphered. The code was cracked within a few days of publication.

The second and third concealed messages comprised the names of the few dozen creators of the cell and an Internet address. The final Easter egg was a set of three apposite quotations. The first was “To live, to err, to fall, to triumph, to recreate life out of life” from the novel
A Portrait of the Artist as a Young Man
by James Joyce.
⁶
The second quotation came from J. Robert Oppenheimer, the so-called father of the atomic bomb: “See things not as they are, but as they might be.” The third was an accidental misquotation of that marvelous phrase from Richard Feynman referred to in chapter 7: “What I cannot build, I cannot understand.”
⁷

These blocks of DNA sequence were designed by hand and assembled in a computer. They had to be flanked by DNA tags that indicated they were not to be translated by the cell, because the code they are written in is entirely new and would have been nonsense. That sequence is not written in the triplet codons of the genetic code; it cannot and will not make a protein. Instead, Venter had designed a storage device with an encrypted code that was invented and new, simply using the alphabet of DNA as code, without reference to its history.

In 2012, Harvard Medical School's George Church drove the commoditization of DNA into a new era with the first publication of an entire book encrypted digitally in DNA (entitled
Regenesis
, it is appropriately enough about synthetic biology). At 53,000 words, eleven images, and one script of software code, it is three-quarters the size of the book you are holding, or around five megabits of information. The code is very simple binary: an
A
or
C
represents a 1, and a
T
or
G
represents a 0. Words were converted into a digital form and then the equivalent sequence of DNA was synthesized in a computer in ninety-six base lengths, each with information tags about the location of the segments and so on—what coders call metadata. Fifty-five thousand of these fragments make up the book, roughly one per word. In translating the whole book into DNA, there are only ten errors in five million bits of data. The whole thing is stored on what's called a DNA chip, which as a means of storing a book is not that different from a page and ink, just much smaller. These chips are small glass plates about the size of a book of matches, coated with chemicals that DNA will stick to. The DNA data fragments are literally sprayed by a tiny nozzle from an inkjet printer. Unlike the information stored in Craig Venter's synthetic bacteria, Church's method never goes near a life-form. The DNA is written in a computer, synthesized by a machine, printed with a printer, and decoded by a machine and software using the very same standard techniques employed by scientists recovering DNA from millennia-dead tissue.