Initial Model Initial Model Generation Generation Workshop on - - PowerPoint PPT Presentation

Workshop on Advanced Topics Workshop on Advanced Topics in EM Structure Determination in EM Structure Determination The Scripps Research Institute The Scripps Research Institute La Jolla, November 2007 La Jolla, November 2007

Initial Model Initial Model Generation Generation

Jiang et al., 2002 Serysheva et al., 2003

The issue: Structures of the IP3 receptor as determined by single particle EM

Sato et al., 2004 Jiang et al., 2003

Structure determination by Structure determination by single particle EM single particle EM

Sample Sample EM specimen EM specimen Set of 2D images Set of 2D images Initial 3D model Initial 3D model Refined 3D model Refined 3D model

Specimen preparation Specimen preparation – – Negative staining Negative staining – – Cryo Cryo-

• negative staining

negative staining – – Vitrification Vitrification Data collection Data collection – – Images of Images of untilted untilted samples samples – – Tilt pairs Tilt pairs – – Tilt series Tilt series 3D reconstruction 3D reconstruction – – Random conical tilt Random conical tilt – – Ab initio Ab initio angle assignment angle assignment – – Tomographic Tomographic reconstruction reconstruction Refinement Refinement Refinement of Refinement of orientation parameters,

• rientation parameters,

magnification magnification, and CTF parameters , and CTF parameters Verification Verification Angular distribution Angular distribution Comparison with 2D data set Comparison with 2D data set Comparison with Comparison with known structures known structures

Sample Sample

Images that are used for a 3D reconstruction Images that are used for a 3D reconstruction have to be of identical molecules ! have to be of identical molecules ! Sample can be heterogeneous Sample can be heterogeneous even if even if it is it is biochemically biochemically homogeneous homogeneous – – conformational heterogeneity conformational heterogeneity – – unstable complexes unstable complexes Before attempting any 3D reconstruction: Before attempting any 3D reconstruction: Understand your sample ! Understand your sample !

We always prepare negatively stained specimens first We always prepare negatively stained specimens first good contrast & often preferred orientations good contrast & often preferred orientations (depends somewhat on preparation method) (depends somewhat on preparation method) We always calculate class averages We always calculate class averages (0 (0° ° images) images) different averages = different conformations different averages = different conformations (but can also be different orientations) (but can also be different orientations) We usually calculate 3D reconstructions (RCT) We usually calculate 3D reconstructions (RCT) different 3D maps = different conformations different 3D maps = different conformations (but can also be deformations) (but can also be deformations)

Sample Sample

Sample Sample -

• Negative staining

Negative staining

Need to record images Need to record images

• f tilted specimens
• f tilted specimens

Particles adopt preferred orientations Particles adopt preferred orientations

• n the continuous carbon film !
• n the continuous carbon film !

Many preparation artifacts Many preparation artifacts (incomplete stain embedding, (incomplete stain embedding, adsorption deformations, specimen flattening upon drying) adsorption deformations, specimen flattening upon drying) Limitation of the resolution to about 20 Limitation of the resolution to about 20 Å Å BUT BUT: : Very useful for Very useful for heterogeneous samples heterogeneous samples 3D reconstruction by 3D reconstruction by RCT RCT is very reliable is very reliable

Radermacher Radermacher et al et al. (1987) . (1987)

Tiltaxis Tiltaxis

Random conical tilt reconstruction Random conical tilt reconstruction

3D reconstruction

• f specimen

BACKPROJECTION

specimen at different tilt angles Different projection views

IMAGING ASSIGN ORIENTATIONAL PARAMETERS x, y and Φ

Random conical tilt reconstruction Random conical tilt reconstruction

Fractions from Superdex-200

M

Crude AMPA 13

14 15 16 17 18 19 20 21

200 116 95 66 45 31 21.5

RCT RCT – – AMPA receptor AMPA receptor

Heterogeneity due to different conformations Heterogeneity due to different conformations

Heterogeneity due to different conformations Heterogeneity due to different conformations

RCT RCT – – AMPA receptor AMPA receptor

Heterogeneity due to different conformations Heterogeneity due to different conformations

RCT RCT – – AMPA receptor AMPA receptor

RCT RCT -

• Sec23/Sec24 complex

Sec23/Sec24 complex

0° ° tilt tilt 60 60° ° tilt tilt Heterogeneity due to different orientations Heterogeneity due to different orientations Negligible deformations Negligible deformations

Bi Bi et al. et al., 2002 , 2002 Lederkremer Lederkremer et al. et al., 2001 , 2001 Heterogeneity due to different orientations Heterogeneity due to different orientations

RCT RCT -

• Sec23/Sec24 complex

Sec23/Sec24 complex

RCT RCT -

• Tf

Tf-

• TfR

TfR complex complex

Heterogeneity due to different orientations Heterogeneity due to different orientations Severe deformations Severe deformations

Face view Face view Top view Top view

Face view Top view

Conventional negative staining Conventional negative staining

RCT RCT -

• Tf

Tf-

• TfR

TfR complex complex

Heterogeneity due to different orientations Heterogeneity due to different orientations Severe deformations Severe deformations

