Limits...
Regulatory mechanisms link phenotypic plasticity to evolvability.

van Gestel J, Weissing FJ - Sci Rep (2016)

Bottom Line: Using individual-based simulations, we compare the RN and GRN approach and find a number of striking differences.Most importantly, the GRN model results in a much higher diversity of responsive strategies than the RN model.The regulatory mechanisms that control plasticity therefore critically link phenotypic plasticity to the adaptive potential of biological populations.

View Article: PubMed Central - PubMed

Affiliation: Groningen Institute for Evolutionary Life Sciences, University of Groningen, P.O. Box 11103, Groningen 9700 CC, The Netherlands.

ABSTRACT
Organisms have a remarkable capacity to respond to environmental change. They can either respond directly, by means of phenotypic plasticity, or they can slowly adapt through evolution. Yet, how phenotypic plasticity links to evolutionary adaptability is largely unknown. Current studies of plasticity tend to adopt a phenomenological reaction norm (RN) approach, which neglects the mechanisms underlying plasticity. Focusing on a concrete question - the optimal timing of bacterial sporulation - we here also consider a mechanistic approach, the evolution of a gene regulatory network (GRN) underlying plasticity. Using individual-based simulations, we compare the RN and GRN approach and find a number of striking differences. Most importantly, the GRN model results in a much higher diversity of responsive strategies than the RN model. We show that each of the evolved strategies is pre-adapted to a unique set of unseen environmental conditions. The regulatory mechanisms that control plasticity therefore critically link phenotypic plasticity to the adaptive potential of biological populations.

No MeSH data available.


Related in: MedlinePlus

Spore production of GRNs in randomly-generated novel environments.The twenty most productive genotypes of the RN model (grey circles) and GRN model (black circles) were exposed to 250 randomly-generated novel environments (Supplementary Table S2 and Supplementary Data S1). For each environment ten replicate colonies were grown per genotype and the average number of spores at the end of colony growth was used for the cluster analysis see Material and Methods). The colours indicate the relative spore production of genotypes in each novel environment: red = relative low spore production, green = relative high spore production. Histrograms on the right show the distribution of relative spore production for the genotypes. Graph on the bottom shows range of absolute spore production in RN model (light grey area) and GRN model (dark grey area) over all 250 environments. The two bars at the bottom compare genotypes from the RN and GRN model directly, by showing where the most productive genotype comes from (grey = a genotype from RN model, black = a genotype from GRN model) and which set of genotypes produce most spores on average (grey = genotypes from RN model, black = genotypes from GRN model).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4834480&req=5

f9: Spore production of GRNs in randomly-generated novel environments.The twenty most productive genotypes of the RN model (grey circles) and GRN model (black circles) were exposed to 250 randomly-generated novel environments (Supplementary Table S2 and Supplementary Data S1). For each environment ten replicate colonies were grown per genotype and the average number of spores at the end of colony growth was used for the cluster analysis see Material and Methods). The colours indicate the relative spore production of genotypes in each novel environment: red = relative low spore production, green = relative high spore production. Histrograms on the right show the distribution of relative spore production for the genotypes. Graph on the bottom shows range of absolute spore production in RN model (light grey area) and GRN model (dark grey area) over all 250 environments. The two bars at the bottom compare genotypes from the RN and GRN model directly, by showing where the most productive genotype comes from (grey = a genotype from RN model, black = a genotype from GRN model) and which set of genotypes produce most spores on average (grey = genotypes from RN model, black = genotypes from GRN model).

