... and finally the third part of the book is about cool stuff, such as nanotech and new materials. Sometimes we aren't quite as clever as Nature, so one of the coolest materials ever, spider silk, is still beyond our grasp:
They may have eight hairy legs, a very small brain and a big image problem, but there is at least one thing that spiders can do much better than we can, namely producing extremely strong fibers.
Nature is in many ways a better engineer than mankind. Whether you look at the ways that diatoms, mussels or snails make their shells, butterflies seem to change their colors, or trees withstand strong winds, there is always a lesson for human engineers to be learnt. And sometimes, nature’s engineering is so clever we still have trouble figuring out how it works. But when we do, the rewards to be gained from the lesson can be enormous.
One of the most spectacular examples of how nature beats our best efforts is spider silk. Like our hair, sheep wool and silk clothes, it consists mostly of protein. But the polypeptide chains are aligned and interwoven in mysterious ways that make the product much stronger than these materials. Optimized by evolution to be able to stop an insect in full flight, this is in fact the strongest material we know in strength per weight terms. If you compare a spider’s thread with a steel wire of the same diameter, they will be able to support roughly the same weight. But the silk is six times lighter, so it is really six times stronger than steel, and the spider wins every time.
So why are suspension bridges still dangling on steel ropes rather than silken ones? The trouble is we can’t make spider silk as nicely as the spider can. Sure, we can express the proteins of which it is made in other organisms, including goats that will have spidroin in their milk (to which I will come back later), and we will soon be able to spin those into some kind of fiber, but given our limited understanding of the processes going on in the spider’s silk gland, the result may not live up to the natural product.
A small scattering of biologists in various labs around the world are trying to get behind the spiders’ secret. First they need the right genes. Until recently, only very few DNA sequences of silk protein genes were known. In 2001, John Gatesy and Cheryl Hayashi with their coworkers at the University of Wyoming in Laramie presented a comprehensive overview of gene sequences from a wide variety of eight-legged silk producers including tarantulas and other animals that separated from the “true spiders” more than 200 million years ago. They showed that the amino acid sequences are extremely diverse between species. About the only thing they all have in common is the occurrence of unusual repetitive sequences following four simple patterns: poly-alanine (An), alternation of glycine and alanine (GA), then combinations of glycine with a small subset of amino acids (X) with or without proline: GGX and GPGXn.
Given that these motifs have been retained (or evolved convergently) over a time span of more than 200 million years, it appears likely that their properties hold important clues to how the silk proteins interact to form silk. So far the only structural information we have is about the finished silk, where it is known that the alanine-rich repeats occur in quasi crystalline domains, while the glycine-rich repeats adopt more disorderly states which are poorly understood.
From these genes, the spider makes the corresponding proteins, called fibroins. Nothing special there - proteins can make hair and wool and the brittle type of silk that insects make their cocoons of. But these materials aren’t tough enough for spiders. To make the special, stronger-than-steel brand of protein filament, spiders have a special silk gland, a complex structure where the magical transformation of protein solution into silk thread takes place. Although this transformation is only poorly understood so far, it is known that it involves a substantial increase in the proportion of the protein chain that is arranged as a beta-pleated sheet.
In a 2001 review in Nature, Oxford zoologists Fritz Vollrath and David Knight, who have been studying the silk production in the orb spider Nephila clavipes for many years, summed up the current knowledge. First, it should be noted that these spiders not only have one kind of silk gland, but seven pairs producing seven different types of silk for different uses, and even the protein composition in these glands is significantly different. The one that has been best characterized is the dragline silk which is produced by the major ampullate gland.
This gland consists of three major regions: a central bag (B zone) flanked by a tail (A zone) and the duct (D) leading towards the exit (Figure XXX ???). The lining of the A and B zones contains the cells which secrete the protein material, the major component of which is a 275 kDa protein containing the polypeptides spidroin I and spidroin II. The A zone specializes on the spidroin protein which forms the strong core of the thread, while the B zone is believed to secrete the as yet poorly understood glycoprotein material that ends up coating it. To be secreted from a cell, proteins must be wrapped up in membrane bubbles called secretory vesicles. In the A zone cells, these vesicles contain protein filaments, the precise structural organization of which is still under investigation. In the B zone vesicles one finds liquid crystals of the coat glycoprotein. Knight and Vollrath think that the liquid crystalline state has an important role to play in the production of the silk thread, to which we shall return later.
