|
New Scientist Living factories When it comes to producing revolutionary new polymers, forget about vast factories belching out fumes and toxic waste. The latest developments in materials manufacturing involve vats of genetically engineered bacteria. Chemists have learnt over the past few years how to equip these organisms with genes that have been specially tailored to turn out designer proteins. The materials have a huge range of potential uses. Chemists have already produced a polymer that can serve as an adhesive for living tissue, and made "smart" plastics that respond to changes in their environment. The future might even hold a nonstick fried egg. The key to this infant industry is the ability of bacteria to turn out complex proteins with absolute precision. Proteins are long-chain molecules made up of amino acids joined to each other by links known as peptide bonds. Some of these natural proteins have properties that are a match for any synthetic polymer. Spider silk is a protein that forms fibres stronger than those made from the high-tech material Kevlar. Keratin, the main fibrous component of hair, skin and claw, is a protein, too. And so is elastin, the tough, elastic component of ligaments, arteries and lungs. The molecules of these structural proteins are composed of quite short sequences of amino acids that repeat again and again along the polymer chains. These repeat sequences - involving up to 20 different types of amino-acid unit - determine way the chains fold up, and this in turn is responsible for the bulk properties of the substance. For instance, keratin is so strong because it contains helical protein chains that twist around each other in a hierarchical way: strands twist into superstrands, and superstrands into clusters and bundles to form exceptionally strong, rope-like fibres. The sequence of amino acid units in a natural protein is defined by the DNA molecules in the cells that produce it - specifically by the sequence of chemical groups called nucleotides strung along the DNA chain. In effect the DNA supplies a template on which the protein is put together, one amino acid at a time. Scientists now have ways of tapping into, and altering, this template by snipping open the double-helical strands of a DNA molecule, and replacing one nucleotide coding unit with another or splicing prefabricated sequences - built from scratch using chemical methods - into the chain. They know, too, how to multiply the products of genetic manipulation to obtain plentiful supplies of any stretch of DNA. Armed with this knowledge researchers have been exploring ways of inserting the blueprints for artificial structural proteins into bacterial DNA. It doesn't matter much to a bacterium whether an artificial gene in its DNA codes for a protein like one that the bacterium produces naturally or one that is entirely different: if its in the genome, the bacterium will make it. In the late 1980s, Joseph Cappello and his coworkers at Protein Polymer Technologies (PPTI) in San Diego, California, pioneered the techniques that are now generally used for making artificial proteins. The first step is to make an artificial DNA sequence that encodes one repeat unit of the protein polymer the researchers are aiming for. These stretches of synthetic DNA are generally made by attaching nucleotides one by one to build up a chain anchored to a solid support. Once these DNA sequences have been synthesised they can be replicated in large numbers using the polymerase chain reaction (PCR). They are then joined end to end to produce the synthetic "gene" that will encode the polymer. The linking process is helped along by the enzymes, called ligation enzymes, that do this job in nature. The next step is to insert the artificial gene into bacterial DNA. Whereas the DNA of multicellular organisms is packaged into chromosomes and sits in a cellular nucleus, a bacterium's DNA takes the form of several double-stranded rings that float about freely in the cell. As well as a main ring, bacteria have smaller rings of DNA called plasmids. It is these that are used as the vehicles for the artificial genes. Using enzymes it is possible to snip open a plasmid and stitch in the artificial gene. When the plasmids are put back inside the bacteria, the cell's molecular machinery then goes about its business of translating the DNA blueprint embodied in the plasmid into individual protein molecules - only now within the blueprint is a plan for a designer protein specified by the altered DNA. It is hard to imagine a cleaner, greener way of making materials. The bacteria are cultured in water at around body temperature and fed on a diet of amino acids. No potentially toxic organic solvents are needed, and the very nature of the process requires conditions in the fermentation vessels that support life. Sticking point In 1990 the PPTI team used this technique to persuade the common gut bacterium Escherichia coli to make an artificial silk-like protein. Cappello and his colleagues then aimed for a more ambitious target: a designer protein that would combine the strength of their artificial silk with the useful