design, principles and building blocks of heteronuclear

Design, principles and building blocks of Design, principles and building blocks of

heteronuclearheteronuclear NMR pulse sequencesNMR pulse sequences

Michael SattlerMichael Sattler

EMBL HeidelbergEMBL Heidelberg

GSFGSF--National Research Center for Environment and HealthNational Research Center for Environment and Health

Technical University Technical University MMüünchennchen

Prog. NMR Spectrosc. (1999) 34, 93-158.http://www.embl.de/nmr/sattler

EMBO Practical Course:EMBO Practical Course:

Structure determination of biological macromolecules by solutionStructure determination of biological macromolecules by solution NMRNMR

BiozentrumBiozentrum Basel, July 6Basel, July 6--13 200713 2007

ContentsContents

•• HeteronuclearHeteronuclear NMR: motivationNMR: motivation

•• Basic pulse sequence elements and 2D correlationsBasic pulse sequence elements and 2D correlations

•• RF pulses: calibration, selective pulsesRF pulses: calibration, selective pulses

•• Sensitivity enhancement, gradients, coherence selectionSensitivity enhancement, gradients, coherence selection

•• WaterWater--flip backflip back

•• SpinSpin--state selection, TROSYstate selection, TROSY

•• Triple resonance experimentsTriple resonance experiments

•• Isotope editing and filteringIsotope editing and filtering

Structure determination by NMRStructure determination by NMR

NOENOE--based assignment strategiesbased assignment strategies

11JJ-- and and 22JJ--couplings in proteinscouplings in proteins

H H| |

C C

Speed/efficiency of magnetization transferSpeed/efficiency of magnetization transfer

H H| |

C C

Homonuclearmagnetization transfer

via 3J coupling

Heteronuclearmagnetization transfer

via 1J couplings

10Hz

35Hz

140H

z

140H

z

100 ms100 ms 35 ms35 ms

S/NS/N ~ ~ NN γγexcexc γγdetdet3/23/2 BB00

3/23/2 NSNS1/21/2 TT22

Assignment based on JAssignment based on J--correlationscorrelations

MultiMulti--dimensional NMR experimentsdimensional NMR experiments

• To resolve signal overlap with increasing molecular weight

• 1/√2 loss of S/N per indirect dimension

• potential problems: time needed for sampling

ω2

ω3

ω1ΩΤ

ΩΤ’

ΩI

ΩS

ΩS’

t1 t2 t3preparation

I S T I

mixing mixing mixing detection

3D FT NMRφS φT

1

3

2

15N

1H

13C

15N

1H

Iz Iz

Ix,y Ix,y cosπΩ t ± Iy,x sinπΩ t

πΩ tChemical shiftChemical shift

PulsesPulses

9090°° pulsepulse

Scalar couplingScalar coupling

Iz Iz

IxSy IxSy

Ix,y Ix,y cosπJISt ± 2Iy,xSz sinπJISt

2Ix,ySz 2Ix,ySz cosπJISt ± Iy,x sinπJISt

πJIS2IzSz t

Iz Iz cosβ ± Ix,y sinβ

Ix,y Ix,ycosβ Iz sinβ

Ix,y Ix,y

+

Iz ± Ix,y

Ix,y Iz

90°y,x

β x,y

Pulse Pulse EductEduct ProductProductx y z & cyclic x y z & cyclic x z x z −−y y permutpermut..

Product operator formalismProduct operator formalism

+

Chemical Chemical shiftshift and and JJ--couplingcoupling evolutionevolution

HowHow to to analyzeanalyze pulse pulse sequencessequences

180o pulses can be combined if coherence order

magnitude is preserved inbetween

( i.e. no 90o pulse applied)

HeteronuclearHeteronuclear polarizationpolarization transfertransfer

S/NS/N ~ ~ NN γγexcexc γγdetdet3/23/2 BB00

3/23/2 NSNS1/21/2 TT22

Basic Basic heteronuclearheteronuclear correlationscorrelations: HMQC: HMQC

δδ((11H): refocusedH): refocused

δδ((1313C): evolution during tC): evolution during t11

J(J(11H,H,1313C): active during C): active during ΔΔ

J(H,H): active !J(H,H): active !

