Constraints Tom JM Van Dooren
[email protected] iEES Paris, team VPA, phenotypic variability and adaptation www.tomvandooren.eu
Some types of evolutionary change seem to have happened repeatedly and with even probabilities among alternative routes
Palmer, A. R. (2004). Symmetry breaking and the evolution of development. Science, 306: 828-833.
Phylogenetic patterns Are genes leaders or followers Conversion of mechanisms
Heteranthera
Schwander and Leimar 2011
Artificial selection: No evolutionary constraints
Atlantic silverside Conover et al. 2009
~ natural selection
Cold adaptation
Trophy hunting
Regulatory mechanism with developmental switches seem to evolve easily, at least the response to the cue provoking the switching Selection experiments on life history traits and morphology usually produce a response
Are there examples wit little or no evolutionary change over long periods?
Evolutionary Constraints
Genetic Variation and Short-Term Evolution Trait two Evolutionarily Convergence Stable Traits
Response
Selective advantage
Selection
Genetic Variation Initial Population
Selective disadvantage
Trait one
Allen, C., Beldade, P., Zwaan, B.J., Brakefield, P.M. (2008) Differences in the selection response of serially repeated color pattern characters: Standing variation, development, and evolution. BMC Evolutionary Biology 2008, 8:94
No response in the short term The breeders equation for selection response R = Gb Two possibilities: Some variation cannot be produced (genetic variance - covariance G is degenerate) (Stabilizing) selection prevents change (selection gradient b = 0)
(consensus paper Maynard Smith et al. 1985)
Evolutionary Constraints No evolutionary response in the short/long term – why?
Mechanisms causing constraints Trade-offs – coupling
G in the breeders equation changes across generations Often assumed that this change is negligible – wrong?
What are causes of constraints in the more long term?
Mechanical/Physical constraints produce allometric patterns
Any organism has to obey the laws of physics and chemistry
• Gravity pulls everything down Meganeura moryi Gigantic proto-Odonata because of different composition atmosphere during Carboniferous (Dudley 1998)
• E.g. limits on body size in organisms that have access to oxygen through trachea
Ecological Constraints
b(E)
Selection depends on the environmental state High (Unavoidable) Cost of Reproduction when 1) Carrying eggs 2) Predators are present 3) Visibility is high • The cost is constrained • It is so in specified environments Daphnia pulex
Levels of organization
Environment
Genes
Developmental Constraints
Phenotype
Performance
Ecological Constraints
Physical Constraints
Fitness
Historical or Phylogenetic Constraints
Some traits evolved already in the past and not recently
Organisms resemble their ancestors
Species are not independent samples
Problem of generalization: contingency on actual species traits
Primates cannot occupy all herbivore niches
Muller et al. 2011
Waved albatross Phoebastria irrorata All Procellariiformes lay a single egg per clutch
Other categories Physiological constraints : The properties of physiologies are not allowing to perform certain tasks or produce some phenotypes
Genetic constraints : genotypic variation cannot produce some phenotypes
Tumbaloflesicodelicomicoso Antonovics, J., & van Tienderen, P. H. (1991). Ontoecogenophyloconstraints? The chaos of constraint terminology. Trends in Ecology & Evolution, 6(5), 166-168.
Classifications of Constraints: What a Mess Physical Constraints
Genetic Constraints
Phylogenetic Constraints
Physiological
Ecological
Constraints (Roff 1992)
Trade- Offs (Roff 2002)
No response – the short term perspective does not always relate easily to these categories of constraints
Species
Variation
Selection
Albatross
Absent
Foraging efficiency
Daphnia
Absent (Cost)
Reproduction
Dragonfly
Aquatic larva known from Permian and …
Performance
Large Meganeuridae exist from the Permian with lower oxygen There are other large insects still present (stick insects) Maybe they were just selected for large size
No response – the short term perspective does not always relate easily to the categories of constraints
Species
Variation
Selection
Albatross
Absent
Foraging mode
Daphnia
Absent
Reproduction
Dragonfly
Aquatic larva known from Permian
Performance
Aim: Combine quantitative genetic and long-term perspectives Estimate both variation and selection on larger timescales
Phylogenetic patterns One often models evolution along a tree assuming responses R = Gb Species traits will change or not, but also the genetic variance-covariance G
Steppan et al. 2002
Phylogenetic patterns One often models evolution along a tree assuming R = Gb Species traits will change or not, but also the genetic variance-covariance G
We are aware that phylogenetic patterns are there, and need models of how they appeared! → Try to reconstruct the emergence of the constraint → Try to reconstruct the patterns of selection
http://cran.r-project.org/web/views/Phylogenetics.html
Reaction norms for egg size ~ maternal size
Austrolebias monstrosus
Austrolebias charrua
What is the phenotype exactly which is constrained?
