3D Modern Techniques - EPJ Web of Conferences

1 downloads 0 Views 5MB Size Report
Dynamics in Complex Sys- tems over .... Heteronuclear correlation building blocks. 1/2J. 1/2J. 1H. 15N .... Water magnetization at z prior to WATERGATE module.
EPJ Web of Conferences 30, 03002 (2012) DOI: 10.1051/epjconf/20123003002 © Owned by the authors, published by EDP Sciences, 2012

Experiment and Modelling in Structural NMR November 28th – December 2nd 2011 INSTN – CEA Saclay, France

Nicolas Birlirakis

ICSN CNRS, France 3D modern techniques [03002]

Organized by Thibault Charpentier Patrick Berthault Constantin Meis

[email protected] [email protected] [email protected]

Article available at http://www.epj-conferences.org or http://dx.doi.org/10.1051/epjconf/20123003002

This is an Open Access article distributed under the terms of the Creative Commons Attribution License 2.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Lectures



ORGANISATION

History, General principles

M. Goldman, CEA Saclay

Formalism and Interactions

H. Desvaux, CEA Saclay

Molecular Modelling

M. Nilges, Institut Pasteur

Anisotropy for Conformational studies

M. Blackledge, CEA Grenoble

Oriented media

D. Merlet, Univ. Paris-Sud 11

Techniques of solid-state NMR

D. Massiot, CNRS Orléans

NMR crystallography

F. Taulelle, UVSQ



Coordinator for the INSTN - CEA Saclay Constantin MEIS, INSTN/DIR



Coordinators for the CEA Patrick BERTHAULT, DSM/IRAMIS Thibault CHARPENTIER, DSM/IRAMIS



Coordinators for the Universities Pierre Richard DAHOO, Versailles Univ., UVSQ Francis TAULELLE, Versailles Univ., UVSQ Eric SIMONI University Paris-Sud 11

Winter-School

“Experiment and Modelling in Structural NMR”

Denis MERLET, University Paris-Sud 11

3D Modern Techniques

N. Birlirakis ICSN - CNRS

Interactions ligandmacromolecules

G. Lippens, University of Lille

Bio-Solid State NMR

H. Oschkinat, University of Berlin

Bioinformatics-protein assembly

R. Guérois, CEA Saclay

Materials Modelling & Simulations

C. Meis, INSTN - CEA, Saclay

DFT calculations of NMR parameters.

J. Yates, Oxford University

Inorganic MaterialsCeramics

S. Ashbrook, University of St Andrews



Secretary, Information and Registration Dominique PARISOT, INSTN - CEA Saclay Bat. 395 - PC N°35 Gif-sur-Yvette F-91191 : +33 1 69084882 email: [email protected]

A limited number of scholarships can be obtained for PhD students

contact : [email protected]

Amorphous materials

T. Charpentier, CEA Saclay

Hyper-polarization

G. De Paepe, CEA Grenoble

Fast NMR

N. Giraud, Univ. Paris-Sud 11

NMR Diffusometry : Dynamics in Complex Systems over many decades

R. Kimmich, University of Ulm





The aim of the school is to provide basic understanding on the coupling of theoretical Modelling and Simulation methods to the Experiments in structural NMR, for biology and material science. It is composed of Seminars in the morning and Practical Training in the afternoon involving both the theoretical and experimental aspects.

November 28 - December 2, 2011 INSTN - CEA, Saclay http://www-instn.cea.fr (events 2011)

3D NMR experiments applied to proteins H

H +

H3N

C*

COO-

+H N 3

R amino acid

C

C

R1

O

H

H

N

C

C

R2

O

H

H

N

C

COO-

Rn

peptide < 50aa < protein

Essential constituents of all living organisms Linear chains consisting of 20 L-amino acids Their folding to particular 3D structures triggers their biological function

Attractive targets for therapeutic applications once their structure-activity relationship is understood

03002-p.2

NMR studies of peptides (> ∆F Typical hard pulse lengths @ 800MHz, max admitted power: 1H → 7-...µs 15N → 39.5µs 13C → 14µs

03002-p.9

Shaped selective pulses Amplitude modulation: first trials Rectangle

Sinc

Triangle

Amplitude

Amplitude

Amplitude

Phase

Phase

Phase

200

3000

400

2000

1000

600

0

800

-1000

-2000

200

[points]

-3000

3000

Hz

400

2000

600

1000

90°

0

800

-1000

-2000

200

[points]

-3000

3000

Hz

400

2000

1000

600

0

-1000

800

-2000

[points]