J(C,C): active !J(C,C): active !

Relaxation during tRelaxation during t11: multiple quantum line: multiple quantum line--narrowingnarrowing

Basic Basic heteronuclearheteronuclear correlationscorrelations: HSQC: HSQC

δδ((11H): refocusedH): refocused

δδ((1313C): evolution during tC): evolution during t11

J(J(11H,H,1313C): active during C): active during ΔΔ

J(H,H): not activeJ(H,H): not active

J(C,C): active !J(C,C): active !

Relaxation during tRelaxation during t11: T1(: T1(11H), T2(H), T2(1313C)C)

Basic building blocks: Basic building blocks: heteronuclearheteronuclear correlationcorrelation

Relaxation during tRelaxation during t11

T2MQ (IxSy) > T2(S)

Methyl TROSY

T1 I-spin (1H)

T2 S-spin (13C)

T2 S-spin (13C)

Bax et al JMR (1990) 86, 304-318

synchronous decoupling!

Transfer Transfer amplitudesamplitudes forfor antiphaseantiphase//inin--phasephaseconversionconversion in CH, CHin CH, CH22 and CHand CH33 spinspin systemssystems

-1

-0.5

0

0.5

1

0 1 2

1/(4J) 1/(2J)

3 4 5

CH

CH2

CH3

6 7 8 9

CH: 2HCH: 2HzzCCyy CCxx sin(sin(ππ11JJH,CH,C ΔΔ’’) )

CHCH22: 2H: 2HzzCCxx CCxx sin(sin(ππ11JJH,CH,C ΔΔ’’) cos() cos(ππ11JJH,CH,C ΔΔ’’) )

CHCH33: 2H: 2HzzCCxx CCxx sin(sin(ππ11JJH,CH,C ΔΔ’’) cos) cos22((ππ11JJH,CH,C ΔΔ’’) )

JJ--coupling evolution during thecoupling evolution during the

second half of a refocused INEPT:second half of a refocused INEPT:

ΔΔ’’ [ms][ms]

ConstantConstant--timetime HSQCHSQC

Set 2T = n/JSet 2T = n/JCCCC to refocus evolution of to refocus evolution of homonuclearhomonuclear C,C couplings during 2TC,C couplings during 2T

JJ--coupling evolution: cos(coupling evolution: cos(ππJJCCCC2T)2T)nn = = −−11nn

BIRD BIRD filterfilter to to suppresssuppress 1212C C magnetizationmagnetization

Excellent 12C suppression and fast acquisition (small, unlabeled molecules!)

Bax Subramanian JMR (1986) 67, 565-9Sattler et al JACS (1992) 114, 1126-7.

BIRD HSQC BIRD HSQC withwith multiplicitymultiplicity editingediting

J(H,C): refocusedJ(H,C): refocused

J(H,C): J(H,C): cos(cos(ππ))nn

Pulse calibration: considerationsPulse calibration: considerations

•• What sample to use for pulse calibration What sample to use for pulse calibration –– HH22O, protein, urea?O, protein, urea?

•• HH22O:O: ☺ large signal, very sensitive, optimize on FIDlarge signal, very sensitive, optimize on FID

radiation damping, different NMR properties than radiation damping, different NMR properties than biomoleculebiomolecule

•• protein:protein: ☺ NMR signals of interestNMR signals of interest

poor S/Npoor S/N

• urea: urea: ☺ high sensitivity, isolated signalshigh sensitivity, isolated signals

different NMR properties than actual sampledifferent NMR properties than actual sample

•• Calibration as 90Calibration as 90ºº, 180, 180ºº, 360, 360ºº pulse?pulse?