Switches Morphologies - size When multiple traits are considered, beta and G can be non-zero and non-degenerate and still pose a constraint
Eco-evo-devo Does the exact nature and determination/development of the trait matter for the detection of patterns of natural selection and constraint?
Pedals got shot ~ the bipedal goat from Utrecht
Internal selection: interactions between developmental modules constrain evolution
Galis et al. 2006
Phylotypic stage
Developmental hourglass Prud’homme and Gompel 2010
Germband Aminoserosa Head region minus gnathal segments Fig. 1. Extended (a) and segmented (b) germband stages in Drosophila. The germband (blue) refers to the part of the embryo that will give rise to the metameric regions: gnathal segments of the head region (Md, mandible; Mx, maxilla; Lb, labium), thoracic segments (T1–3) and abdominal segments (A1–8). The amnioserosa (red) is an extra-embryonic membrane. The extended germband stage starts ~.6.5 h after fertilization and the segmented germband stage ends at ~10.5 h after fertilization.
Von Dassow et al. (2000)
Are phenotypes constrained because they are robust, perturbations are dampened? Not in this case. WT
Null
Df(2)DE
NE2 Null (wg−/−), reduced [Df(2)DE] and partial (NE2) function mutations of the wg gene lead to abnormalities in the larval ectoderm. Expression of wg in the ectoderm (A–D), and cuticular pattern in the ventral (E–H) and dorsal (I–L) larval epidermis (W.T. denotes wild type). In Df(2)DE mutants, wg expression is reduced, and in NE2 mutants, wg transport is hampered (reproduced, with permission, from Ref. [27]).
Galis et al. 2002
Are phenotypes constrained because they are robust? Not in this case.
Hypomorphic Wg-1 mutant showing a failure in the development of antennae, wings, halteres and thorax
Galis et al. 2002
Germband Aminoserosa Head region minus gnathal segments Fig. 1. Extended (a) and segmented (b) germband stages in Drosophila. The germband (blue) refers to the part of the embryo that will give rise to the metameric regions: gnathal segments of the head region (Md, mandible; Mx, maxilla; Lb, labium), thoracic segments (T1–3) and abdominal segments (A1–8). The amnioserosa (red) is an extra-embryonic membrane. The extended germband stage starts ~.6.5 h after fertilization and the segmented germband stage ends at ~10.5 h after fertilization.
Internal selection due to interactions causing effects on many phenotypes
Development as a process with input, control and output
Internal selection is selection due to shapes of multivariate genotype-phenotype maps with interactions between traits I will show the genotype phenotype map is both • A component of variation • A component of the selection gradient
Assume smooth genotype-phenotype maps
Apparent phenotype Y - Underlying trait X
Barbara Stadler has worked out the ingredients to do this analysis for discrete genotype spaces
Apparent phenotype Y - Underlying trait X
Phenotypic trait vector Y underlying traits X of a haplotype
Y depends on X
Y X
Apparent phenotype Y - Underlying trait X
Phenotypic trait vector Y underlying traits X Y
X
Apparent phenotype Y - Underlying trait X
allelic traits organismal traits fitness
devo
eco
evo
The map Y(X) is locally approximately linear
Phenotypic trait vector Y underlying traits X of a haplotype Y
.
. . .
