Electron Crystallography

Henning Stahlberg

http://C-CINA.org

3D model(small

proteins)

3D modelResolution:

12 Å

3D modelResolution: x,y: 2.5 Åz: 3.0 Å

2D surface Resolution:

x,y: 5 Åz: 1 Å

3D modelResolution:

1.5 Å

Nuclear Magn.Resonance

Atomic ForceMicroscopy

X-Ray Diffraction

Single Particles

2DCrystallization

3D Crystallization

Purified (Membrane) Protein

Electron Microscopy

3D modelResolution: x,y: 2.5 Åz: 3.0 Å

2D surface Resolution:

x,y: 5 Åz: 1 Å

3D model(small

proteins)

3D modelResolution:

12 Å

3D modelResolution:

2 Å

Nuclear Magn.Resonance

Atomic ForceMicroscopy

X-Ray Diffraction

Single Particles

3D Crystallization

Purified (Membrane) Protein

Electron Microscopy

2DCrystallization

Single Particle TEM

Expression:His-glpF

in E. coli,

solubilizedwith OG,

purified onNi-Agarose,

neg. stainedsingle particles,

TEM image

GlpF

3D modelResolution: x,y: 2.5 Åz: 3.0 Å

2D surface Resolution:

x,y: 5 Åz: 1 Å

3D model(small

proteins)

3D modelResolution:

12 Å

3D modelResolution:

2 Å

Nuclear Magn.Resonance

Atomic ForceMicroscopy

X-Ray Diffraction

Single Particles

3D Crystallization

Purified Membrane Protein

Electron Microscopy

2DCrystallization

Imaging

Purification

Sample Prep

Image Processing

Model Building

2D Crystallization

Imaging

Bacterial

Yeast

Insect

Mammalian

Cell-Free

Expression

Sample Prep

Image Processing

Model Building

2D Crystallization

Expression

Imaging

lipiddouble layer membrane

protein

detergentmolecule

solubilized membrane protein

detergentmicelles

Solubilization: FOS-12…14, DHPC, DDM

Purification: DMReconstitution: OTG, OG

Detergents

Purification

The Transmission Electron Microscope

-200’000 Volt !

Sample

Viewing Screen

Vacuum !

Negative Stain EM Grid Preparation-e

Proteins

Negative Stain

Carbon-Film

BF-TEM image of negatively stained MloK1 membrane protein

50 nm

Sample Prep

Image Processing

Model Building

2D Crystallization

Expression

Imaging

Chiu et al., Structure 2007

Purification

BF-TEM image of negatively stained MloK1 membrane protein

50 nm

52 Å

Sample Prep

Image Processing

Model Building

2D Crystallization

Expression

Imaging

Chiu et al., Structure 2007

Purification

Purification

Sample Prep

Image Processing

Model Building

Expression

Imaging

2D Crystallization

Jap et al., Ultramicroscopy (1992)

SampleDialysis

Membrane

Detergent removal by dialysis

Detergent capture by adding cyclodextrin

Purification

Sample Prep

Image Processing

Model Building

Expression

Imaging

2D Crystallization

Iacovache et al., J. Struct. Biol. (2010)

(c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

!"(c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

!#(c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

!$(c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

!%(c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

!&(c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

'((c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

'!(c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

(c) Slide from presentation given by Tom Walz, Harvard, on the Electron Crystallography Workshop at UC Davis, 2008

Cryo-EM Grid Preparation-e

Proteins

Vitrified Water

Carbon-Film

2D Crystals of Membrane Proteins

2 µm

Imaging

Electron Diffraction

Goals:

• User-friendly• Fast to install (Linux, OSX)

• Compatible (MRC, CCP4, SPIDER, Chimera, EMDEP, PDB...)

• Easy to learn and to get started• Transparent• Allows manual fine-tuning of every step• Flexible (easy to adapt to your own programs)

• Optionally fully automated• Include new algorithms

Image Processing

“Magic”ImageProcessing

MRC programsRichard Henderson et al.