-3000

90°

90°

Shaped selective pulses Amplitude modulation: peak selection Gaussian_1

Gaussian_1

Gaussian_1 Amplitude

Amplitude

Amplitude

Phase

Phase

Phase

200

3000

2000

400

1000

600

0

90°

-1000

800

-2000

-3000

200

[points]

Hz

3000

400

2000

1000

600

0

800

-1000

-2000

270°

-3000

200

[points]

3000

Hz

400

2000

1000

0

180° 180°y, -y

90° Gaussian δ and J evolution during pulse:

600

270°

90°x

τp

τp

03002-p.10

τp/2

-1000

800

-2000

-3000

[points]

Hz

Hz

Shaped selective pulses Amplitude/phase modulation: band selection (top-hat profile) Gaussian Cascade and BURP famillies Cascade Q3

Cascade Q5

U-BURP 2

Amplitude

Amplitude

Amplitude

Phase

Phase

Phase

200

6000

400

4000

2000

600

0

90°

800

-2000

-4000

-6000

[points]

Hz

200

6000

400

4000

2000

600

0

-2000

800

-4000

-6000

200

[points]

Hz

6000

4000

400

2000

180°

600

0

-2000

800

-4000

-6000

[points]

Hz

90°

Shaped selective pulses Amplitude/phase modulation: band selection (top-hat profile) Gaussian Cascade family - Universal pulses: Q5 (90°, 6.18) and Q3 (180°, 3.41/3.448) - Excitation pulse: G4 (7.82) - Inversion pulse: G3 (3.556) BURP famillies - Universal pulses: U-BURP (90°, 4.666) and RE-BURP (180°, 4.64/5.814) - Excitation pulses: E-BURP1 (4.514) and E-BURP2 (4.952) - Inversion pulses: I-BURP1 (4.498) and I-BURP2 (4.53) Figure of merit (depends on the shape): FM = τp∆Ω Frequency shifted (laminar) pulses : ∆φ = 2π τp ∆F / k

03002-p.11

Examples of selective excitation

Advantages

- of a spectral region

• zoom into a precise spectral region • suppression of indesired signals (solvent) • speed-up experiments • resolution increase - precise measurement of δ et J - refocusing of homonuclear couplings • posibility to direct transfers • possibility to treat an homonuclear system as heteronuclear

- of several peaks

- of two peaks

- of one peak

1D spectrum of strychnine in CDCl3

8.0

7.0

6.0

5.0

4.0

3.0

2.0

ppm

Examples of 2D selective experiments 90°

classical NOESY

90°

t1

90°

t2

tm

F1

1D selective NOESY 90°

90° 90°

t1

90°

tm

90°

t2

semi-soft NOESY F2 90°

soft NOESY

90°

t1

03002-p.12

tm

90°

t2

90°

tm

t2

Selective excitation in triple resonance experiments - Differentiation between 13Cα (or 13Caliph) and 13CO regions Use of rectangular pulses: τ180° = √3 / 2∆Ω and τ90° = √15 / 4∆Ω where ∆Ω is de difference of pulse carrier and the first null point Problems: inhomogenous excitation/inversion and cryoprobe arcing if executed simultaneously high power 13C and 15N pulses Use of universal top-hat pulses (or interleaved G4 and G4tr) 90° pulse comparison: 36µs square and 300µs G4

180° pulse comparison: 33µs square and 200µs Q3

0.80

0.80

0.60

0.60

0.40

0.40

0.20

0.20

-0.00

-0.00

-0.20

CO

-0.20

-0.40

Caliph

CO

-0.60

-0.40

Caliph

-0.60

-0.80

30000 20000 10000

0

-10000 -20000 -30000

ATTENTION Offset-dependent transient Bloch-Siegert phase shifts in homorefocusing

-0.80

Hz

30000 20000 10000

0

-10000 -20000 -30000

Hz

- Selective perturbation of 1HN region (BEST experiments) - Solvent selection (flipback or suppression)

Broadband inversion and decoupling - High field spectrometrs → large 13C spectral width in Hz → imperfect inversion/decoupling → use of adiabatic pulses:

smoothed CHIRP and WURST families

180° pulse comparison: 28µs square and 500µs CA-WURST-4 0.80

Adiabatic decoupling:

0.60 0.40 0.20 -0.00 -0.20

Caliph+arom

-0.40 -0.60 -0.80

30000 20000 10000

0

-10000 -20000 -30000

Use P4M5 supercycle Use bi-level decoupling for reduction of sidebands in 1H-1H-13C and 1H-1H-15N experiments Homonuclear decoupling → offset dependent Bloch-Siegert frequency shifts

