priming

Prev: Spatio-temporal pattern formation as prerequisite for evolving complex (regulatory) individuals Next: Ecosystem based problem solving

TODO List
 * REF Kauffmann .... ???

=Mutational Priming=

In the previous section we have seen the emergence of mutational priming in the evolution of density classifiers, whereby it genomes became structured so that one mutation would allow them to switch between two classifications. //The question we ask here is whether such **mutational priming** can be found in nature?//

To answer this question we take a look at laboratory evolution of yeast as conducted by Ferea et al. (1999). In their experiments yeast was allowed to evolve on a constant glucose poor medium (Note that of course in nature yeast grows on fruit with depleting levels of sugars and therefore will need to shift to other types of metabolism, i.e. the [|diauxic shift]). In this setting Ferea et al. looked at gene expression levels.

After just a few generations there were only few mutations, but already grow had evolved to higher levels. Moreover at least 10% of the genes were significantly differentially expressed compared to the ancestor. Half of those genes where involved in the TCA cycle and were overexpressed. Half were involved in fermentation and were underexpressed. There was also variation between experiments, but this pattern was very consistent. In other words, evolution appears to occur in parrallel to the regulatory diauxic shift, cf attractors in metabolic networks. //The question then becomes: in evolution genotype-phenotype mapping is complex, so what can evolution do for attractors (cf Kaufman paper REF)? Deform the outcome in state space?//

In a paper following that of Ferea et al.(1999), Dunham et al (2002) describe what appears to be mutational priming in yeast in related experiments. They found characteristic genome arrangements in experimental evolution of yeast. They found that many type of mutations (duplications, deletions and cross-chromosomal rearrangements) occur in just a few weeks. Furthermore, these mutations **occur mainly at the same places and at reasonable places** (i.e. in fermentation and respiration genes). Moreover, there were [|transposon]-related sequences at the break-points of the mutations, which is suggestive ....


 * So the bottom line...**
 * short term evolution can change a lot!
 * often the same kind of changes
 * are these constraints?
 * there also appears to be a signal, i.e. break elements: if well located these could help mutation priming to occur .... very suggestive


 * So what can theory help to say about this ..**
 * not to prove this is the case in real systems
 * but to show mechanism
 * without mechanism only have statistics!

//All mutations are equal but some are more equal than others ............//
 * Conjecture:**

Evolution of mutational priming
To further develop the theory about mutational priming and its mechanism Crombach and Hogeweg ([|2007]) studied whether mutational operators in yeast can lead to mutational priming versus that of an artificial system. For this they developed a model as an artificial example, but which had:
 * an organism like coding
 * organism like mutational operators

//Can this then lead to mutational priming?//

Moreover the model allowed:
 * gene duplication and deletion
 * transposon duplication and deletion
 * repeat elements, and breakpoint when deleted
 * double strand break and repairs, with the possibility of error leading to gross-chromosomal rearrangements through repeat elements
 * fast expansion of transposons and dying out (NOT CLEAR)
 * a fluctuating environment (2 states): slow enough to allow adaptation by single gene-duplication
 * space
 * individuals with a long chromosome
 * 20 genes should be expressed normally
 * 20 genes should be either in 1 or 2 copies to cope with environment state 1 or 2
 * fitness and hence selection was then determined by the the fit to the environmental state

Results
Genome evolution in the model follows environmental change whereby fitness drops due to the environmental switch, after which evolution slowly increases fitness through single gene duplications. However later in evolution adaptation appears to occur much faster and occurs in bigger jumps, i.e. by gross-chromosomal rearrangements. In other words there has been a **self-organization of genomes that has lead to the clustering of genes that need to be duplicated!**



Different rates of environmental change also lead to different results. With a slow changing environment most individuals are fit most of the time. With a 10 times faster changing environment:
 * individuals are initially equally distributed over fitness, i.e. can't adapt and there is population-based diversity
 * over time enough selection signal to rearrange genomes and later can adapt to environment, i.e. individual-based diversity
 * now individuals can change fast and track the environment
 * moreover genomes become better organized than for slow change, i.e. many break-point separate duplicated and non-duplicated genes
 * (with even faster environmental change an intermediate solution or generalist evolves because the environment can't be tracked)


 * This means that:**
 * there is a strong selection on **change** (in the environment) **which occurs many times** and this can lead to **mutational priming**
 * whether it occurs or not is another matter and depends on types of environmental change
 * moreover a large enough population is required to get this to get competition between population sweeps.

(Note: at this stage one could ask what the relationship is between information threshold and population-based diversity? In the intermediate environment the population initially cannot adapt because the environment changes too fast, however it evolves to be able to adapt and improves itself relative to the information threshold).


