Kassen 3: Selection and Adaptation 2, Novelty and Cooperation

Two models of mutation - driven adaptation

• Fisher’s geometric model

• Gillespie’s mutational landscape model

Assumptions of models of mutation-driven adaptation

All variation is due to mutation

The ancestor is mal-adapted

SSWM means beneficial mutations occur rarely, and when they do, sweep instantaneously to fixation

Fisher’s geometric model of mutation-driven adaptation

Arrows represent mutational events

The circle represents beneficial phenotypes

The origin is a fitness peak

Gillespie’s mutational landscape model

Does not consider anything that is less beneficial than the ancestral genotype. Most mutations are slightly less (middle of distribution).

Focus on right tail, which looks exponential.

Distribution of Fitness Effects (DFE) among beneficial mutations available to selection

Should contain many small, few great-effect mutations

Most distributions should be L-shaped

Theory of exceedance

A beneficial mutation, has, by definition, a fitness greater than the WT.

Exceedance tells us how much greater

Exceedance events occur in L-shaped, exponential tail of a DFE

The DFE among beneficial mutations contributing to adaptation

Small effect mutations lost to drift

Modestly beneficial mutations are outcompeted by clonal interference (When NU_b »1)

Large effects are rare

So this DFE is hump-shaped, with a mean larger than the DFE (available)

Rates of evolution over time

Decline

As a genotype climbs fitness peaks, there are fewer beneficial mutations

Rates of substitution decrease but don’t stop

Does molecular evolution continue if adaptation slows? And why?

No; mutations always happen, and the environment keeps changing, so populations continuously adapt

How to find targets of selection/ genes contributing to adaptation

Experimental evolution and genome resequencing can serve as a genetic screen

Mutations recovered are different from conventional screens because they arose naturally and have survived drift, clonal interference, and selection

We can therefore let the evolutionary process tell us something about the mutations that contribute to adaptation

What type of mutations are common in adaptation

Loss of function (LOF)

Examples of LOF mutations in adaptation

Domestication mutations eliminate functions no longer needed in the selective environment

• I.e. Pseudomonas lose flagellae

Deregulation mutations lead to constitutive expression of previously inducible functions

Gene amplifications

Can increase the production of growth-limited enzymes

Often costly unless maintained through selection

Gene amplification example

Selection for 200 generations in sulfate limited conditions in yeast leads to amplification of SUL1

Mutations to global gene regulation

Common

Important to note the hierarchical nature of regulation

Novelty

The quality of being new

Can refer to a new gene, new trait, new ecology

Innovation

The act of introducing a novelty

The paradox of novelty

Natural selection works with materials available to it

But changing something that already works usually makes life worse

So how can anything new ever evolve?

Solution - evolution is a “tinkerer” that uses old materials in new ways

Evolutionary novelty

A new ecological function underlain by a qualitatively distinct genetic architecture

Rapid adaptation vs evolutionary novely

Rapid adaptation often involves a single change

• LOF mutations

• Amplifications

• Gene regulation

Novelty involves many more genetic changes

• New genes (built from new mutations)

• New gene arrangements (“structural variation” = a form of mutation)

• New ways of regulating gene expression

Consequently, novelty is rare

How does selection generate a novel gene?

A population in a new environment can barely make a living and uses the enzymes and proteins available

• Through amplification, it creates more of what it already has

Once the population increases to sufficient size, actual novelty can arise

Four steps of novelty evolving

1. Potentiation

2. Exaptation

3. Amplification

4. Divergence

a.k.a. the PEAD model

Potentiation

The evolution of genetic background that affords a lineage access to mutations linked to the novel function that that would otherwise be inaccessible

Exaptation (co-option)

A pre-existing gene function is pressed into service to perform a new role

A trait that takes on a function that it was previously not doing

Promiscuous proteins are exaptations

• They have one main function, but many side functions as well

“Sloppy” protein function can allow an organism to get by in a new environment to which it is not well adapted

Amplification

Genetic changes increasing gene copy number (often duplications) that also increase the quantity of enzyme produced

Gene duplication

Creates two copies of a gene, one of which can continue performing the ancestral function, and the other of which can take on a new function

Paralogs

Genes descended from duplication

Divergence

Having more copies of genes allows them to diverge through selection

PEAD example

Novel trait in E. coli: aerobic citrate metabolism

Citrate added to medium as a chelating agent

Not being able to metabolize citrate aerobically is a diagnostic trait of E. coli

After 33,500 generations, a marked increase in cultural turbidity was caused by citrate metabolism

Potentiation in E. coli example

Clones isolated just before the evolution of cit+ were much more likely to generate cit+ mutants than the ancestor

Exaptation in E. coli example

Machinery for metabolizing citrate (cit promoter plus citT) was present, just not in aerobic conditions

Amplification in E. coli example

Duplication of citT inserted after the rnk promoter, allowing aerobic expression

Divergence in E. coli example

Growth on citrate increased dramatically through conventional selection

Cooperation

Two or more individuals receive a benefit from joint action

• i.e. mobbing in birds, slime molds, parental care

Can be undermined by cheating

Cheating and cooperation

Cooperation is costly, and free riders prosper in the short term by getting benefits without the cost

• i.e. tax evasion

How to get around free riders

Direct the benefits of cooperation to cooperators

• Kin selection

• Reciprocity (game theory)

• Group selection

Levels of selection

Natural selection can operate whenever there is heritable variation among entities

Can be at individuals level, cell level, or group level

Group selection

Selection among groups

Selection within groups favours cheats

Group selection can favour cooperation if cooperation increases fitness

Requirements for group selection

Variation among groups in the frequency among collaborators

A means of “exporting” individuals to other groups

Simpson’s paradox

Producers (P) generate a public good

Non-producers (NP) receive benefits

Paradox: Ps decrease within a group, but can increase globally

Mechanisms behind Simpson’s paradox

NPs increase locally because they get all the benefits of associating with Ps, but none of the costs

NPs decrease globally because the presence of Ps increases the total productivity of the whole group

Effectively, the more Ps in a group, the more that group contributes to the total population

Assumption: the benefits of cooperation are felt by the whole group

Experimental test of Simpson’s paradox

Some E. coli produce Rhl (costly, creates antibiotic resistance)

Growing mixed populations:

• Ps (produce Rhl) decrease within groups

• Subpopulations with higher freq(P) grow faster

• Freq(P) increase globally