A latent capacity for evolutionary innovation through exaptation in metabolic systems

Journal name:
Date published:
Published online

Some evolutionary innovations may originate non-adaptively as exaptations, or pre-adaptations, which are by-products of other adaptive traits1, 2, 3, 4, 5. Examples include feathers, which originated before they were used in flight2, and lens crystallins, which are light-refracting proteins that originated as enzymes6. The question of how often adaptive traits have non-adaptive origins has profound implications for evolutionary biology, but is difficult to address systematically. Here we consider this issue in metabolism, one of the most ancient biological systems that is central to all life. We analyse a metabolic trait of great adaptive importance: the ability of a metabolic reaction network to synthesize all biomass from a single source of carbon and energy. We use novel computational methods to sample randomly many metabolic networks that can sustain life on any given carbon source but contain an otherwise random set of known biochemical reactions. We show that when we require such networks to be viable on one particular carbon source, they are typically also viable on multiple other carbon sources that were not targets of selection. For example, viability on glucose may entail viability on up to 44 other sole carbon sources. Any one adaptation in these metabolic systems typically entails multiple potential exaptations. Metabolic systems thus contain a latent potential for evolutionary innovations with non-adaptive origins. Our observations suggest that many more metabolic traits may have non-adaptive origins than is appreciated at present. They also challenge our ability to distinguish adaptive from non-adaptive traits.

At a glance


  1. Viability on glucose entails viability on multiple other carbon sources.
    Figure 1: Viability on glucose entails viability on multiple other carbon sources.

    a, The binary innovation vector of a hypothetical metabolic network that is viable on glucose. The vector shows that the random network is viable (labelled by 1) on glucose, sorbitol and fructose, but not viable (labelled by 0) on pyruvate and acetate. The innovation index of this network (IGlucose = 2) is the number of additional carbon sources on which the network is viable. b, The distribution of innovation indices for 500 random networks viable on glucose. Only 4% of networks have IGlucose = 0, meaning that they are viable only on glucose.

  2. Innovation varies with respect to the carbon source, C, and the mean metabolic distance between C and Cnew.
    Figure 2: Innovation varies with respect to the carbon source, C, and the mean metabolic distance between C and Cnew.

    a, For each of 50 carbon sources (horizontal axis), the figure indicates the mean innovation index (bar) and its coefficient of variation (vertical line) for 500 random networks required to be viable on that carbon source. Note the broad distribution of the index. Some carbon sources, such as acetate, allow viability on more than nine additional carbon sources, on average, whereas others, such as deoxyadenosine, support viability on fewer than one additional carbon source. The innovation index of glucose (red) is typical compared with other carbon sources. b, A hypothetical carbon source, Cnew, which can be synthesized from another carbon source, C, in one reaction (arrow), and which leads, through multiple further reactions, to the synthesis of biomass. Some metabolic networks may have an alternative metabolic pathway that bypasses Cnew altogether (right-hand sequence of arrows). c, Like b, but with Cnew and C separated by multiple reactions. The fewer reactions separate C and Cnew, the more likely it is that Cnew is not bypassed by some alternative metabolic pathway, and that viability on C therefore implies viability on Cnew. d, Testing the hypothesis in c. The horizontal axis shows the mean number of reactions that separate C and Cnew in networks that are viable on both C and Cnew, binned into integer intervals according to the floor of this number (that is, the greatest smaller integer). The vertical axis shows the fraction of random metabolic networks required to be viable on carbon source C that are additionally viable on Cnew. We note that the potential for innovation decreases with increasing distance. Box edges, 25th and 75th percentiles; central horizontal line in each box, median; whiskers, ±2.7s.d.; open circles, outliers. Data are based on samples of 500 random viable networks for each of 50 carbon sources C (n = 25,000).


