Non-specific DNA-protein interaction: How proteins can diffuse along DNA

Transcription

1 XXIX International Colloquium on Group-Theoretical Methods in Physics, Nankai University, Tianjin, Chine. Non-specific DNA-protein interaction: How proteins can diffuse along DNA Fabien Paillusson, Maria Barbi, Jean-Marc Victor " LPTMC, UPMC - Paris VI Aleksandra Nivina Master student April-July 2012 Marie Jardat, Vincent Dahirel PECSA, UPMC - Paris VI 14/08/12 MARIA BARBI 0/21

2 DNA-binding proteins Example :? 14/08/12 MARIA BARBI 1/21

3 search of target sequences on DNA DNA molecule " ( basepairs) Reaction kinetics k a protein target sequence (5-15 basepairs) 1 τ k d Experimental = k a M 1 s 1 Debye-Smoluchowski theory: 3D diffusion Calculated 1 = k th a 10 8 M 1 s 1 τ th Riggs, JMB, /08/12 MARIA BARBI 2/21

4 search of target sequences on DNA 1. displacement along DNA 1D diffusion (sliding) 3D diffusion (jumps) «intermittent process» «facilitated diffusion» 14/08/12 MARIA BARBI 3/21

5 1. protein 1D diffusion (sliding) experimental evidences (fluorescence microscopy) Bonnet et al NAR 2008 the protein" diffusion " «movie»: DNA extremities protein y x time <x 2 > " (10-3 µm 2 ) diffusion " (MSD x): diffusion coefficient: time (s) D µm 2 /s (300 pb) 2 /s 14/08/12 MARIA BARBI 4/21

6 search of target sequences on DNA 1. displacement along DNA 1D diffusion (sliding) 3D diffusion (jumps) «intermittent process» «facilitated diffusion» 2. sequence recognition interaction (H bonds) with nucleotides «reading» 14/08/12 MARIA BARBI 5/21

7 2. sequence recognition direct interaction:" hydrogen bonding to the bps sides major groove major groove A T G C minor groove minor groove EcoRV and DNA 14/08/12 MARIA BARBI 6/21

8 mobility-specificity paradox energy E(n) sequence dependent interaction (H bonds) rough potential " position n " strongly reduced mobility, " trapping [J Biol Phys 04, PRE 04] " our starting point 14/08/12 MARIA BARBI 7/21

9 Two-state models sliding state / reading state : a possible solution for the paradox how? 1. protein fluctuates between 2 conformations: hypothesis of Mirny lab partially " denaturated fully " folded Slutsky and Mirny" Biophys.J., we propose an alternative mechanism " where the distance between protein and DNA plays a role 14/08/12 MARIA BARBI 8/21

10 physics of the protein-dna interaction electrostatics in solution" [Mol Phys 09, PRL 09, PCCP 11] charge -2e/pb most often positively charged DNA! (σ DNA = - e/nm 2 ) ions DNA-binding protein! (+0.2 σ DNA ) analytical approach: Poisson-Boltzmann (F. Paillusson PhD) MC simulations: Marie Jardat et Vincent Dahirel PECSA-UPMC 14/08/12 MARIA BARBI 9/21

11 oppositely charged surfaces in solution Poisson-Boltzmann equation (1D) : σ(0) 0 σ(l) L x 1. electrostatic potential V(x): " Poisson equation " (ρ(x) = local charge density between plates) 2. ions between plates: " Boltzmann statistics dv(x) dx with B.C.: ρ(x) = q i c 0 βq i V ( x) e i = ρ(x) ε dv dx dv dx σ(0) (0) = ε σ(l) (L) = + ε 14/08/12 MARIA BARBI 10/21

12 oppositely charged surfaces in solution Poisson-Boltzmann equation (1D) : σ(0) σ(l) in monovalent salt dv( x) dx = c0 ( ) ε sinh β q V( x) i 0 L x with B.C.: dv dx dv dx σ(0) (0) = ε σ(l) (L) = + ε V(x) pressure P(L) free energy E(L) [Mol Phys 09] 14/08/12 MARIA BARBI 10/21

13 protein shape curvature effects can be " accounted for by the " Derjaguin approximation: E(L) = A da E ( ) plate-plate 14/08/12 MARIA BARBI 11/21

14 results: protein shape curvature effects can be " accounted for by the " Derjaguin approximation: Free energy (k B T) MC simulations" PB + Derjaguin convex shape: " attraction E(L) = A da E ( ) plate-plate surface-surface distance L (nm) [Mol Phys 09, PRL 09, PCCP 11] 14/08/12 MARIA BARBI 11/21