Hz

- Adiabatic 15N inversion and decoupling is used to reduce power and thus probe heating, important in rapid scan conditions - Adiabatic pulses are tolerant to Rf inhomogeneity - Interesting pseudoadiabatic pulses: Bip and power-BIBOP for inversion and power-BEBOP for excitation

03002-p.13

Pulsed field gradients Linear variation of B0 at a certain direction - Coherence transfer pathway selection: - Artifact suppression per scan ie

1H12C

Σ γi pi Gi = 0

artifacts with short phase cycling (...substraction)

- Solvent suppression and control of water magnetisation (radiation damping) - Translational diffusion measurements - Magnetic Resonance Imaging gradient sensitivity enhanced HSQC (JSS, T1I, T2S) y 1H

1/4J

y

1/4J

1/4J

φ1

1/4J

1/4J

t2

1/4J

±y

t1

15N

cpd

80

±8

Water suppression Strong water (~50M) signal deteriorates spectral dynamic range High fields + cryoprobes → strong radiation damping → large water signal Strategies - Observe an heteroatom - Presaturation: forbitten in the case of fastly exchangable protons - Coherence selection with gradients - Return of water at z axis: polynomial sequences (jump and return...), water selective flip-back pulses, manipulation of radiation dumping - Dynamic filters: based on the differences in relaxation rates or diffusion coefficients - Selective defocusing of water via a gradient echo (WATERGATE, excitation sculpting) - Purging pulses: B0 or Rf gradient y

90°y

-y

1H

90°y

1H 15N

Gr

Gr

03002-p.14

1/4J

1/4J

-Iy -Iy

2IxSz Iy

SLx

2IzSz Iy φ1

The HN(CA)CO experiment y

1

H

∆1

y

-x

∆1 2∆1

2∆1 ∆1

WALTZ-y

HNi 15

y

∆2

N

∆2−t1/2

NHi 13

13



13Cα

∆3

∆3

∆2

15NH ∆3

i

CPD

i

BS

i,i-1

y

φ2

t2/2

CO

1HN t1/2

y

y

∆3

∆1

y

φ1

∆2

WG

t2/2

BS

BS

13CO i,i-1

Gz

INEPT version ∆1 = 1/41JHN = 2.7m ∆2 = 12m (relaxation optimized, 1/4 JNCα ~ 22m) ∆3 = 3.5-4m (relaxation optimized, 1/4 JCOCα = 4.5m)

The HN(CO)CA experiment y

1

H

∆1

y

-x

∆1 2∆1

2∆1 ∆1

WALTZ-y

HNi 15

N

y

∆2

∆2−t1/2

NHi 13

CO



13

t1/2 15NH

∆3 13CO i,i-1

∆3 BS y

φ2

t2/2

∆2

i

13CO i,i-1

t2/2

BS

13Cα

i,i-1

(DQ)

Gz

mixed INEPT / DQ version ∆1 = 1/41JHN = 2.7m ∆2 = 13.5m (relaxation optimized, 1/4 JNCO = 33m) ∆3 = 8m (relaxation optimized, 1/2 JCOCα = 9m) 03002-p.15

∆1 1HN

y

φ1

∆2

WG

i

CPD

The HNCACB experiment y

-x

∆1 2∆1

2∆1 ∆1

y

1

∆1

H

WALTZ-y

HNi

φ1

y

∆2

N

15

NHi

φ3

φ1

∆3

13Cα,β i,i-1 t2/2

∆3

1HN t1/2

∆2

15NH

y

y

t2/2

∆3

∆1

y

∆2−t1/2

∆2

WG

∆3

i

CPD

i



13

13Cα

13Cα

i,i-1

CO

13

i,i-1

BS

BS

Gz

mixed INEPT / RELAY sequence ∆1 = 1/41JHN = 2.7m ∆2 = 12m (relaxation optimized, 1/4 JNCα ~ 22m) ∆3 = 3.6m (relaxation optimized, 1/4 JCαCβ = 7.1m) 1

JCαCβ and 1JCβCγ couplings not refocused during t2

Cα, Cβ peaks of opposite sign

Rapid pulsing - BEST experiments - In 1H-15N correlation experiments 1HN longitudinal dipolar relaxation is accelerated if the rest of the protein protons (aliphatics) and 1H2O are unperturbed (at z axis) - Substitution of 1H hard pulses with shaped ones (top-hat ).

