Robustness (evolution)

In

environmental, recombinational, or behavioral robustness etc.^[6][7][8] Robustness is achieved through the combination of many genetic and molecular mechanisms and can evolve by either direct or indirect selection. Several model systems have been developed to experimentally study robustness and its evolutionary consequences.

Biological processes at the molecular scale are inherently stochastic.[16] They emerge from a combination of stochastic events that happen given the physico-chemical properties of molecules. For instance, gene expression is intrinsically noisy. This means that two cells in exactly identical regulatory states will exhibit different mRNA contents.^[17][18] The cell population level log-normal distribution of mRNA content^[19] follows directly from the application of the Central Limit Theorem to the multi-step nature of gene expression regulation.^[20]

Environmental robustness

In varying

Genomes mutate by environmental damage and imperfect replication, yet they display remarkable tolerance. This comes from robustness both at many different levels.

Organism mutational robustness

There are many mechanisms that provide genome robustness. For example, genetic redundancy reduces the effect of mutations in any one copy of a multi-copy gene.^[21] Additionally the flux through a metabolic pathway is typically limited by only a few of the steps, meaning that changes in function of many of the enzymes have little effect on fitness.^[22][23] Similarly metabolic networks have multiple alternate pathways to produce many key metabolites.^[24]

Protein mutational robustness

Protein mutation tolerance is the product of two main features: the structure of the

Furthermore, in developmental contexts were timing of gene expression in important for the phenotypic outcome, diverse mechanisms exist to ensure proper gene expression in a timely manner.[36] Poised promoters are transcriptionally inactive promoters that display RNA polymerase II binding, ready for rapid induction.^[38] In addition, because not all transcription factors can bind their target site in compacted heterochromatin, pioneer transcription factors (such as Zld or FoxA) are required to open chromatin and allow the binding of other transcription factors that can rapidly induce gene expression. Open inactive enhancers are call poised enhancers.^[39]

Cell competition is a phenomenon first described in Drosophila^[40] where mosaic Minute mutant cells (affecting ribosomal proteins) in a wild-type background would be eliminated. This phenomenon also happens in the early mouse embryo where cells expressing high levels of Myc actively kill their neighbors displaying low levels of Myc expression. This results in homogeneously high levels of Myc.^[41][42]

Developmental patterning robustness

Patterning mechanisms such as those described by the French flag model can be perturbed at many levels (production and stochasticity of the diffusion of the morphogen, production of the receptor, stochastic of the signaling cascade, etc). Patterning is therefore inherently noisy. Robustness against this noise and genetic perturbation is therefore necessary to ensure proper that cells measure accurately positional information. Studies of the zebrafish neural tube and antero-posterior patternings has shown that noisy signaling leads to imperfect cell differentiation that is later corrected by transdifferentiation, migration or cell death of the misplaced cells.^[43][44][45]

Additionally, the structure (or topology) of signaling pathways has been demonstrated to play an important role in robustness to genetic perturbations.^[46] Self-enhanced degradation has long been an example of robustness in System biology.^[47] Similarly, robustness of dorsoventral patterning in many species emerges from the balanced shuttling-degradation mechanisms involved in BMP signaling.^[48][49][50]

Evolutionary consequences

Since organisms are constantly exposed to genetic and non-genetic perturbations, robustness is important to ensure the stability of

Being robust may even be a favoured at the expense of total fitness as an evolutionarily stable strategy (also called survival of the flattest).^[52] A high but narrow peak of a fitness landscape confers high fitness but low robustness as most mutations lead to massive loss of fitness. High mutation rates may favour population of lower, but broader fitness peaks. More critical biological systems may also have greater selection for robustness as reductions in function are more damaging to fitness.^[53] Mutational robustness is thought to be one driver for theoretical viral quasispecies formation.

Emergent mutational robustness

Mutational robustness has been thought to have a negative impact on evolvability because it reduces the mutational accessibility of distinct heritable phenotypes for a single genotype and reduces selective differences within a genetically diverse population.^{[citation needed]} Counter-intuitively however, it has been hypothesized that phenotypic robustness towards mutations may actually increase the pace of heritable phenotypic adaptation when viewed over longer periods of time.^{[64][65][66][67]}

One hypothesis for how robustness promotes evolvability in asexual populations is that connected networks of fitness-neutral genotypes result in mutational robustness which, while reducing accessibility of new heritable phenotypes over short timescales, over longer time periods, neutral mutation and genetic drift cause the population to spread out over a larger neutral network in genotype space.^[68] This genetic diversity gives the population mutational access to a greater number of distinct heritable phenotypes that can be reached from different points of the neutral network.^{[64][65][67][69][70][71][72]} However, this mechanism may be limited to phenotypes dependent on a single genetic locus; for polygenic traits, genetic diversity in asexual populations does not significantly increase evolvability.^[73]

In the case of proteins, robustness promotes evolvability in the form of an excess free energy of folding.^[74] Since most mutations reduce stability, an excess folding free energy allows toleration of mutations that are beneficial to activity but would otherwise destabilise the protein.

In sexual populations, robustness leads to the accumulation of cryptic genetic variation with high evolutionary potential.^[75][76]

Evolvability may be high when robustness is reversible, with evolutionary capacitance allowing a switch between high robustness in most circumstances and low robustness at times of stress.^[77]

Methods and model systems

There are many systems that have been used to study robustness. In silico models have been used to model

• Distribution of fitness effects
• Evolvability
• Canalization
• Neutral network (evolution)
• Epistasis
• Evolutionary capacitance
• Fitness landscape
• Evolutionary developmental biology

References