Evolutionary models of conflict and cooperation. May 13, 2010 Joao Xavier

Transcription

1 Evolutionary models of conflict and cooperation May 13, 2010 Joao Xavier

2 Evolutionary cooperation is central to all life Why evolutionary cooperation is important Simple definitions: Evolution Fitness Social interactions Mathematical models of social evolution Cooperation and conflict in microbial pathogens Cooperation and conflict in cancer

3

4 Evolution is the change of gene frequency within a population Four mechanisms: Mutation Natural selection Drift ( genetic drift, or random drift ) Migration

5 Fitness is a quantitative measure of natural selection Δt (one generation) Y p = p = Y+B The absolute fitness of genotype Y is Y Y +B W(Y) = Y Y The change in frequency is

6 Higher fitness can lead to genotype fixation Evolutionary dynamics Explicit solution of longterm consequences p p=1 (fixation) t

7 Dynamics equation provides evolutionary trajectory Evolutionary dynamics 1 p = 1 (fixation) p 1 p' 0 p p

8 Dynamics equation provides evolutionary trajectory Evolutionary dynamics 1 p 0 p' p = 0 (extinction) p

9 Fitness can be context dependent Example: Cancer cells that have resistance to a drug, but that grow worse than sensitive cells in the absence of drug Context: No drug Context: Drug present p' p' p p

10 Fitness is often frequency dependent When W(Y) and W(B) are not constants, but rather a function of p p' Equilibrium (p*) p

11

12 COOPERATION IS A CENTRAL THEME IN THE HISTORY OF LIFE

13 Biological organization grows on cooperation

14 Traits can be classified according to their social effects An individual carrying out a social trait is an actor The individual affected by that trait is the recipient Effect on recipient Positive Negative Effect on actor Positive Negative Mutual benefit Altruism Selfishness Spite Cooperation Competition West, S.A., Griffin, A.S., Gardner, A. & Diggle, S.P. (2006) Social evolution theory for microorganisms. Nature Reviews Microbiology 4,

15 Public good cooperation is an example of altruism Example: Cancer cells secrete a metalloproteinase that is costly to synthesize (reduces growth rate) Metalloproteinase degrade extracellular matrix and allow cells to proliferate better Cooperators only (Low mixing) Public good W(C) = W 0 C + B Basal fitness Cost of synthesis Benefit from invasion

16 A definition of public good from economics Is available to all individuals within population Is costly to produce Incentive to cheat

17 Cheaters do better than cooperators by exploiting a public good Pure population of cooperators Public good Pure population of cheaters Mix Cheaters do relatively better, but at the expense of the whole group

18 Prisoner s dilemma: a fundamental problem in game theory Prisoner B Cooperate Defect Prisoner A Cooperate Defect Each serves 6 month A: goes free B: 10 years A: 10 years B: goes free A: 5 years B: 5 years

19 Prisoner s dilemma: A metaphor in evolutionary theory Payoff matrix: Cooperator Defector Cooperator b c, b c b/2 c, b/2 Defector b/2, b/2 c 0, 0 Key: b benefit from cooperation c cost of cooperation Nowak MA (2006). Evolutionary Dynamics: Exploring the Equations of Life. Harvard University Press.

20 Evolutionary games are based on pair wise interactions Cooperator Cooperator Defector Defector Cooperator Defector

21 Evolutionary theory can explain the evolution of cooperation Five rules of cooperation (Nowak 2006 Science) Kin selection: r > c/b (r is relatedness) Interactions are random Direct reciprocity: w > c/b (w is chance of repeated encounter) Indirect reciprocity: q > c/b (q is probability of knowing reputation) Network reciprocity: b/c > k (k is average number of neighbors) Group selection: b/c > 1 + (n/m) (n is group size, m is number of groups) Problem: Rules derived from game theory assume pair wise interactions Interactions among cells are seldom pair wise Interactions are structured

22 Diffusible public goods can evolve in spatially structured populations Structured Mixed Nadell et al, FEMS Mic Rev, 2009

23

24 Bacteria rely on multicellular traits Strength by numbers Secretion of virulence factors Biofilm formation Quorum sensing What prevents evolutionary cheating? Public good + Cooperator "cheater"

25 Cells in biofilms are embedded in a matrix of extracellular polymeric substances 100 µm Bacteria (red) and glycoconjugate matrix (green). Courtesy T. Neu, UFZ Germany.

