Transcript Fitness landscapes: a more biological view

Fitness landscapes and
coevolution: from physics to
biology
Frances Taschuk
March 24, 2008
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
Fitness landscapes
• Peaks
represent local
fitness maxima
• Here: fitness
relates to
feeding
efficiency of
crossbills on
different
conifers
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
Kauffman and Johnsen
• Rugged multipeaked fitness landscapes
• NK model: N=number of traits
K=interacting traits within
species
C=interacting traits in other
species
• In trying to maximize their own fitness,
coevolutionary species change one
another’s fitness landscapes
Tuning
• Parameters can be
adjusted so that
coevolving species reach
steady state (“analogue of
Nash equilibrium”)
• More rugged landscapes
(high K/C ratio) take less
time to reach equilibrium
because more local
maxima
Coevolution
Presence of more interconnected species in model makes it take longer to
reach equilibrium
Avalanches
• When optimized for maximum sustained
fitness, model ecosystem is subject to
changes that propagate throughout
When K is close to optimal value
of 10, log-log plot shows straight
line, suggesting a power law
relationship
Adaptive surfaces in biology
• Idea of adaptive landscape has been
around in biology since 1932
• Originally used to describe genetic finetuning, but has also been used to visualize
larger-scale phenotypic evolution
• Little information is available about people
trying to fit real data to landscapes
Poelwijk et al in Nature (2007)
• '...why, if species have descended from
other species by insensibly fine
gradations, do we not everywhere see
innumerable transitional forms?' (Darwin)
• Looks at molecular function to address
question of evolutionary intermediates
1. Modeling adaptation of bacterial β-lactamase
to the antibiotic cefotaxime
–
–
–
Know resistance phenotype, so can reconstruct
intermediates (5 mutations needed)
5! different possible trajectories
Fitness result of a given mutation depends on order
– “balance of structural and functional benefits
•b, c, d unlikely to
happen
•Found18 paths of type
a
2. Evolution of isopropylmalate dehydrogenase
–
–
–
Studied mutational intermediates affecting binding
of cofactors (NADP less optimal than NAD)
In vitro measurements revealed single-peaked
landscape; mutations not order-dependent
In vivo measurements found a valley on fitness
landscape, making some routes from NADP to NAD
inaccessible
NADP
NAD
3. Evolution of regulatory interactions
- In hormone/receptor model, function could
continue while receptors diverged because
there was a ligand that could activate both
- In the case of lac, new interaction offset
deteriorations - robust network allows
evolvability
Decaestecker et al in Nature (2007)
• Observation of “Red Queen” dynamics
between Daphnia and Pasteuria ramosa
(parasite)
• Reasons to expect coevolution:
– Genetic variation in Daphnia resistance
– Genotype-specific interactions
– Daphnia population genetics change during
parasite epidemics
Methods
• Collected dormant Daphnia eggs and
parasite samples from different depths—
“snapshot in the arms race”
• Exposed Daphnia to parasites from
previous, contemporary, and subsequent
growing seasons
Contemporary parasites are more
infectious
References
Kauffman and Johnsen. 1991. Coevolution to the edge of chaos: coupled
fitness landscapes, poised states, and coevolutionary avalanches. Journal
of Theoretical Biology, 149.
Poelwijk et al. 2007. Empirical fitness landscapes reveal accessible
evolutionary paths. Nature, vol. 445.
Decaestecker et al. 2007. Host-parasite “Red Queen” dynamics archived in
pond sediment. Nature, vol 450.
Also:
Benkman. 2003. Divergent selection drives the adaptive radiation of crossbills.
Evolution.
A simple model of the evolution of simple models of evolution
http://daphnia.cgb.indiana.edu/images/Daphnia_DGC.png