I Insights on Antibiotic Translocation : Molecular Dynamics - - PowerPoint PPT Presentation

I

Insights on Antibiotic Translocation : Molecular Dynamics Simulations

Amit Kumar, Eric Hajjar,Enrico Spiga, Francesa Collu, Paolo Ruggerone & Matteo Ceccarelli

Marseille : 11 –04-2008

Outline

1.

Model to assess translocation : What can we compare with experiments ?

2.

Interrelation between Experiments and Simulations

3.

Metadynamics : Different reaction co-ordinates , free energy quantification

4.

Validations for simulation runs ( RMSF , RMSD )

METHODS Results : Successful stories in light with experimental findings :

1 . Carbenicilllin diffusion through WT and D113N_E117Q mutants OmpF

• 2. Ampicillin with Mutants R132A and D113N
• 3. Floroquinilones : Comparison between Moxifloxacin and Enrofloxacin
• 4. Cephalosporins : Comparison between Cefpirome and Cefetamet (preminilary results)

Modeling Translocation From Blockage

 BLM experiment : The

Rate constant k , can be obtained by noise analysis in frequency spectra .

Free energy model : Every 2 blockage corresponds to 1 translocation.

SINGLE MOLECULE EXPERIMENTS :

Interruption in ion flow associated with blockage of antibiotic

Nestorovich et al. PNAS vol . 99 , 15, 9789-9794 2002

Interrelation between experimental and simulation data

 Transition Rate Theory :

7 kB T = 1 ns 14 kB T = 1 µs 21 kB T = 1 ms

K.E.Cooper et al , J. Membrane Biol. 106 95-105 (1988)

E n e r g y A B

∆G

Reaction Co-ordinate

Metadynamics

Metadynamics : Accelerates the process to be observable within the reach of Standard MD Simulations . Main problem: how to choose the reaction coordinates of a given process ? Provides Coarse-garined description for the process in phase space defined by a set of reaction ccordinates . Laio and Parrinello , 99 , PNAS 2002

Reaction Coordinates

0- degree 180- degree

Z- coordinate : The axis of diffusion is defined as center of mass (com )of the antibiotic with respect to com of system along z . Angle : Defined as Orientation of molecule's dipole with respect to com

• f system along z .

H-bond : Defined as number of H- bond between the antibiotic and the protein .

Free Energy Surface Evaluation

Z

• c
• r

d i n a t e Angle

9 Kcal /mol. 1 . Free Energy construction in the phase space of the chosen Reaction co-ordinates . 2 . Each Color Corresponds to 1 Kcal/mol .

• 3. High Intensity of the color

correspond to strong interactions .

Validation of Simulation runs

Root Mean square Fluctuations and Converted B-factors

L1 L5 L4 L3 L2 L8 L7 L6

Validation for Simulation Runs

(ii) Root Mean Square Deviations

Area Calculation along OmpF



Solvent accessible area was calculated using in house built program :Hope for the X-Ray structure of OmpF .



Good agreement with the well known program Hole1 . ( Between Z (-6:6)Å )



Useful to calculate the fluctuation of area (with / without) presence of antibiotic with an effort to correlate with BLM experiments. Z

• a

x i s Area (Ų) X-Ray Structure : 2OMF (PDB ID ) O.S.Smart et al , Biophysical Journal 65 2455-2460 (1993)

I) The case of Carbenicillin

Carbenicillin (CRB) has charge of -2 . The diffusion of CRB through ompf is difficult due to the repulsion between the carboxy group and the residues 113 and 117 near the constriction region.

OmpF Investigated :

• 1. Wild Type (WT) 2.

D113N_E117Q : Double Mutant (DM)

On Screening the repulsion : By Mutation of the residues D113  N113 and E117Q117 can we expect the translocation process for CRB ?

To answer our question :

Free Energy Surface Evaluation

Angle

• Carbenicillin with Wild Type (WT) mutant no translocation .
• Carbenicillin with D113N_E117Q mutant translocation

Movies For WT and D113N_E117Q OmpF

Why No translocation through WT OmpF

The strong minimum at Z = ( 7.8 to 9.8 )Å ang = ( 70 to 90 ) deg

Z-axis

Angle

Mini -1

The antibiotic makes strong H-bonds with residues R42 , and switches its H- bonded interaction between K80 , R168

Observation along the trajectory for WT

• 1. For most part of the total trajectory

( ~20ns) the antibiotic stays in Mini-1.

• Favorable Interaction of oxygens with

K80 , 167, 168 .

• Move towards constriction region :

parallel to axis of diffusion , phenyl group down. (min z= 3.8Å)

• Repulsion effect and low entropy due to

unfavorable environment.

Snapshot for a configuration where CRB arrives near constriction region

Distance Time (20ns)

Figure C on the right depicts distance between D113 – COO (black) and E117 – COO (red)

C

Translocation for CRB with DM

Z-axis Angle

Observation along trajectory of translocation

Inventory Interaction map for Minima's near Constriction region : Mini 2 and Mini 3

 In the two minima's the oxygens of the antibiotic makes H-bonded interactions

with the stacked arginines ( Switching between R132 , R82 , R42 , K16 ) and in particular we observe the oxygens making interactions with N113 and Q117 . Mini -2 Mini -3

Area Calculation for the Minima's for DM

The Mini-2 and Mini-3 which lie near/on the constriction region in presence of antibiotic the area is as low as ~2Ų in the constriction region , closure of pore. Area (Ų) Z-axis (Å)

Comparing WT with DM

 Mini-1 which lies very much above the constriction r

egion , the WT and DM have nearly similar interaction . ( mutations has not much effe ct ) .

