The most compelling argument for the possibility of a radical nanotechnology, with functional devices and machines operating at the nano-level, is the existence of cell biology. But one can take different lessons from this. Drexler argued that we should expect to be able to do much better than cell biology if we applied the lessons of macroscale engineering, using mechanical engineering paradigms and hard materials. My argument, though, is that this fails to take into account the different physics of the nanoscale, and that evolution has optimised biology’s “soft machines” for this environment. This essay, first published in the journal Nature Nanotechnology (subscription required, vol 1, pp 85 – 86 (2006)), reflects on this issue.
Nanotechnology hasn’t yet acquired a strong disciplinary identity, and as a result it is claimed by many classical disciplines. “Nanotechnology is just chemistry”, one sometimes hears, while physicists like to think that only they have the tools to understand the strange and counterintuitive behaviour of matter at the nanoscale. But biologists have perhaps the most reason to be smug – in the words of MIT’s Tom Knight “biology is the nanotechnology that works”.
The sophisticated and intricate machinery of cell biology certainly gives us a compelling existence proof that complex machines on the nanoscale are possible. But, having accepted that biology proves that one form of nanotechnology is possible, what further lessons should be learned? There are two extreme positions, and presumably a truth that lies somewhere in between.
The engineers’ view, if I can put it that way, is that nature shows what can be achieved with random design methods and a palette of unsuitable materials allocated by the accidents of history. If you take this point of view, it seems obvious that it should be fairly straightforward to make nanoscale machines whose performance vastly exceeds that of biology, by making rational choices of materials, rather than making do with what the accidents of evolution have provided, and by using the design principles we’ve learnt in macroscopic engineering.
The opposite view stresses that evolution is an extremely effective way of searching parameter space, and that in consequence is that we should assume that biological design solutions are likely to be close to optimal for the environment for which they’ve evolved. Where these design solutions seem odd from our point of view, their unfamiliarity is to be ascribed to the different ways in which physics works at the nanoscale. At its most extreme, this view regards biological nanotechnology, not just as the existence proof for nanotechnology, but as an upper limit on its capabilities.
So what, then, are the right lessons for nanotechnology to learn from biology? The design principles that biology uses most effectively are those that exploit the special features of physics at the nanoscale in an environment of liquid water. These include some highly effective uses of self-assembly, using the hydrophobic interaction, and the principle of macromolecular shape change that underlies allostery, used for both for mechanical transduction and for sensing and computing. Self-assembly, of course, is well known both in the laboratory and in industrial processes like soap-making, but synthetic examples remain very crude compared to the intricacy of protein folding. For industrial applications, biological nanotechnology offers inspiration in the area of green chemistry – promising environmentally benign processing routes to make complex, nanostructured materials based on water as a solvent and using low operating temperatures. The use of templating strategies and precursor routes widens the scope of these approaches to include final products which are insoluble in water.
But even the most enthusiastic proponents of the biological approach to nanotechnology must concede that there are branches of nanoscale engineering that biology does not seem to exploit very fully. There are few examples of the use of coherent electron transport over distances greater than a few nanometers. Some transmembrane processes, particularly those involved in photosynthesis, do exploit electron transfer down finely engineered cascades of molecules. But until the recent discovery of electron conduction in bacterial pili, longer ranged electrical effects in biology seem to be dominated by ionic rather than electronic transport. Speculations that coherent quantum states in microtubules underlie consciousness are not mainstream, to say the least, so a physicist who insists on the central role of quantum effects in nanotechnology finds biology somewhat barren.
It’s clear that there is more than one way to apply the lessons of biology to nanotechnology. The most direct route is that of bionanotechnology, in which the components of living systems are removed from their biological context and put to work in hybrid environments. Many examples of this approach (which NYU’s Ned Seeman has memorably called biokleptic nanotechnology) are now in the literature, using biological nanodevices such as molecular motors or photosynthetic complexes. In truth, the newly emerging field of synthetic biology, in which functionality is added back in a modular way to a stripped down host organism, is applying this philosophy at the level of systems rather than devices.