•• Problems: radiation damping, B1 Problems: radiation damping, B1 inhomogeneityinhomogeneity, amplifier power drop, amplifier power drop

•• Problems with Problems with cryoprobescryoprobes

•• salt concentration, B1 salt concentration, B1 inhomogeneityinhomogeneity, water suppression, water suppression

Pulse calibration: B1 Pulse calibration: B1 inhomogeneityinhomogeneity

•• Calibration of 180Calibration of 180°° or 360or 360°° pulsepulse

•• offoff--resonance effectsresonance effects

•• B1 B1 inhomogeneityinhomogeneity

Pulse calibration:

Center: [Z-shim detuned] Bulk [Z tuned] ERROR

90deg: 61us (expected) (app. 90)

180deg: 123us (2*90=122us) 132us (66.0us) 8%

360deg: 248us (4*90=244us) 257us (64.3us) 5%

Jerschow & Bodenhausen JMR (1999) 137, 108-115.

180180°°

360360°°

18001800°°

sample

NMR tube

11H H nutationnutation experimentexperiment

ZZ--shim detunedshim detuned0.1% Ethylbenzene

500 MHz Cryoprobe

180180°°

360360°°

Pulse calibrated on bulkPulse calibrated on bulk

sample

Pulse calibration: recommendationsPulse calibration: recommendations

•• calibrate 360calibrate 360°° pulse for pulse for 11HH

bulk pulse closer to bulk pulse closer to ““correctcorrect”” pulse at the center of pulse at the center of rfrf coilcoil

•• use use ShigemiShigemi (solvent susceptibility matched) tubes(solvent susceptibility matched) tubes

reduced effects of B1 reduced effects of B1 inhomogeneityinhomogeneity

reduced problems of higher salt concentrationsreduced problems of higher salt concentrations

•• optimize individual pulses in optimize individual pulses in multipulsemultipulse sequencessequences

i.e. p2 i.e. p2 ≠≠ p1*2, p1*2, ……

Pulse calibrationPulse calibration

ProblemProblem: no phase cycle : no phase cycle 11HH--1212C signals not suppressed!C signals not suppressed!

Phase cycling Phase cycling 11HH--1212C signals are suppressed!C signals are suppressed!

Experimental determination of relative RF phasesExperimental determination of relative RF phases

Why is this important?Why is this important?

RF at different RF power levels may have RF at different RF power levels may have

different relative phase. Such phase different relative phase. Such phase

shifts have to be compensated for in a shifts have to be compensated for in a

pulse sequence.pulse sequence.

How to determine the relative RF How to determine the relative RF

phase?phase?

For calibration, set For calibration, set ψψ = y; this should = y; this should

yield zero signal in the absence of phase yield zero signal in the absence of phase

shifts.shifts.

Adjusting the phase increment Adjusting the phase increment δδ to to

optimize for zero signal. This is then the optimize for zero signal. This is then the

required phase compensation.required phase compensation.

1D 1D 1313C spectrum of a proteinC spectrum of a protein

160.0 120.0 80.0 40.0 [ppm]δ13C

C' Carom. Caliph.

Selective RF pulses: rectangular pulsesSelective RF pulses: rectangular pulses

β = ω1∗ τp = −γB1∗ τp

Flip angle of RF pulse:ΔΩ

13Cα13C’

90° pulse:

βeff = 4∗β = 360° ⇒ (ω1eff)2 = (4ω1)

2 = ω12 + (ΔΩ)2 ⇒ ω1 = ΔΩ/√15

180° pulse:

βeff = 2∗β = 360° ⇒ (ω1eff)2 = (2ω1)

2 = ω12 + (ΔΩ)2 ⇒ ω1 = ΔΩ/√3

τp(90°) = β/ω1 = π/2 ∗ √15 / (2π∗ΔΩ)

τp(180°) = β/ω1 = π ∗ √3 / (2π∗ΔΩ)

BandBand--selective RF pulses: shaped pulsesselective RF pulses: shaped pulses

γB1rect/2π

τp τp

γB1max /2π

β = γB1rect ∗ τp

= γB1rect/2π ∗ τp ∗ 360°

β = γB1max ∗ ΣAi ∗ τp

= γB1rect / ΣAi ∗ τp

Rectangular pulse Shaped pulse

Σ (Ai ∗ τp/NP)

τp = pulse widthNP = number of points in shapeAi = relative intensity, 0…1

BandBand--selective RF pulses: rectangular vs. shaped selective RF pulses: rectangular vs. shaped pulsespulses

Mz

ΔΩ π/2 [kHz]

1.0

1.0

0

0.2

0.2

0.4

0.4

0.6

0.6

0.8

0.8

25 -2520 -2015 -1510 -105 -50

180rectangular

pulse

o

G3-pulse

C'

Caliph.