X
Invasion fitness
fitness of the phenotype of a mutant Y in a population with phenotype Z of the resident allele (genotype)
r (Y , Z ) fitness of a mutant X' in a population of alleles with trait X
X ' , X = r Y X ', Y X
Invasion fitness gradient
' ( X ) =
r (Y X ', Y X ) X ' X '= X
' ( X ) = Y ( X )' r Y X devo
' r ( Z ) =
eco
fitness gradient = phenotypic effects of allele × ecological effects of phenotype
r (Y , Z ) Y Y =Z
Evolutionary Dynamics
Gradual directional evolution by new mutations occurring
d 1 X (t ) = G X (t )Y ( X (t ))' r Y X (t ) dt 2 scaling for devo available variation
eco
d 1 Y (t ) = Y ( X (t ))T G X (t )Y ( X (t ))' r Y X (t ) dt 2 d 1 Y (t ) = GY (t )' r Y X (t ) dt 2
Evolutionary Dynamics
d 1 X (t ) = G X (t )Y ( X (t ))' r Y X (t ) dt 2 selection
d 1 Y (t ) = Y ( X (t ))T G X (t )Y ( X (t ))' r Y X (t ) dt 2 variation
Evolutionary Dynamics
d 1 X (t ) = G X (t )Y ( X (t ))' r Y X (t ) dt 2 selection
d 1 Y (t ) = Y ( X (t ))T G X (t )Y ( X (t ))' r Y X (t ) dt 2 variation
Evolutionarily Stable Configuration
• evolves in the same way in any environment, independent of ecology • evolution driven by internal coherence and system performance • performance is for a proper function (raison d'être)
Example: iguanians use their tongue as a prehensile organ (Wagner and Schwenk 2000) One type of internal selection
Evolutionarily Stable Configuration Y(X*) = 0 for all loci involved
performance is for a proper function (raison d'être) Y is one-dimensional = e.g. capture rate 'r(y) > 0
performance y
x*
tongue traits
Can this framework describe and model internal selection due to interactions between developing traits and early ecological selection on some of them, combined with selection later during life on others?
Annual killifish - Austrolebias
blastula – dispersed cells Diapause I somites
Diapause 2 head formation
Strategy determination is relatively accessible, observable
seemingly ready to hatch Diapause 3
apparently complete but small
An example of how constraints are treated in classical evolutionary ecology and Adaptive Dynamics
Onto the boundary of a feasibility set Fitness often increases in this direction: The more of survival and reproduction, the better, the idea of a feasibility set is an extra assumption that can help to generate predictions
LH Trait two
Feasibility Set
LH Trait one Rueffler et al. 2004
Options - Fitness contours
Begon et al. 2005
Classification of Environments Sensitive to growth InSensitive to growth
Begon et al. 2005
Begon et al. 2005
Some antidote against all this smoothness
Genotype network spaces
The viewpoint of the underlying level is essential to understand constraints
“All complex macroscopic traits comprise microscopic, submicroscopic and molecular traits, down to the level of DNA. Likewise, DNA change can percolate all the way up to macroscopic traits. Although an understanding of the full complexity of this hierarchical organization is beyond current means, important systems can be studied that are necessary to form complex traits and changes therein.”
Wagner 2011
The viewpoint of the underlying level is essential to understand constraints
Example
Variation
Selection
Gene regulation networks
X
X
Metabolic networks
X
X
Macromolecules
X
X
These models provide simple genotype-phenotype maps and ecologies used to investigate constraints
Wagner 2011
No response – Genotype networks Each colour is a phenotype Lines represent possible mutational changes
Selection on robustness
To study effects of: Genotype space structure
G-P mapping
Wagner 2011
Genotype Networks Novelties Connected genotype spaces with different accessible phenotypes promote novelties, Below are three opposites to that:
Genotype networks & plasticity: weakly developed framework
Simplified plasticity: Alternative phenotype If present, can mutate in a new constitutive phenotype
Espinosa-Soto et al. (2011)
Genotype networks and plasticity (a) Populations find a novel genotype network faster when plasticity is allowed. The symbol t*, plast refers to the number of generations that a population of circuits needs to discover a specific genotype network when we allow plasticity. The symbol t*, control refers to the same number, but for populations without plasticity. (b) Plasticity slows the accumulation of individuals in the new genotype network. The symbol t0.25,plast stands for the number of generations that a population in which we allow plasticity needs to have at least 25 percent of its circuits in the new genotype network (after its discovery by a single individual); t0.25,control corresponds to the same number but without plastic phenotypes.
