Diffusion, Reaction, and Biological pattern formation

Transcription

1 Diffusion, Reaction, and Biological pattern formation

2 Morphogenesis and positional information How do cells know what to do?

3 Fundamental questions How do proteins in a cell segregate to front or back? How does an embryo become differentiated into specialized parts? How does an initially uniform tissue become specialized into multiple parts based on chemical signal?

4 Chemical patterns inside cells Back: Rho PTEN Front: Rac PI3K, PIP 2, PIP 3 What process(es) account for segregation of chemicals?

5 Patterns on a larger scale

6 Patterns in development Drosophilla

7 Interpreting a chemical gradient? gradient Morphogen gradient: can it lead to multiple cell types?

8 Wolpert s French Flag

9 How it was proposed to work Spatial gradients of morphogens create the subdivision Threshold concentrations of morphogen trigger gene expression in the cells in the tissue, leading to distinct expression.

10 Thresholds gradient

11 Example: early morphogenesis in the fly (Drosophila)

12 Embryo initially has no major internal boundaries Front (anterior) back (posterior)

13 Stages

14 Stages

15 Later: Patterns of gene products Bergmann S, Sandler O, Sberro H, Shnider S, Schejter E, et al. (2007) Pre-steady-state decoding of the Bicoid morphogen gradient. PLoS Biol 5(2): e46. doi: / journal.pbio

16 What is the question? How can we account for spontaneous creation of such striped patterns of protein activity from the initial fertilized egg? Bergmann S, Sandler O, Sberro H, Shnider S, Schejter E, et al. (2007) Pre-steady-state decoding of the Bicoid morphogen gradient. PLoS Biol 5(2): e46. doi: / journal.pbio

17 Reaction diffusion systems and Patterns

18 On its own, diffusion promotes Diffusion uniformity A 2 t = D A x 2 A t=0 Units: [D]= L 2 /t t>0 x

19 Linked to some reactions, it can CAUSE patterns to form spontaneously Reaction A t = f (A,B) B t = g(a,b) + Diffusion A 2 t = D A x 2 B g=0 A t=0 f=0 A t>0 Units: [D]= L 2 /t x

20 Example: (Schnakenberg) A t = f (A,B) + D 2 A A x 2 B t = g(a,b) + D 2 B B x 2 S 1 A S 2 k4 B B + 2A k 3 3A f (A,B) = k 1 k 2 A + k 3 A 2 B g(a,b) = k 4 k 3 A 2 B B g=0 f=0 A

21 Example: Shnakenberg RD system A 0 time 200 Starting close to the HSS, the system evolves a spatial pattern that persists with time.

22 How does it work? A.M. Turing, The Chemical Basis of Morphogenesis, Phil. Trans. R. Soc. London B237, pp.37-72, 1952 Derived conditions for diffusion-driven pattern formation in a reactiondiffusion system.

23 Why does it work? Basic idea Equations: BC s: sealed domain (no flux, i.e. Neumann)

24 Reaction mixture is stable Assume a stable homogeneous steady state:

25 Consider small perturbations of the homogeneous steady state perturbations: = = + + Substitute into PDEs and use Taylor expansions to linearize the equations

26 Linearized equations Where the coefficients are elements of the Jacobian matrix

27 Some requirements: Stability of the homogeneous steady state:

28 Eigenfunctions and eigenvalues Solutions to the linear equations are: Eigenfunction Eigenvalue Eigenfunctions chosen to satisfy both the PDE and BC s (e.g. no flux at x=0, L): q = nπ / L

29 Interpretation The form of the perturbations Cos(qx) e σt x The value q is the wave number (spatial periodicity) and σ is the rate of growth

30 Spatial instability pattern formation Cos(qx) e σt If the rate of growth σ is >0 then perturbations will grow x

31 Dispersive waves In general, waves of different periodicity will grow or decay at different rates, so the growth rate σ will depend on the wave number q.

32 Condition for pattern formation: There exists some (range of) wavenumber values such σ >0 σ q

33 Substitute the perturbations into the linearized equations What set of (algebraic) equation do you get?

34 Answer:

35 Rearrange terms This is a set of linear eqs in the alpha s which has a unique (trivial) solution unless the system has a zero determinant.

36 Nontrivial perturbations Exist only if determinant = 0 Simplifying leads to a quadratic eqn for σ:

37 Characteristic equation for σ: Eqn σ 2 -βσ +γ =0 Where β = - γ =

38 Technical stuff It is easy to show that for σ to be positive, γ has to be negative. γ= < 0 Look for wave number values such that γ<0

39 When is gamma negative? Gamma depends on q 2 quadratically a 11 /D 1 a 22 /D 2 q 2

40 When is gamma negative? γ = γ a 11 /D 1 a 22 /D 2 q 2

41 Gamma first becomes negative at γ a 11 /D 1 q min 2 a 22 /D 2 q 2

42 Condition for instability Use the criterion that γ (q min2 ) <0 Obtain (details omitted)

43 Conditions:

44 Interpretation (Detailed arguments omitted here.) Case 1: Case 2: In either case, need D 1 < D 2

45 Turing (1952): The chemical basis of morphogenesis Reaction (f,g) + Activator + - inhibitor - Diffusion D A << D B A t = f (A,B) + D 2 A A x 2 B t = g(a,b) + D 2 B B x 2 With appropriate reaction kinetics, an RD system can form spatial patterns due to the effect of diffusion. A.M. Turing, (1952) Phil. Trans. R. Soc. London B237, pp.37-72, 1952

46 Pattern formation in biology Phyllotaxis (Wikipedia) Hans Meinhardt

47 Reaction-Diffusion systems and animal coat patterns J.D Murray, A pre-pattern formation mechanism for animal coat markings, J. Theor. Biol. 88: , 1981

48 Is biological pattern all based on Turing mechanism?

49 Drosophilla Back to the fly

50 Back to the fly: How can we account for spontaneous creation of such striped patterns of protein activity from the initial fertilized egg?

51 Turing system? For some years, it was imagined that this pattern was set up by an activator-inhibitor system.. More recently, this has been overturned..

52 Bicoid synthesized from mrna at one end, then redistributes to form gradient Front (anterior) back (posterior)

53 Bicoid RNA Protein

54 Reading the bicoid gradient 0 L

55 Reading the bicoid gradient 0 L. PLoS Biol 5(2): e46. doi: / journal.pbio

56 Eric Wieschaus: "Cells make choices based on levels of bicoid. Nuclei can measure and make decisions based on those choices. Discovery of bicoid helped emphasize the importance of quantitative aspects of developmental biology.

57 Bicoid diffusion model Equation for the morphogen gradient

58 Bicoid gradient Variability from one embryo to the next how is error in body plan avoided?

59 Bicoid Other genes Feedback interactions sharpen the gradients