Biomimicry

by Janine M Benyus

  Once molecules collide, those that are shaped peg to hole like Lego blocks snap dutifully together. All of this assembly, unlike our building of materials, is energetically “downhill.” It’s order for free. Proteins are amenable to this sort of self-assembly because of their shapes and their “electric” personalities (how their charges are distributed). These precise qualities are set forth by genes—informational templates that contain the code for making proteins. Once gene-templated proteins self-assemble into their accordion sheets, they themselves become the templates for making exquisite shells. The templated becomes the template.

  Which leads to nature’s fourth trick of the trade—the ability to customize materials through the use of templates. Whereas we muddle by in our industrial chemistry with final products that are a mishmash of polymer-chain sizes, with most too long or too short to be of ideal use, nature makes only what she wants where she wants and when she wants. No waste on the cutting-room floor.

  If we want to emulate nature’s manufacturing, we have to get backstage and interview the proteins, those templaters that make precision assembly possible at body temperatures. We have to learn their amino acid sequences and figure out how to produce them in commercial quantities. With the help of these “invisible hands,” the biomimics hope we may be able to sculpt with geometric precision, and do away with “heat, beat, and treat.”

  The Great Protein Sequence Hunt

  Mehmet Sarikaya’s eyes, the color of Turkish coffee, flash a warning to each member of the biomimicry team. “Before we do anything, we’ve got to find the protein sequence.” He is literally straining with impatience, determined to be part of the first team to find that protein-sequencing data. “We are not the only lab working on this,” he confides to me at a harried luncheon meeting, “but we are the only ones on the right track.” As he describes it, the race for a test tube full of honest-to-god, framework-and-wallpaper proteins is furious, and Sarikaya, elbows flailing, wants to win. I briefly imagine him crossing the finish line and renaming the field Biomehmetics. Later, when I tell my joke to someone who works for him, they say they are sure he has already proposed it.

  Right now, Sarikaya is on the warpath because he feels the team is stalled. I am attending a preparatory meeting for an upcoming science conference at which team members will present their work. Rich Humbert, the abalone diver-scientist, is showing pictures of his latest experiments. So far, Humbert has managed to get a random mix of abalone proteins to form “artificial pearls” against the side of a test tube. When the pearls are cut open and magnified, you can see protein (stained orange) crowded into circular layers. This layered “jawbreaker” doesn’t have the exquisite brick-and-mortar architecture of real nacre, but at least it implicates protein in a supervisory role. This has plunged Humbert deep into speculation about how nacre development might have evolved, and he would like to write a paper about it. Sarikaya fumes about the time it will take.

  He wants Humbert to find the abalone proteins responsible for nucleation, so the team can attach them to the surface of an object, dip the decorated object in seawater, then watch the nacre crystallize. The sooner the better. The military is equally interested in this idea of stronger coatings, because it, like the abalone, is often in zones of serious insult and injury, where fracture resistance would be a virtue. To that end, the Office of Naval Research has awarded a three-year grant to the UW team to investigate abalone shell, a study of what they call “layered nanostructures.”

  The team at the University of Washington is wonderfully interdisciplinary, and it is here that I see the future of biomimicry. Engineers and materials scientists are working alongside microbiologists, protein chemists, geneticists, and Renaissance thinkers like Clement Furlong.

  If there is a counterpoint to Sarikaya’s intensity, it is Clem Furlong’s ease and patience. Furlong is Rich Humbert’s supervisor and leader of his own department in medical genetics. Deep in the maze of a huge building, I find him shoehorned into an office that threatens to collapse around and on top of him. Papers are stacked atop filing cabinets all the way up to the high ceilings. Tables are heaped with journals from half a dozen disciplines, and computers lie about in various stages of undress, their circuitry hanging like mattress stuffing. Furlong and his students have just built five computers from mail-order parts this week, and he is positively gleeful about how easily one can assemble a Ferrari of a machine. He finds a piece of blank paper (no small task in that office) and writes up a parts list for me, with exact prices from memory, as if he were writing a recipe for his favorite hors d’oeuvre. For Furlong, I suspect, science is a way to get paid for tinkering.

  Somewhere in those stacks—he points to the dusty neighborhood near the ceiling panels—there are patent certificates for Furlong’s inventions. He has a hefty vita as well—lots of papers on medical genetics—but he seems most proud of the things he has made. A new Furlong invention, in fact, may be instrumental in the team’s quest to mimic abalone shell.

  “Once we sequence the protein,” he says, “we’ll have to find a way to produce lots of it. We can’t continue chopping up the shells.” Besides the risk of overcollecting the species, the grinding is hard on the proteins—it either truncates or destroys them.

  An alternative would be to conscript the trusty E. coli bacteria (found in the human gut) to make those proteins for us. It wouldn’t be the first time we harnessed bacteria to help us make products. For thousands of years, we have used yeast, bacteria, and molds for brewing beer, making wine, leavening bread, and culturing cheese. Today, bacteria grown in vats are persuaded to produce food additives, antibiotics, industrial chemicals, vitamins, and more. We have bred the tiny microbes like livestock, customizing them through artificial selection.