Release from constraints in genotype networks: Exploration of genotype network spaces by selection and drift Plasticity and a new environment revealing cryptic variation Evolutionary novelty facilitated by plasticity Plasticity prevents accumulation on a single phenotype
Many open questions: constraints can arise from specialization?
Heteroecious aphids (Moran 1988)
Galis et al. 2014
Specialists – species selection and constraints on rates of diversification and extinction
New approaches: Design a genetic – regulatory mechanism Synthetic biology Give it the properties that allow theory to be tested / falsified Use theoretical concepts to structure expectations and predictions Experimental evolution Examples: Sander Tans lab Amsterdam – “Optimality in evolution: new insights from synthetic biology” http://tansgroup.amolf.nl/publications.html
sketchy set of references Charnov, E. L. 1993. Life History Invariants. Oxford University Press Conover, D.O., S.B. Munch, and S.A. Arnott (2009) Reversal of evolutionary downsizing caused by selective harvest of large fish. Proceedings of the Royal Society of London. Series B: Biological Sciences 276:2015-2020. Dudley, R. 1998. Atmospheric oxygen, giant Paleozoic insects and the evolution o aerial locomotor performance. Journal of Experimental Biology 201: 1043-1050. Espinosa-Soto, C. Martin, O.C., Wagner, A. (2011) Phenotypic plasticity can facilitate adaptive evolution in gene regulatory circuits. BMC Evolutionary Biology 11:5, doi:10.1186/1471-2148-11-5 Galis, F. , T.J.M. Van Dooren and J.A.J. Metz (2002). Conservation of the segmented germband stage: robustness or pleiotropy? Trends Genet. 18 (10), 504-509. Galis F., T.J.M. Van Dooren, Feuth, H., Ruinard, S., Witkam, A., Steigenga, M.J., Metz, J.A.J., Wijnaendts, L.C.D. (2006). Extreme selection against homeotic transformations of cervical vertebrae in humans.Evolution 60 (12):2643-2654. Galis F, Carrier DR, van Alphen J, van der Mije SD, Van Dooren TJM, Metz JAJ & CM Ten Broek (2014) Fast running restricts evolutionary change of the vertebral column in mammals. Proc. Natl. Acad. Sci. U.S.A. 111: 11401-11406. Maynard Smith, J., R. Burian, S. Kaufman, P. Alberch, J. Campbell et al., 1985. Developmental constraints and evolution. Q. Rev. Biol. 60: 265–287. Muller et al. 2011. Phylogenetic constraints on digesta separation: Variation in fluid throughput in the digestive tract in mammalian herbivores. Comparative biochemistry and physiology. Part A, Molecular & integrative physiology. 06/2011; Nee S et al. The illusion of invariant quantities in life histories. Science. 2005 Aug 19; 309(5738):1236-9 Roff, D.A. 1992. The Evolution of Life Histories: Theory and Analysis. Chapman and Hall, New York. Roff, D.A. 2002. Life History Evolution. Sinauer Associates, Sunderland, MA. G. von Dassow, E. Meir, E. M. Munro, and G. M. Odell (2000) The segment polarity network is a robust developmental module. Nature 406: 188-92. Wagner 2011. Genotype networks shed light on evolutionary constraints. Trends in Ecology & Evolution. doi:10.1016/j.tree.2011.07.001 Wagner, G. P. and K. Schwenk (2000) Evolutionarily Stable Configurations: functional integration and the evolution of phenotypic stability. Evolutionary Biology 31:155-217.