 For WT CRB does not arrive near the constriction region , due to repulsion

and supported low entropy for CRB created by the residues D113 , E117 .

 For DM near constriction region : No repulsion effect + Higher Entropy

= Translocation .

 Interactions with N113 and Q117 essential for translocation .

Comparing with experiments for CRB

 BLM experiments : our partners ( Bremen )

• 1. WT ---------> No Blockage -------> No Translocation.

2. DM----------> No Blockage --------> No Translocation.

 Simulation Results :

• 1. WT --------------> No Translocation
• 2. DM ---------> Translocation.

 Area calculation : in presence of antibiotic area ~2Ų , the pore is blocked

perhaps for a timescale not resolved by exper iments (~10µs) .

The effects of Ompf mutants on ampicillin diffusion

BASIC RESIDUE MUTATION : R132A change in  charge + size ACIDIC RESIDUE MUTATION : D113N change in charge

View of Constriction Zone

Free Energy Surface

Mini Mini

Angle

Z-coordinate

Interaction map for mini's near constriction region

SER125-OH

Observation along trajectory of translocation



For Mutation of R132A –movement of ampicillin towards A132 – finds more space H- bonded interaction with residues localized in neighbour-hood of A132 . The

• rientation is parallel to axis of diffusion. ( phenyl group down)



On Mutation D113N – H-bonded interactions with stacked arginines (42 82 132 ) and E117 . Orientation --- perpendicular . Phenyl group slides along the L3 loop

• --- Phenyl group top

figure on right represents the area sliced for Z= -2 (constriction region) along the simulation time .

AREA CALCULATIONS

Simulation time Area (~ 9 Ų) Area ~ (0 Ų) Area (Ų) Area (Ų) Z-axis (Å) figure on right represents the sas area available along Z- axis , for the minima of R132a , D113n .

Comparison with Experiments

 BLM experiments (Bremen)

• 1. D113N ------- Slight Increase

in background noise .

• 2. R132A -------No Blockage ----- No translocation

 Simulation Results

• 1. D113N ----- Translocation ~ Area ~ 0Ų
• 2. R132A ---- Translocation A

rea ~9Ų

 Liposome swelling assay experiments (Bremen and Porto) : Swelling rate

for mutants R132A and D113N more compared to wild type , confirms translocation of ampicillin for the mutants .

3) Flouroquinolones

Floroquinolones are more bulky and hydrophobic compared to the family of penicillin’s.

Moxifloxacin Enrofloxacin

Moxifloxacin and Enrofloxacin are zwitterionic and charged zero .

Translocation with coo- group up. Translocation with coo- group down.

Free Energy Represenatation

MOXIFLOXACIN ENROFLOXACIN 3

Comparison with BLM experiments

Area calculation for the constriction region all along the trajectory of translocation .

Moxifloxacin Enrofloxacin

Complete closure Complete closure and and conformational change conformational change

• f the pore during the translocation
• f the pore during the translocation

Comparison with fret experiments

) ( ) ( 1 1 ) (

6 6

t R t R t E

+ =

Where R0

6 is the foster radius (E = 0.5 % ) ; and R6 is the

distance between the dipole of Trp and Moxifloxacin. E(t) – measures the energy transfer between the trp and the antibiotic .

E ETrp61

Trp61=0.957

=0.957 E ETrp214

Trp214=0.378

=0.378

Comparison with Fret experiments

High value for Etrp 61 implies difffusion of moxi through OmpF

Further details see poster on Floroquinolones

Moxifloxacin Tryptophane

Conclusions and Prespectives for Floroquinolones

• From Simulations : translocation of Moxifloxacin and Enrofloxacin .
• Thanks to the flouresence property in Floroquinolones , which allow us to compared

diretly with the FRET experiments . We observe a high efficiency of energy transfer between Trp 61 and moxifloxacin . Good agreement with experiments .

• Complete closure of pore and conformational change is the pore (expansion) to

accomadate the bulky antibiotic . (agreement with BLM experiments)

• We extend the methodology from Moxi to Enro and observe similar interactions

between the antibiotic and OmpF during the translocation process . (need overlap integral for Enro (from our partners in Porto to calculate the E(t) )

4) Cephalosporins

CEFPIROME ( CFR) CEFETAMET ( CFT)

• Charge 0

Charge (-1)

Free Energy Representation

I

V11

ANGLE Z-Coordinate

CFR -WT CFT -WT

Preminilary results for CFT

Above the constriction region

H-bond interactions with the stacked arginines , similar conformation for Mini2 , Mini3 and Mini4 . H-contacts with Met-38 Movie : translocation of Cefetamet (CFT)

Work In progress

• In progress

Ceftizoxime

• Preliminary results

(translocation) Cefetamet (chagre –1) In progress In progess In progress Cefepime Analysis to completed Completed (Eric’s talk) Completed (Eric’s talk) Cefpirome

D121A D113A WT

CEPHALOSPORINS

Acknowledgement

1 . Thanks to our Network partners : Experimental data

• 2. Attilio Vargiu : Valuable help Parameterisation of Antibiotics

Computing Facility as CASPUR (Rome) , CINECA ,CYBERSAR (Cagliari ) . Financial support