 * Relationship to real organisms?**
 * this model is more similar to real organisms than the density classification model (which seems good)
 * but in the model the **only** way to evolve is by mutational priming! Other possibilities are excluded. It is therefore only a **proof of principle**, not of occurrence.
 * in the density classification model the system chooses its own coding.

Next: Ecosystem based problem solving


 * Reference**s
 * Crombach ABM & Hogeweg P** (2007) Chromosome rearrangements and the evolution of genome structuring and adaptability.//Mol. Biol. Evol.//, **24**: 1130-1139. [|MEDLINE]. [|DownLoad PDF.]
 * Dunham M et al** (2002) Characteristic genome rearrangements in experimental evolution of Saccharomyces cerevisiae. PNAS 99:16144–16149. [|link]
 * Ferea T et al** (1999) Systematic changes in gene expression patterns following adaptive evolution in yeast. PNAS 96:9721–9726. [|link]

mutational priming, ferrea's experiment, anton's work, CA of CA density classification.

COURSE 2006-2007

In the previous section we have seen the emergence of mutational priming in evolving density classifiers: where with one mutating they can swtich between two ways of classifying.

Q: Can we see this effect in Biological systems?

Ferea et al.: - experimental evolution on yeast on glucose poor medium (constant glucose levels, cf chemostat) (- in nature: yeast grows on fruit with depleting levels of sugars and need to shift of other types of metabolism; diauxic shift) - so in this not natural situation they looked at gene expression levels - after just a few generations: - small number of mutations, but evolved faster growth - 10% of genes significantly differentially expressed compared to ancestor - half of those where in the TCA cycle: overexpressed - half were fermentation genes: underexpressed - there was variation between experiments: but pattern vary consistent - this is in parallel to diauxic shift!: cf attractors in metabolic network - in evolution: genotype-phenotype mapping is complex, what can evolution to do for attractors (Kauffmann paper?): deform outcome in state space?

In later paper: mutational priming in yeast (Dunham et al 2002 PNAS) - they found characteristic genome arrangements in experimentally evolution of yeast - mutations: type of mutations mainly duplications / deletions / cross-chromosomal rearrangements (in few weeks!) - and AT THE SAME PLACES!, but also REASONABLE PLACES (fermentation + respiration genes) - transposon-related sequences at the break-points .... suggestive

Bottom-lines: - short term evolution: a lot can change! - but also the same kind of changes: - constraints? - but also see signal: break elements: chromosomal rearrangments, put break elements at correct place helps mutational priming (-does not prove mutational priming, but very suggestive ....)

So what can theory help to say about this: - not to prove this is the case in real systems - but show mechanism - without mechanism only have statistics!

Conjecture: All mutations are equal but some are more equal than others ............

Crombach and Hogeweg (2006?): Next step theory

Study whether mutational operators in yeast can lead to mutational priming (vs artificial system)

Model: Artificial example, but: - organism like coding - organism like mutational operators Q: can this lead to mutational primiing?

- gene duplication / deletion - transposon duplication / delection - repeat elements, breakpoints when deleted - double strand breaks + repairs: can lead to errors: gross-chromosomal rearrangments through repeat elements - transposons expand fast and die out ??? - selection: fitness criterion - forget gene regulation for the moment, assume more genes give more product (duplication) - fluctuating environment - space - individuals: long chromosome - 20 genes should be expressed normally - 20 genes shouuld be either in 1 or 2 copies to cope with ENV 1 or 2 - loose genotype: bad - gene too much or little: fitness difference

Results - as environment changes fitness drops (change still slow enough to allow adaptation by single gene duplication) - evolution then increases fitness: adaptation by single gene duplications later: - adaptation is much quicker: goes to target in bigger jumps, i.e. by gross-chromosomal rearrangements - SO: self-organization of genomes by clustering genes that need to be duplicated!

Slow change environment: - most individuals are fit most of the time Slow times 10: - initially equally distributed over fitness: can't adapt: population heterogeneity (based diverstiy) - then over time: enough signal to rearrange genome and later can adpat to environment! - now they can change fast and so track environment - here genomes are actually better organized than for slow change: many breakpoints separate dupl. and non-dupl. genes Even faster: - evolution to intermediate solution because environment can't be tracked

So: - strong selection on change in the environment which occurs many times: CAN LEAD TO MUTATIONAL PRIMING - whether it occurs or not is another matter and depends on types of environmental change - NEED LARGE ENOUGH POPULATION! to get competition between population sweeps.

(Q: What is the relationship between information threshold and population-based diversity? Previously population can't adapt because environment change too fast, but evolves to be able to adapt, improves info threshold??)

Relationship to real organisms? - better than density classification (so that is good?) - but ONLY way to evolve is by mutational priming! (other possibilities are excluded): i.e. proof of principle, not of occurance - in density classification system chooses its own coding.