The Latest from the NMR Collaboration Bicelles n’ Stuff

Group Meeting

October 23, 2008

20 amino acid antimicrobiel peptide.

Part of the immune system (for fungi).

Forms channels in membranes which allow water and ions to go through.

Destroys membrane potential.

Alamethicin

Motivation:

Potentially a good and simple membrane channel model

Insight into this family of proteins could be valuable in the development of antibiotics

Gln7

Glu18

Gln19

How does alamethicin interact with the membrane?

Liquid NMR Setup

1 M stock solutions of DMPC and DHPC in 10 mM phosphate buffer at pH 6.6.

Mixed with lyophilized alamethicin, repeated cycles of vortexing and centrifugation.

DMPC and phosphate buffer added to yield a solution with 300 mM total lipid concentration and 15 mM peptide.

10% D2O added for field lock. Vortexed and centrifuged several times until clear solution.

NOE : lipid-peptide hydrogen interactions PRE : Paramagnetic Relaxation Enhancement. Gadodiamide (gadolinium

complex) remains in solution. Affects the relaxation in a distance dependent manner => tells where the different hydrogens are positioned in the bilayer.

Sample Preparation

Output

AA MD of Alamethicin in DMPC 25 peptides and 330 lipids + water ~125,000 atoms. Energy-minimized and equilibrated for 200 ps with peptides fixed. Energy-minimized and equilibrated for 1 ns in the NPT ensemble

(T = 323 K, P = 1 atm, NAMD and CHARMM FF) Production run of 100+ ns in the NPzAT ensemble

(T = 323 K, Pz = 1 atm, A: fixed at end of eq.)

0 ns (after eq.)124 ns

NOE Stuff

Difficult to Interpret Spectra Absence of Hα protons and NMR-resolvable side-chain

protons in Aib

Two prolines => lack of amide protons

Lipid acyl groups (positions 4-13) are degenerated to one resonance at 1.28 ppm => Only first and second CH2 groups of the lipid acyl chains as well as the terminal methyl group are separated

AA

NOESY spectrum of alm in bicelle

NOE Connectivities

(CH

2 ) 4-13(C

H2 ) 4-13

ring-H

20, H20

ba

ckb

on

e (N

H)3

-19

CH3

(NH2)7

(NH2)19

(NH)1

(CH3)3

(CH3)3

(CH

2 ) 4-13(C

H2 ) 4-13

ring-H

20, H20

ba

ckb

on

e (N

H)3

-19

CH3

(NH2)7

(NH2)19

(NH)1

(CH3)3

(CH3)3

MD Connectivities

ChoCH3

4 4 2 1 1 0 0 0 0 0 0 0 0 0 0 1 1 2 16 22 17

Acy

C2

7 2 3 2 2 1 2 0 1 2 0 0 1 0 1 4 2 0 30 24 16

AcyC3

7 2 2 2 2 1 2 0 1 1 0 0 1 0 1 2 1 0 25 27 14

Acy

C4-13

19 4 4 9 15 8 16 9 21 41 16 5 6 4 16 7 3 2 56 100 83

Acy

CH3

5 0 0 1 3 2 3 6 6 8 7 3 1 2 3 2 1 0 8 17 10

Aib1Aib3

Ala4 Ala6 Aib8 Aib10 Leu12 Val15 Aib17 Gln19 Gln7 Phl20Aib5 Gln7 Val9 Gly11 Aib13 Aib16 Glu18 Phl20 Gln19

0-1 1-3 3-6 6-10 10-15 15-20 20-30 30-50 50-100

backbone HN

NH2 HPhe

Should be redone with larger rlimit

strange

# proton contacts weighted by r -6

Helix Tilt (MD)

0.000

0.005

0.010

0.015

0.020

0.025

0.030

0.035

0.040

0.045

0 10 20 30 40 50 60 70 80 90

helix tilt / deg.

PRE Stuff

PRE results

< 0.1 0.1-0.2 0.2-0.3 0.3-0.4 0.4-0.6 0.6-0.8 > 0.8 undet. non-protons

Close to water interface

Far from water interface

s-1

Is this alamethicin structure found from the same type of

experiment?

An average structure could be constructed from the MD sim as well.