  1. Darwin, C. On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life 6th edn (Murray, 1872)
  2. Gould, S. J. & Vrba, E. S. Exaptation: a missing term in the science of form. Paleobiology 8, 415 (1982)
  3. True, J. R. & Carroll, S. B. Gene co-option in physiological and morphological evolution. Annu. Rev. Cell Dev. Biol. 18, 5380 (2002)
  4. Zákány, J. & Duboule, D. Hox genes in digit development and evolution. Cell Tissue Res. 296, 1925 (1999)
  5. Keys, D. N. et al. Recruitment of a hedgehog regulatory circuit in butterfly eyespot evolution. Science 283, 532534 (1999)
  6. Tomarev, S. I. & Piatigorsky, J. Lens crystallins of invertebrates. Diversity and recruitment from detoxification enzymes and novel proteins. Eur. J. Biochem. 235, 449465 (1996)
  7. Pievani, T. & Serrelli, E. Exaptation in human evolution: how to test adaptive vs exaptive evolutionary hypotheses. J. Anthropol. Sci. 89, 923 (2011)
  8. Schuster, P., Fontana, W., Stadler, P. F. & Hofacker, I. L. From sequences to shapes and back: a case study in RNA secondary structures. Proc. R. Soc. Lond. B 255, 279284 (1994)
  9. Lipman, D. J. & Wilbur, W. J. Modelling neutral and selective evolution of protein folding. Proc. R. Soc. Lond. B 245, 711 (1991)
  10. Cowperthwaite, M. C., Economo, E. P., Harcombe, W. R., Miller, E. L. & Meyers, L. A. The ascent of the abundant: how mutational networks constrain evolution. PLoS Comput. Biol. 4, e1000110 (2008)
  11. Ferrada, E. & Wagner, A. A comparison of genotype-phenotype maps for RNA and proteins. Biophys. J. 102, 19161925 (2012)
  12. Samal, A., Matias Rodrigues, J. F., Jost, J., Martin, O. C. & Wagner, A. Genotype networks in metabolic reaction spaces. BMC Syst. Biol. 4, 30 (2010)
  13. Poulsen, T. S., Chang, Y.-Y. & Hove-Jensen, B. D-allose catabolism of Escherichia coli: involvement of alsI and regulation of als regulon expression by allose and ribose. J. Bacteriol. 181, 71267130 (1999)
  14. Meijnen, J.-P., De Winde, J. H. & Ruijssenaars, H. J. Engineering Pseudomonas putida S12 for efficient utilization of D-xylose and L-arabinose. Appl. Environ. Microbiol. 74, 50315037 (2008)
  15. Feist, A. M. et al. A genome-scale metabolic reconstruction for Escherichia coli K-12 MG1655 that accounts for 1260 ORFs and thermodynamic information. Mol. Syst. Biol. 3, 121 (2007)
  16. Neidhardt, F. & Ingraham, J. Escherichia Coli and Salmonella Typhimurium: Cellular and Molecular Biology Vol. 1, 1316 (American Society for Microbiology, 1987)
  17. Nam, H. et al. Network context and selection in the evolution to enzyme specificity. Science 337, 11011104 (2012)
  18. Aguilar, C. et al. Genetic changes during a laboratory adaptive evolution process that allowed fast growth in glucose to an Escherichia coli strain lacking the major glucose transport system. BMC Genomics 13, 385 (2012)
  19. Kim, J., Kershner, J. P., Novikov, Y., Shoemaker, R. K. & Copley, S. D. Three serendipitous pathways in E. coli can bypass a block in pyridoxal-5′-phosphate synthesis. Mol. Syst. Biol. 6, 436 (2010)
  20. Martin, O. C. & Wagner, A. Multifunctionality and robustness trade-offs in model genetic circuits. Biophys. J. 94, 29272937 (2008)
  21. Ancel, L. W. & Fontana, W. Plasticity, evolvability, and modularity in RNA. J. Exp. Zool. 288, 242283 (2000)
  22. Amitai, G., Gupta, R. D. & Tawfik, D. S. Latent evolutionary potentials under the neutral mutational drift of an enzyme. HFSP J. 1, 6778 (2007)