properties of a biological protein called fibronectin, which helps cells to stick together in the body. The researchers hoped to come up with a way to make cells grown in tissue culture stick to normal human cells. Artificially grown tissue is already proving valuable in promoting wound healing, and it might eventually be possible to grow entire organs this way to replace damaged or malfunctioning ones. Existing methods for making human cells stick to synthetic substrates such as cultured tissue cells use either natural cell-adhesion proteins extracted from blood or from animals, or purely synthetic peptides which contain certain key amino acid sequences. But there are difficulties with both approaches: natural products are much stickier than the artificial ones, but they are also much less stable. The PPTI team hoped to make a protein that combined the stability of artificial agents with the stickiness of natural ones. To do this, they devised an artificial protein in which silk-like structural blocks of six amino acids appeared repeatedly along the chain, interrupted every ninth block by a single fibronectin-like block of 16 amino acids. This entire sequence was itself repeated 13 times to form the complete protein molecule containing more than 900 amino acid units. To polymer chemists this is a large, complicated structure, and it would be extraordinarily difficult to reproduce it exactly by standard chemical synthesis. But compared with the complex proteins that bacteria make routinely, the structure is pretty simple. Perhaps too simple. For reasons not yet understand, cells do not take kindly to repetitive DNA sequences in their genomes. The cell's molecular machinery has a tendency to rearrange such sequences by cutting and pasting the DNA, thereby scrambling the repetitive structure. So Cappello's team took advantage of the fact that the genetic code has a degree of redundancy: because there are only 20 types of amino acid in natural proteins but 64 possible three-nucleotide combinations of the four nucleotides that encode them, each amino acid can be represented by more than one triplet. So by varying the triplet combination that represents a given amino acid, it is possible to generate repetitive amino acid sequences from a DNA sequence that is not itself quite as repetitive. Home and dry After mastering tricks like this, and by providing the bacteria with various "promoter" molecules to enhance the efficiency with which they convert the DNA to proteins, Cappello and his colleagues were able to obtain their artificial protein as a dry powder that was about 85 per cent pure. The PPTI researchers found that a coating of their synthetic protein would indeed enable cells to become attached to a variety of substrates, sometimes giving better results than fibronectin itself. The structural blocks, meanwhile, ensured that the protein stayed stable in temperatures at which the natural proteins would lose their activity. The protein is now marketed commercially by PPTI as an adhesive for attaching mammalian cells to tissue cultures, under the name ProNectin F. Other research groups have since used this approach to make a variety of synthetic proteins with structures similar to those of natural fibrous proteins. Last year, David Kaplan at the US Army's Natick Research, Development and Engineering Center in Massachusetts used bacteria to express artificial genes based on the ones that the orb-weaving spider uses to make its silk. From this it might be possible to make silk-like materials much more plentiful, as they could be grown in fermentation vats rather than being harvested from the spiders themselves. Spider silk is extremely strong, and the possible applications for artificial materials with silk-like properties range from aircraft engineering to reinforcement for bullet-proof vests. Dan Urry at the University of Alabama, meanwhile, has used the bacterial method to make synthetic proteins similar to elastin. The molecules of the natural protein have a coiled structure, and their spring-like behaviour makes the material elastic. But this structure changes with temperature: as the protein is warmed, water is stripped from around the chains and they collapse into a dense, sticky mass. Urry has made artificial polypeptides built up from the five-unit aminoacid sequence valine-proline-glycine-valine-glycine - the pattern that is characteristic of elastin. He found that, like elastin, these artificial peptides change their structure at certain temperatures. When heated, the loosely coiled chains switch to a more ordered, more tightly coiled state, in which much of the water surrounding the chain is squeezed out; this causes sheets of the material to contract. As the peptide chains coil up they exert a substantial force - some of Urry's proteins can lift more than 1000 times their dry weight as they contract. The temperature at which the conformation of the chains switches depends on their precise amino acid composition, so by tinkering with the genes that make the protein Urry can vary the synthetic polymer's properties. He believes that his synthetic elastin-like polymers might