Mentions: Next, we exposed the twenty most productive RN and GRN genotypes to 250 novel environments, which were generated by randomly varying seven model parameters: (1) nutrient diffusion rate, (2) signal diffusion rate, (3) signal degradation rate, (4) nutrient consumption rate, (5) probability of cell division, (6) duration of sporulation and (7) energy consumption during sporulation (Supplementary Table S2 and Supplementary Data S1). Although many of these parameters are not strictly environmental, since they can also be influenced by cellular or molecular factors (e.g. the signal diffusion rate depends on the molecular weight of the signal and the ambient temperature), changes in these parameters do affect the environment to which cells are exposed and the optimal timing of sporulation. As such, they confront cells with unseen environmental conditions. The parameters are varied such that colony growth would not exceed the surface area. In each environment, we determined the relative and absolute spore production of a genotype by growing ten replicate colonies. To evaluate the responses of the evolved genotypes over the 250 environmental conditions, we performed a cluster analysis (see Material and Methods). Genotypes were clustered with respect to their relative spore production over the 250 environments and environments were clustered with respect to the absolute spore production of the 40 genotypes. Genotypes that appear close in the cluster analysis have approximately the same relative spore production – i.e. relative fitness with respect to the other genotypes – in all novel environments. Environments that cluster together have a similar effect on the absolute spore production of all genotypes. Figure 9 shows that the relative spore production of each genotype varies strongly between the novel environments (see histograms on the right of the cluster diagram in Fig. 9). Moreover, the differences in spore production between genotypes from the GRN model are much more pronounced than the differences between genotypes from the RN model. Thus, the higher diversity of responsive strategies in the GRN model (Figs 4 and 6) is translated to a higher diversity in spore production among the tested novel environments (Figs 8 and 9). Even though there are many novel environments for which genotypes from the RN model on average perform better than the genotypes from the GRN model, the best performing genotypes nearly always come from the GRN model. In the GRN model, each genotype has a distinct profile along the 250 novel environments. In other words, each responsive strategy is pre-adapted to unique set of environmental conditions. Some genotypes perform relatively well under most novel environments, but never have the highest relative spore production (e.g. genotype 10 from the GRN model). Others perform very well in some environments, but badly in others (e.g. genotype 1 from the GRN model). Importantly, none of the genotypes performs best in all environmental conditions. The diversity in responsive strategy is therefore important for the capacity of a population to cope with many potential future environments.


Regulatory mechanisms link phenotypic plasticity to evolvability.

van Gestel J, Weissing FJ - Sci Rep (2016)

Spore production of GRNs in randomly-generated novel environments.The twenty most productive genotypes of the RN model (grey circles) and GRN model (black circles) were exposed to 250 randomly-generated novel environments (Supplementary Table S2 and Supplementary Data S1). For each environment ten replicate colonies were grown per genotype and the average number of spores at the end of colony growth was used for the cluster analysis see Material and Methods). The colours indicate the relative spore production of genotypes in each novel environment: red = relative low spore production, green = relative high spore production. Histrograms on the right show the distribution of relative spore production for the genotypes. Graph on the bottom shows range of absolute spore production in RN model (light grey area) and GRN model (dark grey area) over all 250 environments. The two bars at the bottom compare genotypes from the RN and GRN model directly, by showing where the most productive genotype comes from (grey = a genotype from RN model, black = a genotype from GRN model) and which set of genotypes produce most spores on average (grey = genotypes from RN model, black = genotypes from GRN model).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4834480&req=5