26 Biofilm dynamics emerge from interactions among cells end start 1. Growth 2. Division 3. Spreading growth division 1 cell 2 cells

27 Gradients produce heterogeneous growth Diffusion Reaction Vertical cross-section through biofilm Gradients of oxygen concentration Liquid y Biofilm x Solid substratum

28 Producers and cheaters compete within the biofilm Boundary layer at constant distance from biofilm Liquid in boundary layer Xavier and Foster, PNAS (2007)

29 With oxygen gradients OFF cheaters win Xavier and Foster, PNAS (2007)

30 With oxygen gradients ON cheaters lose Xavier and Foster, PNAS (2007)

31 Lower polymer density Xavier and Foster, PNAS (2007)

32 Extracellular polymers can work as a competitive trait Cooperative trait Open to exploitation Competitive trait Benefit own lineage Xavier and Foster, PNAS (2007)

33

34 Swarming: collective motility in Pseudomonas aeruginosa

35 Swarming: collective motility in Pseudomonas aeruginosa

36 Pseudomonas aeruginosa a model for bacterial pathogenesis Sequenced genomes Non redundant libraries of transposon mutants Affymetric GeneChip microarrays It is an opportunistic pathogen Forms biofilms in cystic fibrosis lungs that are hard to treat with antibiotics

37 Swarming benefits the colony but requires biosurfactant synthesis by cells What prevents evolutionary cheating?

38 Biosurfactant synthesis is well characterized Glucose 6 phosphate Ketocyl ACP Acyl ACP Glucose 1 phosphate type II fatty acid cycle (FASII) dtdp D glucose Hydroxyacyl ACP Enoyl ACP dtdp 6 deoxy D 4 hexulose dtdp 6 deoxy L lyxo 4 hexulose RhlA Wild-type dtdp L rhamnose 3 (3 hydroxyalkanoyloxy) alkanoates (HAAs) RhlB Mono rhamnolipids RhlC rhla - (Non-cooperator) Di rhamnolipids

39 Different genotypes are distinguishable using neutral colors GFP RFP mix 10 2x10 cells in colony 10 1x10 GFP RFP mix

40 Biosurfactants are a public good

41 Biosurfactant secretion is uncheatable Non cooperators do better than when alone but at expense of wildtype Not enough to distinguish who wins, WT or rhla - Measured relative fitness:

42 Reporter fusions allow monitoring gene expression dynamics P. aeruginosa PA14 rhlab-gfp rhlab expression OFF rhlab expression ON [h]

43 Quorum sensing is necessary yet not sufficient High density Low density rhlab ON rhlab OFF [h]

44 Expression of biosurfactant synthesis is favored at lower nitrogen source levels Carbon source: Glycerol (C 3 H 5 (OH) 3 ) Nitrogen source: (NH 4 ) 2 SO 4

45 rhla regulation ensures metabolic prudence C C C C C N Medium both carbon and nitrogen but carbon is in excess C C N C C C C C Cells grow while there s nitrogen then use excess carbon to secrete rhamnolipids

46 Inducible rhlab bypasses metabolic prudence mechanism No inducer (behaves like noncooperator) Inducer present (strict cooperator)

47 Biosurfactant secretion in strict cooperator is cheatable Day 1 Day 2 Day 3 Day 4

48

49 Summary Bacteria rely on multicellular traits for many tasks Multicellular cooperative traits are open to exploitation and therefore must have evolved with mechanisms for robustness We can find the mechanisms stabilizing bacterial multicellularity: Physical or biological mechanisms setting populations structure Molecular mechanisms (metabolic prudence, quorum sensing, more?) Can lead to new therapies

50

51 Cancer is a disease of clonal evolution within the body Nowell. "Clonal evolution of tumor-cell populations" (1976) Science

52 Can cooperation evolve de novo in cancer progression? Hanahan and Weinenberg (2000) The hallmarks of cancer Six traits necessary to lead a neoplasm to cancer malignancy Hypothesis: If some of these hallmarks are cooperative (benefit other cells) lineage complementation can greatly speed up the appearance of malignancy Hanahan and Weinenberg (2000) The hallmarks of cancer

53 Cell cell communication between cancer and tumorassociated macrophages With Johanna Joyce (MSKCC Cancer Biology & Genetics)