prove useful for wound repair. More prosaically, there is the possibility of producing a super-absorbent material for disposable nappies. One apparent limitation of the bacterial method is that it is restricted to just the 20 amino acid building blocks found in nature. Or is it? David Tirrell of the University of Massachusetts at Amherst is interested in using unnatural fluorinated amino acids to form polypeptide chains that might have properties similar to polytetrafluoroethylene (PTFE), the basis of the nonstick coating Teflon. Chemical groups containing fluorine are usually keen to avoid water. So Tirrell reasoned that if a polypeptide could be made with a backbone of water-soluble units interspersed with fluorinated amino acids, it might fold up into a shape in which the fluorinated groups sat on the outside, yielding a kind of nonstick protein. To put this plan into practice, however, requires a touch of bacterial husbandry as well as genetic engineering, because most bacteria will turn up their noses at fluorine-containing compounds. Tirrell's answer has been to breed a new strain of bacteria that could use the fluorinated amino acids for protein synthesis and that would not synthesise their own unfluorinated amino acids instead. The carrier the researchers chose for the fluorine atoms was phenylalanine, which has a benzene ring sidegroup to which fluorine atoms can be attached. They starved the bacterial culture of phenylalanine and fed it the fluorinated variant instead. Sure enough, the resulting polymer turned out to be insoluble in water, suggesting that the fluorinated groups did indeed migrate to the surface of the folded polymer. While he was pondering the possible uses for such a material, Tirrell discovered on his desk one day a mischievous suggestion from his graduate students. The note showed a picture of a fried breakfast in the making: an egg frying in a nonstick pan. Next to it was a picture of a breakfast of the future - which was being made in an old iron frying pan, because the egg itself was nonstick. Its protein component, the albumin that makes up most of egg white, had been fluorinated. Although Tirrel doesn't expect to be making non-stick fried eggs in the near future, he sees huge opportunities in this new-found ability to redirect the protein-synthesising abilities of bacteria. He imagines tagging natural proteins with a synthetic polypeptide chain, by inserting the DNA sequence for an artificial polypeptide right next to the sequence for the natural protein in the bacterial plasmid. For example, a protein might be given an artificial side chain that will fold up into a crystalline film (Tirrell's group has already built synthetic proteins that will do this). Then such hybrid proteins might spontaneously form thin films on a substrate, which would immobilise them for use in devices such as bioreactors and biosensors. Or a protein might be equipped with an optically or electrically active substituent, made up of suitably modified, non-natural amino acids, that could make it possible to control the protein's function with light or electricity. Tirrell's researchers have already made a synthetic protein containing 3- thienylalanine in place of the natural amino acid alanine. The sulphur- containing thienyl group was used because it is the building block of polythiophene, a polymer that conducts electricity. Tirrell aims to link up the thienyl groups protruding from a film of this synthetic protein to make a biocompatible conducting material for possible use in biomedical devices. Ultimately, the researchers hope that the worlds of synthetic and natural materials will start to merge, as they find ways to extract the best from both of them. Plastics with a life of their own Bacteria have already chalked up a success as a source of polymers on a commercial scale. The British biotechnology company Zeneca, a subsidiary of ICI, manufacturers and markets a biodegradable polymer called Biopol, which is synthesised in fermentation vats by a bacterium Alcaligenes eutrophus. Biopol is used in moulded plastic items such as bottles and for the controlled release of drugs. The bacteria are fed a diet of propionic acid and glucose, which they turn into a polyester. This acts as an energy reserve for the bacteria and is stored as granules inside the cell, just as our own cells store fat. When extracted from the cells and collected, it provides a flexible material similar to polypropylene - except that it is fully biodegradable into harmless compounds. ICI is now using A. eutrophus and other bacteria to develop a range of different biodegradable plastics. While researchers can exercise some control over the
nature of these plastics by altering the chemicals
in the feedstock, the synthetic procedure within the
cells is dictated by the bacterial enzymes and
cannot easily be changed. Furthermore, the products
will always be something of a mxied bag, with slight
variations in the detailed molecular structure,
since the bacteria are concerned only with producing
an energy store, and don't need to be perfectionist
about the precise structure of the polymer chain.
|