15 protein shape matters! results: concave shape: " repulsion for L < 0.5 nm Free energy (k B T) convex shape: " attraction surface-surface distance L (nm) [Mol Phys 09, PRL 09, PCCP 11] 14/08/12 MARIA BARBI 11/21

16 protein shape results: concave shape: " repulsion for L < 0.5 nm osmotic origin : trapped ions protein DNA protein DNA Free energy (k B T) convex shape: " attraction ions 0 ion density/bulk 30 pertinent for «DNA-binding " proteins», concave surface-surface distance L (nm) [Mol Phys 09, PRL 09, PCCP 11] 14/08/12 MARIA BARBI 11/21

17 implications on sliding osmotic repulsion + electrostatics + hydrogen bonds DNA-protein distance L (nm) [Mol Phys 09, PRL 09, PCCP 11] 14/08/12 MARIA BARBI 14/21

18 implications on sliding osmotic repulsion + electrostatics + hydrogen bonds single H-bond = Morse potential number n of H-bonds dependent " on the position z along DNA E H (z, L) = n(z) e H (L) DNA-protein distance L (nm) (different positions z) [Mol Phys 09, PRL 09, PCCP 11] 14/08/12 MARIA BARBI 15/21

19 implications on sliding osmotic repulsion + electrostatics + hydrogen bonds E (k B T) barrier sliding mode minimum" (secondary) 30 bp E (k B T) n max -1 reading mode minimum" (primairy) DNA-protein distance L n max : target" sequence [Mol Phys 09, PRL 09, PCCP 11] «facilitated sliding»!! 14/08/12 MARIA BARBI 16/21

20 «facilitated sliding»: a proof of principle Aleksandra Nivina master internship original landscape (charged convex protein) : discrete version a. original landscape random walk on the energy profile (Arrhenius law) two «control» landscapes : b. no interaction landscape c. no repulsion landscape" (spherical protein) 14/08/12 MARIA BARBI 17/21

21 «facilitated sliding»: a proof of principle very preliminary results: no significant role on the time of target search 4000 Mean time of target search (arbitrary units) Complete landscape Flat landscape No repulsion landscape a b c 250 proteins 300 base pairs 8000 arbitrary units " observation time 14/08/12 MARIA BARBI 18/21

22 «facilitated sliding»: a proof of principle very preliminary results: increase the chances of finding the target within a fixed time % of proteins that have found the target 50,00% 45,00% 40,00% 35,00% 30,00% 25,00% 20,00% 15,00% 10,00% Complete landscape Flat landscape No repulsion landscape 5,00% 0,00% 42,04% 24,36% 18,96% a b c 250 proteins 300 base pairs 8000 arbitrary units " observation time 14/08/12 MARIA BARBI 19/21

23 «facilitated sliding»: a proof of principle very preliminary results: more DNA visited (regardless of distance L) % of positions visited along DNA (L) 50,0% 45,0% 40,0% 35,0% 30,0% 25,0% 20,0% 15,0% 10,0% Complete landscape Flat landscape No repulsion landscape 5,0% 0,0% 44,6% 34,7% 21,4% a b c 250 proteins 300 base pairs 8000 arbitrary units " observation time 14/08/12 MARIA BARBI 20/21

24 perspectives 1. confirm these results with different parameters and/or different charges and shapes 2. compare to new structural data (different properties for different protein classes?) 3. pass to a 3D model with helical path to obtain realistic quantitative results 4. compare to experimental data (association/dissociation constants, diffusion coeff ) Y ADN protéine X 14/08/12 MARIA BARBI 21/21

25 M3V group modélisation multi-échelle de la matière vivante multiscale modeling of living matter Laboratoire de Physique Théorique de la Matière Condensée Université Pierre et Marie Curie - Paris VI Annick LESNE Julien MOZZICONACCI Jean-Marc VICTOR Maria BARBI Fabien PAILLUSSON Pascal CARRIVAIN Christophe LAVELLE Hua WONG Aleksandra NIVINA Poisson-Boltzmann for oppositely charged bodies: an explicit derivation Paillusson, Barbi, Victor, Molecular Physics, 107, 1419 (2009) Non-specific DNA-protein interaction: Why proteins can diffuse along DNA Dahirel, Paillusson, Jardat, Barbi, Victor, Phys. Rev. Lett, 102, (2009) Effective interaction between charged nanoparticles and DNA Paillusson, Dahirel, Jardat, Victor, Barbi, Phys. Chem. Chem. Phys. 13, (2011) 14/08/12 MARIA BARBI