1.0

Mxy

ΔΩ π/2 [kHz]

0

0.2

0.4

0.6

0.8

25 -2520 -2015 -1510 -105 -50

90rectangular

pulse

o

G4-pulse

C'

Caliph.

Selective excitation Selective inversion

OffOff--resonance pulses: phaseresonance pulses: phase-- and amplitude modulationand amplitude modulation

Phase modulation:Phase modulation:

φφiimodmod = = φφii −− (2(2ππ ∗∗ ΔΩΔΩ ∗∗ττpp∗ ∗ i/NP)i/NP)

Amplitude modulation:Amplitude modulation:

AAiimodmod = A= Aii ∗∗ 2cos(22cos(2ππ ∗∗ ΔΩΔΩ ∗∗ττpp∗ ∗ i/NP)i/NP)

G3, 500G3, 500μμssno modulationno modulation

Mz

Mz

Mz

ΔΩ [Hz]

BandBand--selective RF pulsesselective RF pulses

GaussianGaussian(90(90°° / 270/ 270°°))

EE--BURPBURP

G4G4

Q5Q5

Selective inversion,Selective inversion,refocusing (180refocusing (180°°))

Selective excitationSelective excitation(90(90°°))

II--BURPBURP

RERE--BURPBURP

G3G3

Q3Q3

Adiabatic inversionAdiabatic inversion

HyperbolicHyperbolicsecantsecant

WURSTWURST

CHIRPCHIRP

References (shaped pulses and bandReferences (shaped pulses and band--selective decoupling):selective decoupling):

JMR (1991) 93, 93; Chem. Phys. Lett. (1990) 165, 469; JMR (1992) 97, 135;

JMR (1992) 100, 604; JMR (1993) A102, 364; JMR (1995) A115, 273; JMR (1996) A118, 299 .

shaped pulse: universal rotation

100

50

0

-50

[kHz]

[%]

2 1 0 -1 -2 -3

simulation

4.66 ms 2174.2 Hz 6.18 ms 742.1 Hz

[kHz] 2 1 0 -1 -2 -3

e-Burp2

Mz -> My

100

50

0

-50

[%]

G4

Q5u-Burp

7.82 ms 598.5 Hz4.94 ms 829.2 Hz

My -> Mz

time rev.My -> Mz

W. Bermel, Bruker

BandBand--selective RF pulses selective RF pulses -- adiabatic pulsesadiabatic pulses

G3, 250μs, γB1max/2π ≈ 14kHz

WURST, 2.7ms, γB1max/2π ≈ 3kHz

Mz

ΔΩ [Hz]

WURST

Amplitude Phase

•• AdiabaticAdiabatic””fastfast”” passage: keep magnetization and passage: keep magnetization and rfrf field field colinearcolinear: |: |ddζζ/dt/dt| | «« ωωeffeff

•• NonNon--linear pulse phase modulation linear pulse phase modulation frequency sweep: frequency sweep: ddΔωΔω/dt/dt

•• broadbroad--band inversion with low powerband inversion with low power

•• problem: problem: adiabaticityadiabaticity requires long requires long ττPP

ζ

BlochBloch--SiegertSiegert phase shifts (BSP)phase shifts (BSP)

Chem. Phys. Lett. (1990) 165, 469

Off-resonance: ΔΩ » ω1

Precession around ω1eff :

