Mechanistic basis of equal plasmid spacing in E. coli: a general mechanism for sizing and spacing

Transcription

1 Mechanistic basis of equal plasmid spacing in E. coli: a general mechanism for sizing and spacing CANES, Kings College London November 11 th, 2015 Martin Howard Computational & Systems Biology John Innes Centre, UK

2 Low copy number of plasmids causes a problem! Usually small pieces of DNA separate to main chromosome Plasmids sometimes present at very low copy numbers Causes a problem at cell division: how to ensure plasmid are transmitted reliably to both daughter cells? Active segregating mechanism needed Importance: Intrinsically interesting basic biology Many plasmids encode important functions e.g. antibiotic resistance or virulence Method of segregation prior to cell division is a primitive ancestor of mitotic apparatus in eukaryotes

3 Plasmid pb171 equipositioned across nucleoid Plasmid encodes ParABC machinery... Ietswaart et al, PLOS Comput. Biol. (2014)

4 Plasmid pb171 equipositioned across nucleoid Ietswaart et al, PLOS Comput. Biol. (2014) Very far from random!

5 parabc partitioning system ParA: ATPase that binds nonspecifically to DNA potentially polymerizes on nucleoid surface ParB: protein that binds to DNA at parc regions together form partition complex ParB interacts with ParA functioning as adaptor between ParA and parc ParB stimulates ATPase activity of ParA ParA dynamics require ParB and parc centromere-like site Plasmids/Par protein dynamics occurs along nucleoid surface Might remind you of the MinCDE system but there are differences! Howard et al, PRL (2001)

6 ParA structure retraction correlates with plasmid motion Effective pulling dynamics Ringgaard et al, PNAS (2009)

7 Possible mechanisms: passive or active? Diffusionimmobilization Directedmotion Both cases: plasmid moves up ParA concentration gradient Ietswaart et al, PLOS Comput. Biol. (2014)

8 parabc system predicted to enhance plasmid mobility par system enhances mobility! Can explain rapid plasmid segregation events Rules out diffusionimmobilization

9 How are plasmids equally spaced? Imagine a single plasmid, positioned away from mid-nucleoid: [ParA] Plasmid/ParB ParA fluxes position ParA can bind anywhere but can only unbind at a plasmid In order to balance fluxes, higher concentration to left of plasmid Plasmid moves towards higher concentration of ParA System self-organizes to equally spaced positioning

10 How are plasmids equally spaced? Imagine a single plasmid, positioned away from mid-nucleoid: L [ParA] plasmid ParA fluxes x position Integrate: J = Jx L J + J(L x) = L Only balance when x = L/2!

11 Simulations of directed motion model Ietswaart et al, PLOS Comput. Biol. (2014) 1d model gives robust equipositioning Independent of precise concentrations or degree of polymerization Any mechanism where plasmids are pulled up ParA concentration gradient will work! (µm)

12 Model predicts and experiments confirm lack of oscillations Unbiased image analysis: Normalized asymmetry measure: close to one if highly oscillatory Not the case for model or experiments

13 Where does 1d order come from? All modelling so far is 1d Why do plasmids and ParA line-up? WT Perhaps nucleoid imposes 1d order through its own internal organization If so, expect plasmid equipositioning defects in nucleoid organization mutants

14 Conclusions I Low copy number plasmids exhibit active partitioning mechanisms Diffusion/immobilization mechanism is ruled out experimentally Equal plasmid spacing results from self-organized competing ParA structures that direct plasmids on a structured nucleoid Mechanism relies on flux balance Modelling successfully predicts that - ParA oscillations infrequent and caused by perturbed plasmid spacing - Disrupted nucleoid architecture perturbs plasmid positioning

15 Is flux balance a general mechanism for sizing/scaling? Possible application to cell size control

16 How do cells regulate their size? Basic question, studied for many years Mechanism unclear in any cell type Marshall et al, BMC Biology (2012) Lots of hypotheses, e.g. - make proteins in a pulse - dilute out with growth - divide when concentration falls below threshold Easy to come up with a plausible hypothesis hard to verify experimentally Potentially multiple control mechanisms