  There’s a difference, however, between this kind of bioprocessing and the modern version, called biotechnology. With biotechnology, we genetically alter a bacterium’s manufacturing processes by splicing in a gene from another species. To make insulin, for instance, we take the human gene for insulin manufacture and splice it into E. coli. By cutting and splicing, genetic engineers assure me, they are simply imitating a technology that bacteria themselves have long practiced. Genes from one species of bacteria are freely transferred to completely different species of bacteria. That’s how the global microcosm has been able to adapt so quickly to cataclysmic change. But human genes to bacteria? Abalone genes to bacteria?

  No matter how many times I hear scientific assurances of safety, I can’t shake the feeling that it is the height of hubris for us to cross that interphylum line, to take a gene from one class of animal and insert it into another. I tell them I would be more comfortable if we could culture whole cells from the abalone in a vat, and milk protein from those cells. For many reasons, they tell me, this is not yet practicable.

  So I am left with a dilemma that cropped up often in researching this book. Counterbalancing my real fear about genetic engineering is my real desire for us to find more benign ways of manufacturing. With my ears open and my caution up, I learned what I could about this technique, all the while hoping the problems with cell culturing would be ironed out soon.

  Once the protein is sequenced (soon, says Humbert), the dip-and-coat procedure for making nacre will be halfway home. Knowing the protein’s makeup, team members will use a machine to synthesize a segment of DNA that is the recipe template for “how to make nacre protein.” They’ll insert this DNA into E. coli, and hope for the best. With luck, the E. coli will follow the coded instructions and use its own cellular machinery to manufacture the proteins to order. It will essentially be a farming operation, where bacteria, like so many milch cows, produce a continuous stream of ceramic-crafting proteins.

  That’s where Clem Furlong’s latest device will come in handy. Furlong’s bioreactor will house the E. coli and provide them with food, water, and air, thus automating the production of proteins. The prototype bioreactor looks like a small shoebox with glass walls. Ten or twelve t
ransparent partitions slide into the box like slices of bread in a loose loaf. On each glass partition there are thousands, millions, of immobilized E. coli capable of producing one perfect protein after another. A flow of liquid nutrients surrounds them, and oxygen bubbles up from the bottom.

  As Furlong explains, “The same flow that carries in nutrients will, at the other end of the box, flush and carry off the protein they are producing. This protein—call it protein A—will flow into a beaker. But say you wanted an assembly of two proteins. You could engineer one strain of E. coli to produce protein A, another to produce protein B, and then place them in fifty-fifty proportions on the glass slides. You’d then have proteins A and B flowing into solution, finding one another, and self-assembling in the beaker. Want a different combination of proteins? Put a different slice of protein factories in.”

  The proteins can be anything the biomimic might imagine—proteins that would nucleate an even harder coating than abalone, or perhaps a thin film of crystals with electrical or optical qualities. While Furlong dreams of how we might use the bioreactor, Humbert and company are trying to find the abalone proteins that will take the shakedown cruise.

  Rich Humbert describes this protein identification, sequencing, and cloning strategy as if he’s telling me how to cook a roast. First you extract a stew of proteins from the intervening layers of the nacre and try to separate out and identify as many proteins as you can. Most of them turn out to be insoluble (they won’t stay dissolved in solution), and as such, they aggregate at the bottom of a vial and can’t be separately named. Those that do dissolve in an acetic acid solvent are all you have to work with; to separate them, you first run them through an electrified gel.

  To prepare for this gel electrophoresis, you add detergent to the proteins, which neutralizes their charges and equalizes their shapes. You then pour the soapy proteins near the top of a slab of polymer gel and throw the switch, shooting an electric charge through the gel. This starts the proteins shimmying down through the gel, moving at different speeds depending on how heavy they are (the lighter they are, the faster they are). After a while you see a banding effect as the proteins settle to certain locations in the gel.

  Each band represents a different protein. You transfer these bands to a paperlike sheet and literally cut out the bands of purified proteins or fragments of proteins and place them in separate vials. Then you take each vial and expose the proteins to another lab technique called protein sequencing. Using enzymes that are specially designed to chew off one amino acid at a time, you figure out the lineup of amino acids in each protein. Then you congratulate yourself, take a deep breath, and put on more coffee, because you’ve still got a ways to go.

  Fishing for Templates

  One of the key discoveries in molecular biology is the procedure that enables scientists to find the gene or the portion of a gene that is responsible for producing a particular protein. Like contestants in a game of Jeopardy!, gene hunters work backward. They are given the answer—protein—and they have to find the question that would have generated that answer.

  That question—the code for protein—is a carefully crafted DNA segment sitting in the cells of abalone. To find this particular strand of nucleic acid in the huge abalone genome, you make yourself a probe: a piece of DNA that will match, and stick to, the DNA you want to find.