φoff-reson = ω1eff∗ τp

Instead of precessionaround z-axis with:φfree = ΔΩ∗ τp

ω = 2ω1 1effθ

ω1

ΔΩ = √3ω1

ΔΩ = 3√ ω1

Mx180°-pulse:τπ =π/ω1

0 −ΔΩ

φBSP(0) = −ΔΩ * τπ = −√3 ω1 * π/ω1 = −√3π ≈ 48.2°

1313CC’’ 1313CCαα

rotating coordinate system at :−ΔΩ

x

y

ΔΩ τπx

y

x

y

ΔΩ τπx

y

ΔΩ τπ

180°-pulse at −ΔΩ

360° on-resonance=̂

free evolution

((1313CCαα))

Bloch Siegert shift: inversion simulationMz -> Mz

[%]

100

50

0

-50

[kHz] 4 0 -4 -8

Q3 (2 msec, 1650.4Hz, freq: 2kHz, -3kHz)

W. Bermel, Bruker


[%]

100

50

0

-50

[kHz] 4 0 -4 -8


W. Bermel, Bruker

Bloch Siegert shift: inversion simulationphase

[°]

200

100

0

-100

[time] T/2 T

Q3 (2 msec, 1650.4Hz, freq: -3kHz)

without pulse, with Q3 pulse, difference

My (freq: 2kHz)

W. Bermel, Bruker

BSP compensationBSP compensation

ΔΩ/2π [kHz]

BSP [°]

120

-80

-60

-40

-20

0

20

40

60

80

100

10 12 14 16 18 20 22 24 26 28 30

Bloch-Siegert phasefor a band-selective(± 5kHz) G3 pulse

Determine phase empirically(0th and 1st order)

Use amplitude-modulatedinversion pulse (0th order)

Use additional pulse(all orders are corrected)

φ+BSP(0)

δCα

C'

G3, 250 s, 18 kHzphase modulation

μ

TT t12

t12

Intrinsic correction(all orders are corrected)

φδ'

G3, 700 s, 18 kHzamplitude modulation

μ

TT t12

t12

BSP

φTT t1

2t12 δCα

C'

TT t12

t12

G3 shape withadded BSP correction

φBSP = ω12/(2ΔΩ)*τP

Calculate φBSP for each subpulse

of a shaped pulse and add it to

the phase to compensate for the

BSP.

JMR (1992) 100, 604; JMR (2000) 146, 369.

Experimental determination of Bloch Experimental determination of Bloch SiegertSiegert PhasePhase

•• Zero order phase correction Zero order phase correction δδ00 applied to the phase of a flanking 90applied to the phase of a flanking 90oo pulsepulse

•• First order phase correction by addition of a delay First order phase correction by addition of a delay δδ11


[%]

100

50

-50

[kHz] 4 0 -4 -8


with BS compens.

without

W. Bermel, Bruker

BSP

φTT t1

2t12

ΔΩ/2π [kHz]

BSP [°]

120

-80

-60

-40

-20

0

20

40

60

80

100

10 12 14 16 18 20 22 24 26 28 30

Bloch-Siegert phasefor a band-selective(± 5kHz) G3 pulse

BSP compensation (intrinsic or add. pulse)BSP compensation (intrinsic or add. pulse)

GradientsGradients in in heteronuclearheteronuclear NMR NMR experimentsexperiments

x x

Gz

I

y

x

y

xa

b

y

xa

b

a

b

y

x

Spin echo with gradients

Δ2

Δ2

x x y

Gz

S

I

I: Iz

Iz

-Iy

-Iy 2I Sx z -2I Sz y2I Sz z2I Sz z

Iy Iy

I-S:

y

x

Spoil/purge gradients

Coherencerejection

Coherenceselection

Water Water suppressionsuppression methodsmethods

Concentration [Concentration [11H] in HH] in H22O O ≈≈ 110 M, concentration 110 M, concentration biomoleculebiomolecule ≈≈ 1010--33 MM

PROBLEMS: PROBLEMS: dynamic range (receiver); radiation dampingdynamic range (receiver); radiation damping