Cryo Cryo-

• negative staining

negative staining

Addition of glycerol Addition of glycerol minimizes adsorption artifacts minimizes adsorption artifacts prevents specimen flattening prevents specimen flattening serves as serves as cryo cryo-

• protectant

protectant Carbon sandwich Carbon sandwich reduces incomplete stain embedding reduces incomplete stain embedding Freezing Freezing prevents specimen flattening prevents specimen flattening

Conventional negative staining Conventional negative staining 0° ° tilt tilt 60 60° ° tilt tilt Cryo Cryo-

• negative staining

negative staining 0° ° tilt tilt 60 60° ° tilt tilt

RCT RCT -

• Tf

Tf-

• TfR

TfR complex complex

Heterogeneity due to different orientations Heterogeneity due to different orientations Severe deformations Severe deformations

Face view Top view

Cryo Cryo-

• negative staining

negative staining

RCT RCT -

• Tf

Tf-

• TfR

TfR complex complex

Heterogeneity due to different orientations Heterogeneity due to different orientations Severe deformations Severe deformations

x y γ α β

5 parameters 5 parameters to determine to determine

Single particles in ice Single particles in ice

Angular reconstitution ( Angular reconstitution (Imagic Imagic) )

Serysheva Serysheva et al. et al., 1995 , 1995

• 2. add in further projections and keep
• 2. add in further projections and keep

refining refining

• 1. chose 3 projection images that are
• 1. chose 3 projection images that are

perpendicular views of the particle perpendicular views of the particle (anchor set) (anchor set) van Heel, 1987 van Heel, 1987

Serysheva Serysheva et al. et al., 1995 , 1995

Angular reconstitution Angular reconstitution -

• Ryanodine

Ryanodine receptor receptor

Angular reconstitution Angular reconstitution -

• Tf

Tf-

• TfR

TfR complex complex

Vitrified ice Vitrified ice

Angular reconstitution Angular reconstitution -

• Tf

Tf-

• TfR

TfR complex complex

Angular reconstitution Atomic model

Angular reconstitution (IMAGIC) Angular reconstitution (IMAGIC)

Angular reconstitution is best for: Angular reconstitution is best for: large specimens with symmetry, large specimens with symmetry, and available and available structural information structural information (can be obtained from (can be obtained from random conical tilt) random conical tilt) Structure depends critically on the anchor set Structure depends critically on the anchor set (these should ideally be 3 perpendicular views, (these should ideally be 3 perpendicular views, which is hard to know with an unknown molecule which is hard to know with an unknown molecule) ) Structure also depends on the order in which Structure also depends on the order in which additional projections are included additional projections are included

OP command (SPIDER) OP command (SPIDER)

A common A common-

• lines based method for determining orientations

lines based method for determining orientations for for N N > 3 particle projections simultaneously > 3 particle projections simultaneously Penczek Penczek et al. et al., 1996 , 1996 Applied to 70S ribosome from Applied to 70S ribosome from E. coli

• E. coli

(instead of (instead of trying to determine the Euler trying to determine the Euler angles based on pair angles based on pair-

• wise

wise angles of common lines in the projections angles of common lines in the projections’ ’ planes, one assumes that planes, one assumes that rotation matrices are known, finds set of angles of common lines rotation matrices are known, finds set of angles of common lines and and computes the overall discrepancy along these lines) computes the overall discrepancy along these lines)

Essentially the opposite of standard common lines approach Essentially the opposite of standard common lines approach

OP command OP command -

• Tf

Tf-

• TfR

TfR complex complex

OP reconstructions OP reconstructions

OP reconstruction (33 classes) OP reconstruction (160 classes) Atomic model

3D reconstruction in EMAN 3D reconstruction in EMAN

Also uses a set Also uses a set of projections

• f projections

to generate an initial model to generate an initial model

Baumeister et al. (1999)

Electron Electron tomographic tomographic reconstruction reconstruction

± 90° 2° steps ± 60° 2° steps ± 90° 5° steps ± 60° 5° steps

Electron tomography Electron tomography -

• Clathrin

Clathrin-

• coated vesicles

coated vesicles

QuickTime™ and a Cinepak decompressor are needed to see this picture.

Electron tomography Electron tomography -

• Clathrin

Clathrin-

• coated

coated vesicles vesicles

Model refinement Model refinement -

• Tf

Tf-

• TfR

TfR complex complex

Angular reconstitution Atomic model Refined density map (FREALIGN)

Reference model (face view in cryo-negative stain) Atomic model filtered to 40 Å Alignment of 500 class averages

Model refinement Model refinement -

• Tf

Tf-

• TfR

TfR complex complex

Model refinement Model refinement -

• Tf

Tf-

• TfR

TfR complex complex

Reference model (top view in cryo-negative stain) Atomic model filtered to 40 Å Alignment of 500 class averages

Model refinement Model refinement -

• Tf

Tf-

• TfR

TfR complex complex

Reference model (face view in conventional negative stain) Atomic model filtered to 40 Å Alignment of 500 class averages

Model refinement Model refinement -

• APC

APC

Anaphase promoting Anaphase promoting complex ( complex (S.