  You can make a DNA probe from scratch using a machine that automatically strings together designated sequences of nucleotide bases, the subunits of DNA. You simply dial up an A (adenine), T (thymine), G (guanine), or C (cytosine), and the machine drips the base out of a vial and welds it to the end of a growing string called an oligio. (What’s amazing to me is how scientists know which bases to dial to code for a particular protein. We know, Humbert explains, because we know the DNA code representing each of the twenty common, natural amino acids that occur in all life-forms. This genetic code, one of the truly amazing findings of our time, is simple enough to be printed on a 3-inch by 3-inch chart. Most labs keep it taped right on the oligio machine.) Using other beguilingly simple genetic engineering techniques, you make millions of copies of this probe. Now you’re ready to go fishing.

  The other part of the process is building a fishing pool of segments of complementary DNA (cDNA) that you derive from the abalone. This process is called making a cDNA library. From one of the big scientific supply houses, you order a kit that essentially takes the tissue from the abalone and transforms the messenger RNA found in the cells into complementary DNA. Then you go fishing in this pool of cDNA, trolling until your DNA probe finds a complementary strand and sticks to it.

  The matchup is possible because of the laws of complementarity. That is, if you have a base A on your DNA strand, it will always match up with a base T on the cDNA, a C will always bond with a G, and so on. Chances are, your relatively small fishing probe will hook onto a much larger segment of cDNA, thereby calling attention to the whole gene—the one that holds the instructions for how to make an abalone shell protein. If all this works, you fish out that abalone gene, convince an E. coli to accept it, and cross your fingers in the hope that it will produce, or “express,” the protein for you.

  To find out whether the E. coli has cooperated, you need some way of seeing which colonies (out of thousands spread onto petri dishes) are producing abalone protein. The best way to do this is to go fishing again with another biological probe, this time a molecule that excels at recognizing proteins: an antibody. Our immune system produces antibodies by the millions when we are invaded by a foreign molecule. Like attack troops, the antibodies recognize this foreign object by its shape, then glom on and interfere with its functioning. What Humbert and company need are antibodies that will glom on to the shell proteins in a plate of E. coli. For this trick, they pull out a rabbit.

  After Humbert purifies the protein from nacre, he will inject some of this mollusk protein into a rabbit. The rabbit’s immune system, unused to mollusk proteins, will see them as foreign and create antibodies shaped to fit them. Humbert will then extract these antibodies from the rabbit’s blood and modify them so that the next time they attach to a protein, the attachment will trigger an effect that Humbert will be able to detect with his instruments. Thus labeled, the antibodies are then spread onto the dishes of E. coli, and if abalone proteins are anywhere on the plate, the antibodies will head right over and stick to them. Using instruments to detect a “score,” Humbert can then pluck out those E. coli colonies that are expressing the mollusk protein and let them reproduce to their heart’s content. They and their offspring will be the new tenants of Clem Furlong’s condo by the sea—the bioreactor.

  But what happens when we do find a way to produce abalone proteins to our heart’s content? Will our crystals grow as well as abalone’s do? Can we use slightly different proteins and produce slightly different, custom-made crystals? These questions can be answered only by going through the motions—setting out proteins or protein analogues and letting them grow crystals.

  Growing Crystals Nature’s Way

  Galen Stuckey, Department of Chemistry, and Daniel Morse, Department of Molecular, Cellular and Developmental Biology, University of California at Santa Barbara, have learned as much as they need to know about abalone proteins, and they’re moving on. Like the Washington team, they found it difficult to break the stalemate of insoluble proteins, those that lump together in the bottom of the beaker instead of yielding to water. Even those that could be dissolved rarely revealed their complete amino acid sequence. Rather than wait for a complete sequence, Stuckey and Morse decided to bank on the one large clue that kept turning up: the preponderance of acidic amino acid groups in all the proteins they could measure. They made themselves a protein analogue—a simple chain of acidic amino acids—as a stand-in for the real thing.

  Hoping to see mineralization in the act, they first had to convince the protein analogue to embed itself on a surface that would act like the walls and floors and ceilings of the abalone’s scaffolding
. The surface they chose is called a Langmuir-Blodgett, or L-B, film. Basically, it’s a slick of tadpole-shaped molecules that float atop a pan of water. Each molecule’s bulbous head is a charged group and the fatty tail is neutral. Because water is slightly charged, the charged head is attracted, while the neutral tail is repelled. To create an L-B film, these molecules are spread onto a shallow tray of water, and then herded together by a boom that moves across the surface. The boom actually squishes the molecules together until they “stand up”—with water-loving heads buried in the surface, and tails extending above. In the cartoon sketches that scientists have drawn for me, an L-B film looks like a putting green of grass blades.

  To get crystals to grow from this ceiling of molecules, Morse pours some zigzag, accordion-sheet proteins into the tray of water. With the help of chemical hooks, the neutral side of the protein sheet embeds itself in the fatty film ceiling, while the negatively charged pleats hang down into the water, creating a wallpaper of landing sites, just as in the abalone’s “rooms.” He then adds mineral ions to the water and lets crystals grow like stalactites from the ceiling. By being able to control the placement of the nucleation sites, Morse has found he can essentially direct what kind of crystal will form. He is now on to stage two, trying to identify the “pruning” proteins that are also present in abalone, thought to float around in the abalone “rooms” and terminate crystal growth.