Self-organization, Natural Selection, and Evolution: Cellular Hardware and Genetic Software

Articles Self-organization, Natural Selection, and Evolution: Cellular Hardware and Genetic Software Self-organization is sometimes presented as an alternative to natural selection as the primary mechanism underlying the evolution of function in biological systems. Here we argue that although self-organization is one of selection’s fundamental tools, selection itself is the creative force in evolution. The basic relationship between self-organization and natural selection is that the same self-organizing processes we observe in physical systems also do much of the work in biological systems. Consequently, selection does not always construct complex mechanisms from scratch. However, selection does capture, manipulate, and control self-organizing mechanisms, which is challenging because these processes are sensitive to environmental conditions. Nevertheless, the often-inflexible principles of self-organization do strongly constrain the scope of evolutionary change. Thus, incorporating the physics of pattern-formation processes into existing evolutionary theory is a problem significant enough to perhaps warrant a new synthesis, even if it will not overturn the traditional view of natural selection. Keywords: natural selection, self-organization, complexity theory, adaptation, evolution A diverse group of researchers in mathematics, physics, and several branches of biology have argued that self-organization should be placed alongside natural selection as a complementary mechanism of evolution (Nicolis and Prigogine 1977, Kauffman 1993, Camazine et al. 2001, Denton et al. 2003, Kurakin 2005, 2007, Newman et al. 2006, Karsenti 2008, Wills 2009). Ignoring these calls, most evolutionary research has continued along traditional lines. This article is an attempt to put the rich experimental and theoretical work on self-organization in biological systems into intuitive terms that demonstrate the role of self-organization in the evolutionary process. First, we briefly review evolutionary theory with an emphasis on those aspects critical to our discussion of self-organization. We then introduce self-organization with a general definition and several examples from biological systems. Following this introductory material, we explore the intersection between natural selection and self-organization. Here, we have two goals: (1) to clear up the misunderstanding that self-organization competes with natural selection as the organizing force in evolution, and (2) to explore the myriad ways that self-organization affects the evolutionary process. In the course of these discussions, we argue that evolutionary biology historically has dealt with the gradual evolution of structures and plans, whereas the evolution of controlled but largely spontaneous processes is the chief unanswered question of today. Further, as several authors have suggested (Kauffman 1993, Kurakin 2005, 2007, Wills 2009), evolution at the macroscopic level and evolution dependent on self-organization at the molecular level are so different that the resulting expansion of evolutionary theory ultimately may be considered a new synthesis. As the first modern synthesis incorporated genetics into natural selection, this new synthesis seeks to incorporate the physics of complex systems (Nicolis and Prigogine 1977, Kauffman 1993, Camazine et al. 2001, Newman et al. 2006, Karsenti 2008). A brief historical preface Although evolution by natural selection is one of biology’s most well-supported theories, it is important to keep in mind that the theory’s common derivation was fully formed before the rise of molecular biology. We cannot overemphasize this fact because it sets the stage for our basic problem, which is that evolution by natural selection was conceived using data at the macroscopic level. This does not imply that principles derived from macroscopic studies will not be applicable to the molecular world (they often are); it simply means that we could sometimes face an apples-and-oranges problem when we apply traditional evolutionary principles to the evolution of molecular mechanisms. When molecular biologists suggest that traditional approaches to natural selection do not seem fully appropriate to their systems, an evolutionary biologist should not be skeptical. In fact, given the history, it would be surprising if a major new approach to evolution were not necessitated by data on life at the molecular level. BioScience 60: 879–885. ISSN 0006-3568, electronic ISSN 1525-3244. © 2010 by American Institute of Biological Sciences. All rights reserved. Request permission to photocopy or reproduce article content at the University of California Press’s Rights and Permissions Web site at www.ucpressjournals.com/ reprintinfo.asp. doi:10.1525/bio.2010.60.11.4 www.biosciencemag.org December 2010 / Vol. 60 No. 11 • BioScience 879 Brian R. Johnson and Sheung Kwan Lam Articles 880 BioScience • December 2010 / Vol. 60 No. 11 The third concept follows from the second, and concerns the nature of evolutionary trajectories. Does selection quickly form new adaptations by favoring radically different variations of phenotype, or does selection lead to gradual changes in phenotypes that vary only quantitatively (Maynard-Smith 1986)? The last concept has to do with the relationship between genotype and phenotype and is involved in the basic principle of heritability (Maynard Smith 1986, Müller 2007). Here the question is, How expansive and complex are the genetic bases of traits? Are traits controlled by a few genes that work independently, or are genes organized into large networks that contribute to one or more modules of biological organization (Müller 2007, Monteiro and Podlaha 2009)? If genes work independently, then the effect of differential reproductive success, on average, will be to shape traits independently of one another. If genes often affect multiple traits, however, then selection acting on one trait can affect many others as well. This last concept is partly encompassed by the notion of the evolutionary constraint, as it explores proximate explanations for the constraint of variation. Self-organization In contrast to conservative systems, in which energy is conserved, self-organization occurs in dissipative systems through which energy is flowing (Nicolis and Prigogine 1977, Kauffmann 1993, Camazine et al. 2001). Such systems produce what are called dissipative structures. A dissipative structure is one that breaks down without the continual input of energy (Nicolis and Prigogine 1977, Maynard Smith 1986). A dissipative structure is thus not a structure at all, but a metastable pattern. For instance, many weather patterns, such as clouds and hurricanes, are dissipative structures (Whitesides and Grzybowski 2002, Arsenyev et al. 2004); in fact, we are surrounded by complex, purely physical self-organization patterns. In this section, we introduce the nature of selforganization in biological systems. For reasons (...truncated)