• S. pombe

pombe) ) in negative stain in negative stain

RCT reconstruction in negative stain Reconstruction with vitrified ice data using CNS RCT as reference model RCT reconstruction in cryo-negative stain

Model refinement Model refinement -

• APC

APC

RCT reconstruction in negative stain Reconstruction by aligning ice data to RCT map in cryo-negative stain Reconstruction by aligning raw images to RCT map in negative stain Reconstruction by aligning class averages to RCT map in negative stain

Model refinement Model refinement -

• APC

APC

Model verification Model verification -

• APC

APC

3D variance map Penczek et al., 2006 Reconstruction

• f vitrified ice data

using CNS RCT as reference model Reconstruction

• f vitrified ice data

using OP command

Model verification Model verification -

• APC

APC

Clathrin Clathrin cages in vitrified ice cages in vitrified ice

Model verification Model verification -

• Clathrin

Clathrin cages cages

700 Å

Model verification Model verification -

• Clathrin

Clathrin cages cages

Model verification Model verification -

• Clathrin

Clathrin cages cages

EM X-ray

mGluR1 mGluR1

Kunishima Kunishima et al. 2000 et al. 2000 (K. (K. Morikawa Morikawa) )

KcsA KcsA

Doyle et al. 1998 Doyle et al. 1998 (R. MacKinnon) (R. MacKinnon)

GluR2 GluR2

Armstrong et al. 2000 Armstrong et al. 2000 (E. (E. Gouaux Gouaux) )

Model verification Model verification -

• AMPA receptor

AMPA receptor

Model verification Model verification -

• AMPA receptor

AMPA receptor

Model verification Model verification -

• AMPA receptor

AMPA receptor

5 nm 90 90˚ ˚ 90 90˚ ˚ 90 90˚ ˚ 5 nm 90 90˚ ˚ 90 90˚ ˚ 90 90˚ ˚

S.

• S. pombe

pombe Cdc5p complex Cdc5p complex Mammalian C complex Mammalian C complex

( (Jurica Jurica et al. et al., 2004 , 2004) )

Model verification Model verification -

• Spliceosome

Spliceosome

Model verification Model verification -

• Spliceosome

Spliceosome

S.

• S. pombe

pombe Cdc5p complex in vitrified ice Cdc5p complex in vitrified ice using cryo using cryo-

• negative stain

negative stain reconstruction as initial model reconstruction as initial model

Model verification Model verification -

• Spliceosome

Spliceosome

Conclusions Conclusions Never just believe your initial model ! Never just believe your initial model !

There is currently no general way There is currently no general way to generate a reliable initial model to generate a reliable initial model

Calculate random conical tilt reconstructions Calculate random conical tilt reconstructions

• f (cryo
• f (cryo-
• )

)negatively stained specimens ! negatively stained specimens !

The density map is probably distorted, but it is The density map is probably distorted, but it is a good basis to interpret subsequent reconstructions a good basis to interpret subsequent reconstructions

Use your Use your “ “biological intelligence biological intelligence” ” ! !

Some Literature Some Literature

Frank, J. (2006) Three Frank, J. (2006) Three-

• Dimensional Electron Microscopy of Macromolecular Assemblies

Dimensional Electron Microscopy of Macromolecular Assemblies Oxford University Press, Inc. Oxford University Press, Inc. Radermacher, M., Radermacher, M., Wagenknecht Wagenknecht, T., , T., Verschoor Verschoor, A. and Frank, J. (1987) , A. and Frank, J. (1987) Three Three-

• dimensional

dimensional reconstruction from a single reconstruction from a single-

• exposure, random conical tilt series applied to the 50S ribosoma

exposure, random conical tilt series applied to the 50S ribosomal l subunit of Escherichia coli. subunit of Escherichia coli. J.

• J. Microsc

Microsc. . 146 146: 113 : 113-

• 136

136 Van Heel, M. (1987) Van Heel, M. (1987) Angular reconstitution: a posteriori assignment of projection di Angular reconstitution: a posteriori assignment of projection directions for rections for 3D reconstruction. 3D reconstruction. Ultramicroscopy Ultramicroscopy 21 21: 111 : 111-

• 123

123 Serysheva, I. I., Serysheva, I. I., Orlova Orlova, E. V., Chiu, W., Sherman, M. B., Hamilton, S. L. and , E. V., Chiu, W., Sherman, M. B., Hamilton, S. L. and van Heel, M. (1995) van Heel, M. (1995) Electron Electron cryomicroscopy cryomicroscopy and angular reconstitution used to and angular reconstitution used to visualize the skeletal muscle calcium release channel. visualize the skeletal muscle calcium release channel. Nat.

• Nat. Struct
• Struct. Biol.

. Biol. 2 2: 18 : 18-

• 24

24 Penczek, P. A., Zhu, J. and Frank, J. (1996) Penczek, P. A., Zhu, J. and Frank, J. (1996) A common A common-

• lines based method for determining

lines based method for determining

• rientations for N > 3 particle projections simultaneously.
• rientations for N > 3 particle projections simultaneously. Ultramicroscopy

Ultramicroscopy 63 63: 205 : 205-

• 218

218