•• PresaturationPresaturation

depends on Bdepends on B00 homogeneity (shimming)homogeneity (shimming)

signals with near solvent frequency are suppressed as well (e.g.signals with near solvent frequency are suppressed as well (e.g. HHαα in proteins)in proteins)

reduces S/N of exchangeable protons due to saturation transferreduces S/N of exchangeable protons due to saturation transfer

•• JumpJump--andand--return / binominal sequencesreturn / binominal sequences☺ waterwater--flipflip--back intrinsic/possibleback intrinsic/possible

nonnon--optimum excitation profile optimum excitation profile

difficult to combine with triple resonance/multidifficult to combine with triple resonance/multi--pulse sequencespulse sequences

•• SpinSpin--lock, gradient spoil pulses, WATERGATElock, gradient spoil pulses, WATERGATE☺ can be combined with watercan be combined with water--flipflip--backback

suppression of signals near watersuppression of signals near water

•• HeteronuclearHeteronuclear gradient echoesgradient echoes☺ excellent waterexcellent water--suppression with sensitivity enhancement, combine with watersuppression with sensitivity enhancement, combine with water--flipflip--backback

•• PostPost--acquisitionacquisition

☺ apply lowapply low--pass filters to eliminate signals at 0 pass filters to eliminate signals at 0 ±± ωω, i.e. water on, i.e. water on--resonanceresonance

suppresses signal near water as wellsuppresses signal near water as well

RadiationRadiation dampingdamping

τp=20.5μs (<180°)

τp=20.8μs (>180°)

1H

FID

FID

detect

Δ [s]

Δ

0.050 0.100 0.150 0.200 0.250 0.300

Δ

1H

1H

Grad

Mxy

Warren, Hammes, Bates J.Chem. Phys (1989) 91, 5895Chen, Mao, Ye JMR (1997) 124, 490-494

WaterWater--flipflip--backback

1, 1, --1 Jump1 Jump--returnreturn

excitation null onexcitation null on--resonanceresonance

WATERGATE with waterWATERGATE with water--flipflip--backback

HSQC HSQC withwith WATERGATE & WATERGATE & waterwater--flipflip--backback

HH22O:O: zz --yy yy zz --zz zz zz

Suppress radiation damping Suppress radiation damping

of Hof H22O signal during tO signal during t11

waterwater--flipflip--backback

SensitivitySensitivity enhancementenhancement

•• Optimized coherence transfer, coherence order selective Optimized coherence transfer, coherence order selective

transferstransfers

•• sensitivity enhancement (coherence order selective coherence trasensitivity enhancement (coherence order selective coherence transfer)nsfer)

•• double sensitivity enhancementdouble sensitivity enhancement

•• TROSYTROSY

•• MQ lineMQ line--narrowing (methyl TROSY)narrowing (methyl TROSY)

•• Simultaneous acquisition, i.e. Simultaneous acquisition, i.e. 11HH--1313C, C, 11HH--1515N correlationN correlation

•• WaterWater--flipflip--back (for amides affected by solvent exchange)back (for amides affected by solvent exchange)

•• Longitudinal relaxation optimization (LHSQC, LTROSY)Longitudinal relaxation optimization (LHSQC, LTROSY)

•• Fast data acquisitionFast data acquisition

SensitivitySensitivity enhancementenhancement

RSH

amplitude modulation

Echo/anti-echo

Phase modulation

SensitivitySensitivity enhancementenhancement / / gradientgradient coherencecoherence selectionselection

Standard HSQCno gradient coherence selection

S/N = 1

Standard HSQCgradient coherence selection

S/N = 1/√2

Sensitivity enhanced HSQCgradient coherence selection

S/N = √2

S/N=: I/√n (I: Σ(intensities), n: # signals)

SensitivitySensitivity--enhancedenhanced HSQC HSQC withwith waterwater--flipflip--backback