Aib1Aib3

Gln7

Ala4

Aib5

Ala6

Val9

Aib8

Gly11

Aib10

Leu12

Val15Aib13

Aib16 Aib17

Phl20

Glu18

Gln19

Aib1Aib3

Gln7

Ala4

Aib5

Ala6

Val9

Aib8

Gly11

Aib10

Leu12

Val15Aib13

Aib16 Aib17

Phl20

Glu18

Gln19

MD Peptide-Water Contacts

If I made this data for the lipids as well, I bet it would not look as nice as the NMR results

# peptide proton-water oxygen contacts weighted by r--6

75 40 45 44 15 8 4 0 1 4 1 0 0 0 2 4 3 3 79 100 47

Aib1Aib3

Ala4 Ala6 Aib8 Aib10 Leu12 Val15 Aib17 Gln19 Gln7 Phl20Aib5 Gln7 Val9 Gly11 Aib13 Aib16 Glu18 Phl20 Gln19

0-1 1-3 3-6 6-10 10-15 15-20 20-30 30-50 50-100

backbone HN

NH2 HPhe

Close to water interface

Far from water interface

But what about the cross peak to DHPC??

Main Classes : Nonpolar (N), Charged (Q), Apolar (C), Polar (P)Subclasses (N,Q) : Hydrogen bond donor (d), acceptor (a), both (da), none (0)Subclasses (C,P) : Degree of polarity (1-5)

MARTINI CG

Class C – apolar

NAMD:Class Nda – nonpolar and hydrogen bond donor and acceptor

MARTINI:Class P – polar

Class Na – nonpolar and hydrogen bond acceptor

Class Qa – charged and hydrogen bond acceptor

Class Q0 – charged

4

41

212

4

2

1

DPPC4Ctail

DLPC3Ctail

DHPC2Ctail

DMPC (C1-14) DHPC (C1-6)

Bicelles – how do they really look? 240 DHPC (2Ctail) 120 DMPC (3Ctail)

Simulations so far q = [DMPC]/[DHPC] = long/short = 0.5

120 3Ctail ; 240 2Ctail ; 12000 W

6 Alm ; 120 3Ctail ; 240 2Ctail ; 12000 W

120 3Ctail ; 240 2Ctail ; 48000 W

120 4Ctail ; 240 2Ctail ; 12000 W

Previous slide

1Ctail ??

A Volume Argument

bz

d / Å h / Å bz / Å Vbil/nm3 Vrim/nm3

140 40 0 616 310

140 40 10 531 395

140 30 10 398 233

90 30 10 150 154

100 50 0 393 374

bilayer part

rim part

DMPC (C1-14)DHPC (C1-6)

V(DHPC) > V(DMPC)/2

Simulations so far q = [DMPC]/[DHPC] = long/short = 3.2

768 3Ctail ; 240 2Ctail ; 31500 W

768 4Ctail ; 240 2Ctail ; 31500 W

384 4Ctail ; 120 2Ctail ; 16000 W

~half size

Simulations so far

360 2Ctail ; 12000 W

360 4Ctail ; 12000 W

Random Thoughts so far Short and long lipids mix!!

I would like to see cases where they actually don’t q = 0.5 => bicelle/micelles q = 3.2 => ~bilayers What about other q’s ? Can we find the limits where bicelle go to bilayer and so on. The more lipid and/or water included in the simulation, the slower convergence to low

energy structure => can we just say it is fair to look at small systems? Equal distribution of the short lipids to relieve potential stress? Does the distribution of the lipids change over time? Relevant simulations should be repeated many times. Is the CG model of DHPC a good description – check the literature Is there a problem in representing DHPC with only one bead in each lipid tail?? Artifacts from number of lipids involved in simulation should be investigated. Dependence of the amount of water should be investigated. Reverse CG or simply redo in AA for the most interesting set-ups to validate CG Could something beside lipids in the NMR mixture affect size/shape/segregation? Could the NMR signals be explained by something else than the perfect-model bicelle

based on the simulations? What type of experiments have been done on bicelles, and what can we really be sure of? In lab it takes several hours for the lipid mixture to obtain the ”right” macro structures =>

several close minima that actually have different physical properties?