f9: Spore production of GRNs in randomly-generated novel environments.The twenty most productive genotypes of the RN model (grey circles) and GRN model (black circles) were exposed to 250 randomly-generated novel environments (Supplementary Table S2 and Supplementary Data S1). For each environment ten replicate colonies were grown per genotype and the average number of spores at the end of colony growth was used for the cluster analysis see Material and Methods). The colours indicate the relative spore production of genotypes in each novel environment: red = relative low spore production, green = relative high spore production. Histrograms on the right show the distribution of relative spore production for the genotypes. Graph on the bottom shows range of absolute spore production in RN model (light grey area) and GRN model (dark grey area) over all 250 environments. The two bars at the bottom compare genotypes from the RN and GRN model directly, by showing where the most productive genotype comes from (grey = a genotype from RN model, black = a genotype from GRN model) and which set of genotypes produce most spores on average (grey = genotypes from RN model, black = genotypes from GRN model).
Mentions: Next, we exposed the twenty most productive RN and GRN genotypes to 250 novel environments, which were generated by randomly varying seven model parameters: (1) nutrient diffusion rate, (2) signal diffusion rate, (3) signal degradation rate, (4) nutrient consumption rate, (5) probability of cell division, (6) duration of sporulation and (7) energy consumption during sporulation (Supplementary Table S2 and Supplementary Data S1). Although many of these parameters are not strictly environmental, since they can also be influenced by cellular or molecular factors (e.g. the signal diffusion rate depends on the molecular weight of the signal and the ambient temperature), changes in these parameters do affect the environment to which cells are exposed and the optimal timing of sporulation. As such, they confront cells with unseen environmental conditions. The parameters are varied such that colony growth would not exceed the surface area. In each environment, we determined the relative and absolute spore production of a genotype by growing ten replicate colonies. To evaluate the responses of the evolved genotypes over the 250 environmental conditions, we performed a cluster analysis (see Material and Methods). Genotypes were clustered with respect to their relative spore production over the 250 environments and environments were clustered with respect to the absolute spore production of the 40 genotypes. Genotypes that appear close in the cluster analysis have approximately the same relative spore production – i.e. relative fitness with respect to the other genotypes – in all novel environments. Environments that cluster together have a similar effect on the absolute spore production of all genotypes. Figure 9 shows that the relative spore production of each genotype varies strongly between the novel environments (see histograms on the right of the cluster diagram in Fig. 9). Moreover, the differences in spore production between genotypes from the GRN model are much more pronounced than the differences between genotypes from the RN model. Thus, the higher diversity of responsive strategies in the GRN model (Figs 4 and 6) is translated to a higher diversity in spore production among the tested novel environments (Figs 8 and 9). Even though there are many novel environments for which genotypes from the RN model on average perform better than the genotypes from the GRN model, the best performing genotypes nearly always come from the GRN model. In the GRN model, each genotype has a distinct profile along the 250 novel environments. In other words, each responsive strategy is pre-adapted to unique set of environmental conditions. Some genotypes perform relatively well under most novel environments, but never have the highest relative spore production (e.g. genotype 10 from the GRN model). Others perform very well in some environments, but badly in others (e.g. genotype 1 from the GRN model). Importantly, none of the genotypes performs best in all environmental conditions. The diversity in responsive strategy is therefore important for the capacity of a population to cope with many potential future environments.

Bottom Line: Using individual-based simulations, we compare the RN and GRN approach and find a number of striking differences.Most importantly, the GRN model results in a much higher diversity of responsive strategies than the RN model.The regulatory mechanisms that control plasticity therefore critically link phenotypic plasticity to the adaptive potential of biological populations.

View Article: PubMed Central - PubMed

Affiliation: Groningen Institute for Evolutionary Life Sciences, University of Groningen, P.O. Box 11103, Groningen 9700 CC, The Netherlands.

ABSTRACT
Organisms have a remarkable capacity to respond to environmental change. They can either respond directly, by means of phenotypic plasticity, or they can slowly adapt through evolution. Yet, how phenotypic plasticity links to evolutionary adaptability is largely unknown. Current studies of plasticity tend to adopt a phenomenological reaction norm (RN) approach, which neglects the mechanisms underlying plasticity. Focusing on a concrete question - the optimal timing of bacterial sporulation - we here also consider a mechanistic approach, the evolution of a gene regulatory network (GRN) underlying plasticity. Using individual-based simulations, we compare the RN and GRN approach and find a number of striking differences. Most importantly, the GRN model results in a much higher diversity of responsive strategies than the RN model. We show that each of the evolved strategies is pre-adapted to a unique set of unseen environmental conditions. The regulatory mechanisms that control plasticity therefore critically link phenotypic plasticity to the adaptive potential of biological populations.

No MeSH data available.


Related in: MedlinePlus