HH22O:O: zz --yy yy zz --zz --yy yyyy --yy --zz zz

Suppress radiation damping Suppress radiation damping

of Hof H22O signal during tO signal during t11

waterwater--flipflip--backback

WaterWater--flipflip--backback

SE HSQCSE HSQC

2D: S/N*2D: S/N*√√22

WATERGATE HSQC WATERGATE HSQC

H2O: +zH2O: dephasedH2O: -z

H2O: +zH2O: dephasedH2O: -z

Longitudinal Longitudinal relaxationrelaxation optimizationoptimization

•• faster longitudinal relaxationfaster longitudinal relaxation

•• bandband--selective inversion of spins, i.e. amide protonsselective inversion of spins, i.e. amide protons

•• partially already achieved by waterpartially already achieved by water--flipflip--backback

•• faster repetition ratesfaster repetition rates

Pervushin et al JACS (2002)124, 12898Schanda & Brutscher JACS (2005), 127, 8014Attreya & Szyperski PNAS (2004) 101, 9624.Deschamps & Campbell JMR (2006) 178, 206

Double Double sensitivitysensitivity enhancementenhancement

Double sensitivity enhanced HCCHDouble sensitivity enhanced HCCH--TOCSYTOCSY

J. Biomol. NMR (1995) 6, 11-22.

•• can be recorded in Hcan be recorded in H22O due to excellent water suppression by O due to excellent water suppression by heteronuclearheteronuclear gradient echogradient echo

SimultaneousSimultaneous 1313C/C/1515N,N,11H HSQCH HSQC

JBN (1994) 4, 201-213; JMR (1994) B103, 197-201.

•• 11H H ↔↔ X transfer can be optimized simultaneously for X transfer can be optimized simultaneously for 1313C and C and 1515NNbut: some relaxation loss for but: some relaxation loss for 1313C due to longer delay.C due to longer delay.

•• poor waterpoor water--suppression, since E/AE cannot be implemented without sensitivitsuppression, since E/AE cannot be implemented without sensitivity loss.y loss.

•• building block for simultaneous 3D/4D NOESY experiments.building block for simultaneous 3D/4D NOESY experiments.

SimultaneousSimultaneous sensitivitysensitivity enhancementenhancement

SimultaneousSimultaneous sensitivitysensitivity enhancementenhancement

JBN (1995) 5, 97-102.

SimultaneousSimultaneous sensitivitysensitivity enhancedenhanced HSQC in HHSQC in H22OO

JBN (1995) 5, 97-102.

Basic building blocks: spinBasic building blocks: spin--statestate--selective filtersselective filters

JMR (1998) 131 373.

IP AP

ΔΣ


J. Biomol. NMR (1997) 10, 181; JMR (1997) 128, 92; J. Biomol. NMR (1998) 12, 435.

Short filter, but sensitive to JShort filter, but sensitive to J--mismatch!mismatch!


JBN (1997) 10, 181; JBN (1998) 12, 435.


JMR (1999) 139, 439.

Basic building blocks: TROSYBasic building blocks: TROSY

Experiments to Experiments to measuremeasure RDCsRDCs

Criteria to consider for measuring Criteria to consider for measuring RDCsRDCs

•• spinspin--statestate--selection, sensitivity to J+D variationsselection, sensitivity to J+D variations

•• ωω1 lines usually narrower 1 lines usually narrower measure splitting in measure splitting in ωω11

•• Large molecules: splitting from TROSY and Large molecules: splitting from TROSY and decdec. HSQC. HSQC Bax JMR (2000) 143, 184

•• Sensitivity and resolution (2D vs. 3D) Sensitivity and resolution (2D vs. 3D)

Pulse sequencesPulse sequences1H-15N splittings

•• IPAP combined with E.COSY for passive spinsIPAP combined with E.COSY for passive spins Bax JMR (1998) 131, 373

•• interleaved interleaved α/βα/β ((ωω1 or 1 or ωω2) TROSY2) TROSY Sorensen JBN (1999) 13, 175; JMR (1999) 140, 259;

Otting JBN (1998) 12, 435.

•• DSSE HSQC DSSE HSQC Grzesiek JBN (1999) 13, 175.