Let us follow the route of a spidroin molecule from the secretion through to the finished thread. On leaving the cells of the A zone (as the secretory vesicles merge with the cell membrane and empty their contents to the outside), our protein finds itself in a small spherical droplet with lots of other spidroin molecules. The protein concentration in the whole gland is around 50% - higher than in most protein crystals. Most proteins would aggregate into insoluble lumps at much lower concentrations. This highly viscous protein mass flows down the A-tail into the bag (B-zone), where it gets coated by the secretion of the B zone cells. At the exit of the bag the liquid is funneled into the much narrower duct (D). During this transition, the droplets are slowly distorted into long thin shapes aligned with the direction of the flow. It is assumed that a similar transformation happens to the molecules. Initially they must have been in a rather compact conformation to avoid aggregation, but as they move into the duct, they are stretched out and aligned in a way which will eventually allow them to form those intermolecular links that hold the thread together.
The spinning solution or dope is now in a liquid crystalline state, with proteins aligned in an orderly fashion, but still able to slide past each other. This is thought to be an important part of the spider’s secret weapon. As the material is slowly and gently narrowed down in the first two legs of the tripartite duct, molecules have time to orient themselves in a favorable way so they can eventually form intermolecular beta sheet interactions and possibly disulphide bonds when it comes to making the actual thread. This step happens at a point ca. four millimeters before the exit, and it happens quite suddenly. Although the molecular details are far from clear, it is thought that as the dope is drawn out to a thin thread that retracts from the walls of the duct, the molecules align even further and form hydrogen bonds defining the complex beta sheet patterns found in the finished product. In the process, the protein becomes more hydrophobic and practically ejects some of the water content it has been carrying until this point. Finally, most of the water is stripped off the surface when the thread leaves the exit spigot, helping the spider to avoid water loss and making its thread even tougher.
This broad picture drawn by Vollrath and Knight combines anatomical with some structural data. However, the exact details of the crucial structural transitions are far from understood. The trouble is that the most powerful tools for the determination of protein structures, X-ray crystallography and NMR, require a protein crystal or a homogeneous solution, respectively. As yet, there is no method that could give you the atomic detail structure of a protein molecule flowing down the duct of a spider’s silk gland.
And yet, even in the absence of a full understanding in molecular terms, maybe one could copy the spider’s technique on a microscopic scale, by supplying a dope with the right protein composition and passing it through a spinning device modeled on the spider’s gland? The only comparable man-made material, aramid (KEVLARTM, the fiber used in bullet-proof vests), is spun from hot sulfuric acid. Thus, an ambient temperature process leading to something similar would be very attractive, even if the resulting fiber turned out only as good as Kevlar, and not quite as good as authentic spider silk.
But first, you need to produce the proteins in reasonable quantities. Unlike silk moths, spiders have an aggressive territorial behavior, which means they won’t be cooperating with any ideas of high-throughput farming. Expressing the silk proteins in bacteria or yeasts doesn’t work either. The curious repetitive nature of their sequences invites the microbes to take shortcuts and produce abridged versions of the protein chains.
Thus, if you want to use the silk to catch fighter jets rather than flies, you’d better get an animal that can produce more than a few milligrams of the precious material. The company Nexia Biotechnologies Inc. at Montreal, Canada, first succeeded in breeding goats that are genetically modified in a way that they secrete spidroin protein in their milk. It turned out that the secretory cells of mammary glands aren’t that different from those of silk glands, only there are a lot more of them in a goat, which makes milking goats a lot more economical than milking spiders.
Since the summer of 2000, Nexia boasts the possession of two African dwarf goats, Peter and Webster, who have been shown to carry the appropriate spider gene. A couple of breeding generations later, there will be a flock of females producing spidroin in their milk by the gram. Nexia keep mum about which way exactly they want to spin that milk-silk protein into strong fibers on an industrial scale. As soon as they can do that, however, applications ranging from surgical threads through to missile protection and aviation security will be conquered rapidly by the new material.
Although some of the applications envisaged are substantially scaled up in comparison to a spider’s web, there is also a case of scaling down from there. In an attempt to turn a visible thread into an invisibly thin nanowire, the group of Michael Stuke at the Max-Planck-Institute for Biophysical Chemistry in Göttingen stripped spider silk down to the core, using ultraviolet laser technology. They obtain very strong nanowires, currently with as little as 100 nm diameter. Plans for the future include coating this thread in metal to make it conductive.
But even when we can copy the spider’s thread and use it on various length scales, the hairy little arthropods can still do one better. As Stefan Schulz and his coworkers at the Technical University of Braunschweig, Germany, reported in 2000, the female tropical spider Cupiennius salei leaves a thread marked with sex pheromones, which induce any male of her species to vibrate excitedly. The vibrations are transmitted through the thread, which rapidly switches from a role of odor dispenser to that of a phone line. The female vibrates back, and you can figure out the rest for yourself. I wonder whether anybody wants to set up a company banking on that technology ...
J.Gatesy et al. Science, 2001, 291, 2603.
F.Vollrath, D.P.Knight, Nature, 2001, 410, 541.
M. Papke et al., Angew. Chem. Int. Ed. 2000, ...
What happened next
I recently did a feature on spider silk for Oxford Today.