Abstract
Living systems, as created initially by the transition from assemblies of large molecules to self-reproducing information-rich cells, have for centuries been studied via the empirical toolkit of biology. This has been a highly successful enterprise, bringing us from the vague non-scientific notions of vitalism to the modern appreciation of the biophysical and biochemical bases of life. Yet, the truly mind-boggling complexity of even the simplest self-sufficient cells, let alone the emergence of multicellular organisms, of brain and consciousness, and to ecological communities and human civilizations, calls out for a complementary approach.
In this editorial, we propose that theoretical physics can play an essential role in making sense of living matter. When faced with a highly complex system, a physicist builds simplified models. Quoting Philip W Anderson's Nobel prize address, 'the art of model-building is the exclusion of real but irrelevant parts of the problem and entails hazards for the builder and the reader. The builder may leave out something genuinely relevant and the reader, armed with too sophisticated an experimental probe, may take literally a schematized model. Very often such a simplified model throws more light on the real working of nature....'
In his formulation, the job of a theorist is to get at the crux of the system by ignoring details and yet to find a testable consequence of the resulting simple picture. This is rather different than the predilection of the applied mathematician who wants to include all the known details in the hope of a quantitative simulacrum of reality. These efforts may be practically useful, but do not usually lead to increased understanding.
To illustrate how this works, we can look at a non-living example of complex behavior that was afforded by spatiotemporal patterning in the Belousov–Zhabotinsky reaction [1]. Physicists who worked on this system did not attempt to determine all the relevant chemical intermediates, nor collect data on the kinetics of the many complex reactions. Instead, the focus was on formulating two- or three-component reaction–diffusion equations (e.g. the Oregonator), which could explain such generic features as the existence of rotating spiral waves (and their instability), the transition to chaos, the control of the reaction by light etc. By stressing mechanism instead of meticulous detail, one could understand the system even if there were still components and interactions waiting to be cataloged and quantified.
In living systems, this way of thinking is even more crucial. A leading biologist once remarked to one of us that a calculation of in vivo cytoskeletal dynamics that did not take into account the fact that the particular cell in question had more than ten isoforms of actin could not possibly be correct. We need to counter that any calculation which takes into account all these isoforms is overwhelmingly likely to be vastly under-constrained and ultimately not useful. Adding more details can often bring us further from reality. Of course, the challenge for models is then falsification, i.e., finding robust predictions which can be directly tested experimentally. The most severe criticism, to quote Pauli, remains that 'your model is not even wrong.'
Is this approach proving successful? In many cases it is too early to tell, but simple models have already proved useful in understanding protein folding, directed cell motility, gene expression variability and even laboratory-scale Darwinian evolution. One could argue as well that the extremely influential Hodgkin–Huxley approach to the action potential in neurons is a vastly oversimplified description and that is why it is tractable and compelling. There are other cases, however, where the model was too simple—the simple Turing instability does not account for Drosophila segment formation [2] and the binary state relaxational dynamics of the Hopfield model may prove incapable of explaining memory capacity of more intrinsically non-equilibrium neural networks. But we learn by failure what is essential, as opposed to having so many parameters to fit that we are able to reproduce the data and thereby mask our incorrect hypotheses.
Recognizing the opportunity and growing community of physicists working on living systems in ways that are different from traditional biology, the National Science Foundation Physics Division has developed a program called 'Physics of Living Systems (PoLS)', located in the Mathematical and Physical Sciences Directorate. The program has as its primary goal to fund physics research on living systems ranging from single cells to communities of organisms. The synopsis of the program can be found at www.nsf.gov/funding/pgm_summ.jsp?pims_id=6673&org=DAS. It is also focused on how inanimate organic molecules cooperatively form living cells and how their properties are modified inside the cells. This effort is shared with Molecular Biophysics program in the Directorate for Biological Sciences (www.nsf.gov/funding/pgm_summ.jsp?pims_id=504858), which supports research at the molecular level and hence is complementary to PoLS. Recently, these two programs have initiated and funded a network of universities across the US for sharing students and resources in the field of PoLS. This concept has now been spread to the international arena, where several countries have formed the International PoLS Network (iPoLS) (http://pols.rice.edu). iPoLS is currently one of the three Science Across Virtual Institutes projects of NSF. A new community is being formed by the graduate students across the network, who collectively will transform our understanding of living systems.
Physical Biology stands as one of the best journals dedicated to quantitative physics research addressing all biological scales. It strikes a good balance between demanding that models have significant biological foundations and allowing physicists to explore conceptual mechanisms that cannot as yet be completely grounded in known molecular interactions. As the physics of living systems comes of age, we expect to see many seminal ideas that are first broached in these pages.
References
[1] Field R J and Noyes R M 1974 Oscillations in chemical systems: IV. Limit cycle behavior in a model of a real chemical reaction J. Chem. Phys. 60 1877–84
[2] Nüsslein-Volhard C and Wieschaus E 1980 Mutations affecting segment number and polarity in Drosophila Nature 287 795–801
Export citation and abstract BibTeX RIS