This kind of synthetic biology is informed by what’s essentially an engineering sensibility – it is sufficient to get the system to work in a predictable and controllable way. Some physicists, though, might want to go further, taking inspiration from Richard Feynman’s slogan “What I cannot create I do not understand”. Will it be possible to have a biomimetic nanotechnology, in which the design philosophy of cell biology is applied to the creation of entirely synthetic components? Such an approach will be formidably difficult, requiring substantial advances both in the synthetic chemistry needed to create macromolecules with precisely specified architectures, and in the theory that will allow one to design molecular architectures that will yield the structure and function one needs. But it may have advantages, particularly in broadening the range of environmental conditions in which nanosystems can operate.
The right lessons for nanotechnology to learn from biology might not always be the obvious ones, but there’s no doubting their importance. Can the traffic ever go the other way – will there be lessons for biology to learn from nanotechnology? It seems inevitable that the enterprise of doing engineering with nanoscale biological components must lead to a deeper understanding of molecular biophysics. I wonder, though, whether there might not be some deeper consequences. What separates the two extreme positions on the relevance of biology to nanotechnology is a difference in opinion on the issue of the degree to which our biology is optimal, and whether there could be other, fundamentally different kinds of biology, possibly optimised for a different set of environmental parameters. It may well be a vain expectation to imagine that a wholly synthetic nanotechnology could ever match the performance of cell biology, but even considering the possibility represents a valuable broadening of our horizons.
4 thoughts on “Right and wrong lessons from biology”
I think that one of the important aspects of biological design paradigm that you did not mention is material flow. Its my understanding that inside a cell the structures that make up the cell are constantly being created and destroyed. I have never seen a nano engineered design that does that.
Sure I have see designs that process information and energy and even some that process materials but never in the self consuming /self creating way that biology does.
But on the more general question Is biology the upper limit on what is possible with nanotechnology? I am still firmly in the engineers camp because biology is limited by systems that can evolve and engineering isn’t. The example you gave about electron flow at the nanoscale is a prime example. (The recent work on using voltage to alter the light absorption of graphene is something that may turn out to be incredibly useful.) Another nanoscale phenomenon that is not used in biology is super lubricity. And if you leave a warm water environment and move the the diamondiod / vacuum system the way heat flows through the system can be totally different. In the water system heat will radiate outward but in the diamond / vacuum system you can create “heat” circuits.
Another striking design constraint which for some reason binds biology is the use of repetitive elements in designs. Many components are made out of a single protein sticking to itself repeatedly to form a chain or ring; and many more are made out of an ABAB… pattern of proteins, or similar variations on this theme. I imagine that this is related to the need for self-assembly, that it is easier to make that work by using a limited number of parts. But at the same time I can’t help but imagine that it is extremely difficult to design a protein that can both be functional and also will stick to itself in just the right way to form a macromolecular complex with the necessary properties.
I am always struck at how much Drexlerians dismiss Humanities ignorance of how biology makes complex machines. I believe that this attitude lead to poor communications between the camps.
Therefore I have a suggestion to make.
If a great artist was to created representations of the glorious complexities of life itself, it would in my opionion advance the debate, improving communication between the differing camps.
Unfortunately, most artists tend to be leftist types who are unthinkingly critical of Science.
Jim, indeed, that is a fundamental difference. Biological systems are driven, far from equilibrium systems that are constantly turning over both matter and energy. There are few or no analogues of this in a human-made technology (even the situations in which self-assembly is exploited are generally close to equilibrium). I’m not sure that superlubricity is so crucial, though. Actually, I think that’s an unfortunate name, as it invites analogies with the phenomena of superconductivity and superfluidity that aren’t really there – in the latter two phenomena, resistance (to material flow or current) genuinely goes to zero because of quantum weirdness, while superlubricity simply refers to friction dropping to a low, but still finite value because a major cause of energy dissipation ceases to be operative. Biology (probably) has some mechanisms of ultra-low friction available to it too, as demonstrated by the experiments of Jacob Klein on the rubbing together of polymer brush decorated surfaces.
Hal, there are undoubtedly many constraints on biological “design”. Protein folding is a very obvious place to start – the fraction of possible sequences that are able to fold to a single ground state is almost vanishingly small, so chunks of foldable sequence tend to be frequently reused.
Zelah, have you seen the work of David Goodsell?
Comments are closed.