17 How does it work in fission yeast? Data is consistent with cells growing to a fixed size before dividing Wood & Nurse, Cell Cycle (2013) Sizer mechanism How is this controlled?! Fission yeast is the almost the only organism for which we have enough molecular understanding to try to answer this question

18 Most work has focused on the properties of two proteins: Pom1 and Cdr2

19 Pom1 is an inhibitory intracellular gradient Pom1: cell division regulator Forms intracellular concentration gradient Padte et al, Curr. Biol. (2006) Paoletti et al, J. Cell Sci. (2006) Mechanism of (noisy!) gradient formation now quite well understood Saunders et al, Dev. Cell (2012) Howard, Trends Cell Biol. (2012) Hachet et al, Cell (2011)

20 Pom1 gradient as a cell-size sensor? Pom1p Mid1p, Cdr2p Wee1p Moseley et al, Nature (2009) Martin & Berthelot-Grosjean, Nature (2009)

21 Great idea, but gradient not used in this way! Pom1 levels at medial cortex almost constant for different cell lengths in population of cells Same results holds when single cells are monitored Pan et al, elife (2014)

22 What is sensing cell size?! Could it be related to medial accumulation of cdr2? Cdr2-GFP Pom1-dTomato Width of cdr2 nodal region increases only slightly with cell length Total cdr2 nodal intensity rises much more

23 What is sensing cell size: nodal cdr2! Medial density of cdr2 rises with cell length! Cell population Individual cells

24 Could cdr2 be a dose-dependent cell size regulator? Put cdr2 under inducible control Expect over/under expression of cdr2 to cause alterations in length at cell division Observed experimentally! Over-expression effect is subtle

25 So how is size sensing achieved? What is the mechanism behind rise in medial cdr2 density? Global information about cell size converting into local information Idea: cdr2 is an area sensor

26 Nodal cdr2 is highly dynamic Recovery on timescale of a few minutes Much quicker than cell cycle timescales Justifies quasi-steady-state approach

27 Cytoplasmic cdr2 concentration is constant

28 So how could area sensing be achieved? Allow cdr2 to bind all over membrane Then cdr2 moves to nodes Nodes have roughly fixed area Flux balance creates bottleneck leading to rising local density Pan et al, elife (2014) Related to microtubule length control antenna model

29 Model for cdr2 dynamics Constrain parameters from experiments (e.g. FRAP) Cdr2 density scaling from model agrees with experiments

30 Model predictions: membrane binding Cdr2 should be able to bind to the membrane everywhere before transiting to nodal region Needs to sense area! Is this true? Yes!

31 Cdr2 density predicted to scale with area Model predicts that cdr2 is an area sensor Is this true?! Normally difficult to test: in fission yeast, length, area and volume are all proportional! Use rga4, rga2 cells to perturb width!

32 Does size at mitosis scale with area? Not perfect but much better correlation than with length or volume

33 Conclusions II What is the mechanism of cell size control?! Surprisingly pom1 is not key size sensor Instead involves cdr2 accumulation in medial zone of approximately fixed area Key feature may be area sensing employing a flux balance mechanism Plenty of open questions: How general is area sensing?

34 Conclusions II Not completely general: variety of solutions available: e.g. E. coli appears to add a constant length, ( adder ) regardless of birth length Other cases (e.g. budding yeast) lie between an adder and a sizer Jun & Taheri-Araghi, Trends Microbiol. (2015)

35 Acknowledgements Cell Size: Timothy Saunders (John Innes Centre/EMBL/National Univ of Singapore) Giuseppe Facchetti (John Innes Centre) Plasmids: Robert Ietswaart (John Innes Centre) Cell size: Kally Pan, Ignacio Flor Parra, Fred Chang (Columbia Univ) Plasmids: Florian Szardenings, Kenn Gerdes (Newcastle Univ, Copenhagen Univ) :