BEST-HSQC (JSS, T1I, T2S)

SOFAST-HMQC ( J SS, T2I, T2S)

y 1H

1/2J

1/2J

t2

1H

φ1 15N

1/4J

1/4J

1/4J

1/4J

t2

φ1

t1

cpd

15N

t1

: PC9 pulse (flip-angle 120°, Ernst angle) band-selective excitation pulse

: PC9 pulse (flip-angle 90°) band-selective 90° pulse

: band-selective universal 180° pulse

: band-selective universal 180° pulse

03002-p.16

cpd

Rapid pulsing - BEST experiments SOFAST-HMQC

10

9

BEST-HSQC

8

7

105

105

110

110

115

115

120

120

125

125

130

130

ppm

10

90° pulse → 1770µs Q5.1000

1

180° pulse → 1260µs Q3.1000

9

8

7

ppm

H carrier frequency @ 7.999ppm

P5M4 15N decoupling using 2ms CA-WURST-2 adiabatic pulse 1scan, 128 transients, AQ=70ms, RD=300ms, Total Experimental Time: 55s

Fast out and back experiments: BEST-HNCA

y

∆1 1

H

∆1

y

∆1

2∆1 y

y

15

N

∆2

y

φ1

t1/2

∆2

∆1

∆2−t1/2

∆2

CPD

y

φ2

t2/2

t2/2



13

CO

13

BS

BS

Gz

: 180° broadband inversion pulse → 155µs power-BIBOP pulse 10kHz inversion, tolerant to 20% inhomogeneity

03002-p.17

3D spectra for some BEST out and back experiments BEST-HNCO: 4scans, 58min

F2[ppm] 115

110

pm] F1 [p 40 50 60 130 125 120

115 pm] F1 [p 174 176 125 120

130 8

7

F3[ppm]

9

10

8

7

F3[ppm]

BEST-HN(CA)CO: 4scans, 59min

10

7

F3[ppm]

10

7

F3[ppm]

115 125

190 130

F1 170

] [ppm

120

115 pm] F1 [p 54 56 58130 125 120 8

8

110

F2[ppm] 110

F2[ppm] 110 115 120 125

9

9

BEST-HN(CO)CACB: 4scans, 4h 9min

F2[ppm]

9

180 130

pm] F1 [p

110

F2[ppm] 110 115 pm] F1 [p 55 60 65130 125 120

10

BEST-HN(CO)CA: 4scans, 2h

10

BEST-HNCACB: 4scans, 3h 54min

F2[ppm]

BEST-HNCA: 4scans, 1h 53min

9

8

7

F3[ppm]

10

9

8

7

Conclusions on BEST out and back experiments - Usefull for protonated proteins < 150aa (above, fractional deuteration is required) - Other fast methods are based on reduced sampling at the indirect dimensions: Spectrum compression through folding Exponential sampling and maximum entropy Projection reconstruction (radial sampling) Hadamard transform Filter Diagonalisation Method Sharc NMR ...

03002-p.18

F3[ppm]

Backbone and side chain assignment via transfer 3D experiments HBHA(CBCACO)NH H

In H2O:



N





O

H

C

N

Hα Cα

(H)CC(CO)NH-TOCSY / H(CCCO)NH-TOCSY O

H

C

N













i-1

i





O

H

C

N

Hα Cα

O C



i-1

i

In D2O: HACACO H N



HACACO-TOCSY

H



C

i-1

O

N

H

Hα Cα

C

i

O

N

Hα Cα

O

H

C

N

i-1

H(C)CH-TOCSY / (H)CCH-TOCSY O

H



C

N







H

C

N



O



C























i



O

i-1

i

The HCCH-TOCSY experiment δ → t1 J → 2∆1

H

1

∆1

y

∆1

t1/2 t1/2 ∆1 y

13

C

13CO

∆1/2

t2/2

∆1/2 t2/2

∆1

-y FLOPSI

∆1/2

∆1/2

BS

Gr purging

Refocused INEPT / z-TOCSY Constant time at both indirect dimensions ∆1 = 1/21JHC = 37m (optimized for CH2) Purging of residual water magnetization with B0 and RF gradients

03002-p.19

CPD

NMR studies of large proteins (>200aa) - Decrease transverse 13CH relaxation rates through deuteration: Expression in D2O → up to 70% deuteration (isotopic effect) Expression in D2O+U[13C, D]-glucose → up to ~100% deuteration Use of pyruvate or α-ketobutyrate/α-ketoisovalerate for production of methyl protonated perdeuterated proteins. (back-exchange of ND to NH, introduction of D decoupling) - Decrease transverse 15N1H relaxation rates through TROSY effect at high field: Selection of the slower relaxing component due to the destructive interference of cross-correlated relaxation (DD - CSA) Decoupled HSQC

Non-decoupled HSQC

TROSY 15 N

1H

- At high fields transverse 13CO relaxation rates increase due to CSA Use i-HNCA instead of HN(CO)CA

03002-p.20