Available from Jude Smith, MRC, Cambridge, UK

BryantGipson

MarcelArheit

Extract the information from the noiseFFTNoisy

image

Filtered FFT

Filtered image

Fourier Filtering

Image

FFT

Fourier Filtering

Image

FFT

Lattice Indexed(in FFT)

Fourier Filtering

Image

FFT

Lattice Indexed(in FFT)

Masked(in FFT)

Filtered Image

Fourier Filtering

Image

FFT

Lattice Indexed(in FFT)

Masked(in FFT)

Reconstruction (2D)

Reconstruction (3D)

Filtered Image

CTF world

Basic Unbending Algorithm

FFT

FFT MaskedFFT

mask= 1px

masktran

MaskedFFT

20px

PerfectFFT

UnbentImage

RawImage

Reference

FFT of ReferenceCC-Map

quadserchccunbendProfile

CCP4ctfapply

Amps&

Phases origtilt

Amps&

Phases

Collection of Amp&Phs

Amps&

PhasesAmps

&Phases

Amps&

PhasesAmps

&Phases

Amps&

PhasesAmps

&Phases

Amps&

Phases

Amps&

Phases

mmbox

maketran

(ApplyDefocus)

More Complete Unbending Algorithm

RawFFT

MaskedFFT

(24px)

UnbentImage

(III)

FFT of UnbentImage

(III)

Amps&

Phases(APH file)

CTF-corr.Amps &Phases

(APH file) Collectionof

APH files

MergedAmps &Phases

(APH file)

From other images

RawImage

MaskedFFT

(1 px)Reference

FFTCropped

Reference(400px)

Cross-Cor. Map

1

DistortionVectorField

1MaskedFFT

(20px)

MaskedFFT

(2 px)Reference

FFT

CroppedReference

(300px)Distortion

VectorField

2MaskedFFT

(22px)

UnbentImage

(I)

FFT of UnbentImage

(I)

MaskedFFT

(2 px)Reference

FFT

CroppedReference

(300px)Distortion

VectorField

3

UnbentImage

(II)

FFT of UnbentImage

(II)

Cross-Cor. Map

2

Cross-Cor. Map

3

UNBEND I

UNBEND II (1st iteration)

UNBEND II (2nd iteration)

Cryo-EM

Continous C-film

-180°C

Before!Image Processing

Cryo-EM

Continous C-film

-180°C

Before!! After Image Processing

Cryo-EM

Continous C-film

-180°C

Before!! After Image Processing

Real Space Fourier Space

x

y

z

The Third Dimension

Real Space Fourier Space

Diffraction Spots in 2D

The Third Dimension

Real Space Fourier Space

Diffraction Spots in 2D

Lattice Lines in 3D

x

y

z

Complex Values for Amplitude and Phase

in Z-direction

The Lattice Line

Fourier Space

0

Resolution in Z-direction0

The Radon projection theorem for 3D reconstructions

Fourier-Space Method

The Missing Cone in Fourier SpaceFourier Space

Diffraction Spots in 2DLattice Lines in 3D

Real Space

3D Reconstruction Cryo EMGlpF

A first 3D Map of a Waterchannel

AQP1

3D cryo TEM3.8 Å resolution

Mitsuoka et al. J. Struct. Biol. (1999)

Atomic model

3D cryo EM, 3.8 Å resolution, Murata et al. Nature 2000

Interpretation

3D cryo EM, 3.8 Å resolution, Murata et al. Nature 2000

Conclusions• Water channel is

amphipathic

• H20 traverses in a 1D string

• 2 Asn dissect string, positive residues and helix dipoles: all inhibit permeation of H3O+

Grotthus “hop (I) and turn (II)” proton conduction mechanism

De Grotthuss, C.J.T. (1806). Mémoire sur la décomposition de l’eau et des corps qu’elle tient en dissolution a l’aide de l’électricité galvanique. Ann. Chim. 58, 54-74

Molecular dynamics

Bert de Groot & Helmut Grubmüller, Science (2001)300 K, 4 AQP1, 271 lipid, 25’891 H2O10 ns molecular dynamics simulation

The Nobel Prize in Chemistry 2003

"for discoveries concerning channels in cell membranes"

Peter Agre

Johns Hopkins University School of Medicine Baltimore, MD, USA

b. 1949

Roderick MacKinnon

Rockefeller University, Howard Hughes Medical Institute

New York, NY, USA b. 1956

"for the discovery of water channels"

"for structural and mechanistic studies of ion

channels"