  23. Isalan, M. et al. Evolvability and hierarchy in rewired bacterial gene networks. Nature 452, 840845 (2008)
  24. Price, N. D., Reed, J. L., Palsson, B. & Ø Genome-scale models of microbial cells: evaluating the consequences of constraints. Nature Rev. Microbiol. 2, 886897 (2004)
  25. Kauffman, K. J., Prakash, P. & Edwards, J. S. Advances in flux balance analysis. Curr. Opin. Biotechnol. 14, 491496 (2003)
  26. Goto, S., Nishioka, T. & Kanehisa, M. LIGAND: chemical database of enzyme reactions. Nucleic Acids Res. 28, 380382 (2000)
  27. Goto, S., Okuno, Y., Hattori, M., Nishioka, T. & Kanehisa, M. LIGAND: database of chemical compounds and reactions in biological pathways. Nucleic Acids Res. 30, 402404 (2002)
  28. Kanehisa, M. & Goto, S. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 28, 2730 (2000)
  29. Kanehisa, M., Goto, S., Furumichi, M., Tanabe, M. & Hirakawa, M. KEGG for representation and analysis of molecular networks involving diseases and drugs. Nucleic Acids Res. 38, D355D360 (2010)
  30. Barve, A., Rodrigues, J. F. M. & Wagner, A. Superessential reactions in metabolic networks. Proc. Natl Acad. Sci. USA 109, E1121E1130 (2012)
  31. Matias Rodrigues, J. F. & Wagner, A. Evolutionary plasticity and innovations in complex metabolic reaction networks. PLoS Comput. Biol. 5, e1000613 (2009)
  32. Henry, C. S. et al. High-throughput generation, optimization and analysis of genome-scale metabolic models. Nature Biotechnol. 28, 977982 (2010)
  33. Matias Rodrigues, J. F. & Wagner, A. Genotype networks, innovation, and robustness in sulfur metabolism. BMC Syst. Biol. 5, 39 (2011)
  34. Binder, K. & Heerman, D. W. Monte Carlo Simulation in Statistical Physics (Springer, 2010)
  35. Koskiniemi, S., Sun, S., Berg, O. G. & Andersson, D. I. Selection-driven gene loss in bacteria. PLoS Genet. 8, e1002787 (2012)
  36. Ochman, H., Elwyn, S. & Moran, N. A. Calibrating bacterial evolution. Proc. Natl Acad. Sci. USA 96, 1263812643 (1999)
  37. Wagner, A. & Fell, D. A. The small world inside large metabolic networks. Proc. R. Soc. Lond. B 268, 18031810 (2001)
  38. Ma, H.-W. & Zeng, A.-P. The connectivity structure, giant strong component and centrality of metabolic networks. Bioinformatics 19, 14231430 (2003)
  39. Moore, E. in Proc. Internat. Symp. Theory Switching, Ann. Comput. Lab. Harvard Univ. 285–292 (Harvard Univ. Press, 1959)
  40. Sokal, R. R. & Michener, C. D. A statistical method for evaluating systematic relationships. Univ. Kansas Sci. Bull. 28, 14091438 (1958)

Download references

Author information


  1. Institute of Evolutionary Biology and Environmental Sciences, Building Y27, University of Zurich, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland

    • Aditya Barve &
    • Andreas Wagner
  2. The Swiss Institute of Bioinformatics, Bioinformatics, Quartier Sorge, Bâtiment Genopode, 1015 Lausanne, Switzerland

    • Aditya Barve &
    • Andreas Wagner
  3. The Santa Fe Institute, 1399 Hyde Park Road, Santa Fe, New Mexico 87501, USA

    • Andreas Wagner


A.B. and A.W. designed the research; A.B. performed the research; A.B. contributed new reagents and analytic tools; A.B. and A.W. analysed the data; and A.B. and A.W. wrote the paper.

Competing financial interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to:

Author details

Supplementary information

PDF files

  1. Supplementary Information (2 MB)

    This file contains Supplementary Results, Supplementary Figures 1-12, Supplementary Tables 1-4 and additional references.

Additional data