•• QuantitativeQuantitative--J J Prestegard JMR (1996) B112, 245; Bodenhausen JBN (2002) 23, 195.

HN-N, HN-C’, N-C’, C’-Cα, HN-Cα splittings:

•• IPAP combined with E.COSY for passive spinsIPAP combined with E.COSY for passive spins Bax JACS (1998) 120, 7385

•• 3D HNCO based E.COSY experiments3D HNCO based E.COSY experiments Kay JBN (1999) 14, 333; JBN (1998) 12, 325.

•• Methyl Methyl 11HH--11H, H, 11HH--1313CC Otting JACS (2001) 123, 1770; Brutscher JACS (2002) 124, 14616

•• Aromatic Aromatic 1313CC--1313C, C, 11HH--1313CC Sattler JMB (2003) 327, 507

MultiMulti--dimensionaldimensional NMR experimentsNMR experiments• To resolve signal overlap with increasing molecular weight• 1/√2 loss of S/N per indirect dimension, but increased resolution

ω2

ω3

ω1ΩΤ

ΩΤ’

ΩI

ΩS

ΩS’t1 t2 t3preparation

I S T I

mixing mixing mixing detection

3D FT NMRφS φT

1

3

2

15N

1H

13C

15N

1H

sin2(π 1JHα,Cα Δ') exp(−Δ'/T2Hα)

* sin(π 1JCα,N 2τ1) cos(π 2JCα,N 2τ1) cos(π 1JCα,Cβ 2τ1) exp(−2τ1/T2Cα)

* sin(π 1JCα,N 2τ2) cos(π 2JCα,N 2τ2) exp(−2τ2/T2N)

* sin2(π 1JN,HNΔ) exp(-Δ/T2HN)

Transfer amplitude Transfer amplitude including Tincluding T22 relaxationrelaxation

sin4(π 1JN,HN Δ) exp(-2Δ/T2HN)

* sin2(π 1JCα,N 2τ) cos2(π 2JCα,N 2τ) exp(−4τ/T2N)

* cos(π 1JCα,Cβ t2) exp(−t2/T2Cα)

OutOut--andand--back vs. transfer experimentsback vs. transfer experiments

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

0 2τ12 /3τ1

2τ1-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

0

t1 evolution:

∫(cos(πJ2τ1) exp(−2τ1/T2))dt1

∫(cos(πJt1) exp(-t1/T2))dt1

Constant Time (CT) vs. Real Time (RT) evolutionConstant Time (CT) vs. Real Time (RT) evolution

0 2 +tτ2 2max2τ2

0

0.2

0.4

0.6

0.8

1

2 +tτ2 2max2τ20

0

0.2

0.4

0.6

0.8

1

t2 evolution:

HNCO HNCO –– purging of dispersive purging of dispersive lineshapelineshape

JBN (1999) 14, 273-276.

• 2J(HN,C’) are ~ 4Hz

• If 2τ is not matched to 1/(21JN,C’) dispersive lineshape in F2 or F3 are observed

• Dispersive signals due to 2J(HN,C’) couplings can be removed

i) by a 90 deg C’ pulse or ii) by 13C’ decoupling during acquisition

H2O: z -y y z -x -y yy -y -z zz

2NyC’z2HzNy sin(π1JN,C’2τ) cosΩNt2 cos(π2JHN,C’t2) cosΩHNt2

4HzNyC’z cos(π1JN,C’2τ) sinΩNt2 sin(π2JHN,C’t2) cosΩHNt2

HN(CO)CAHN(CO)CA

HN N C’ Cα (t1) C’(t2) N(t2) H(t3)

H(N)COCA H(N)COCA -- CO chemical shift evolutionCO chemical shift evolution

HN(CA)COHN(CA)CO

HN N Cα C’(t2) Cα N(t2) H(t3)

HN(CA)COHN(CA)CO

Basic triple resonance experiments: CBCA(CO)NHBasic triple resonance experiments: CBCA(CO)NH

Transfer amplitude in a 13C spin system Cα−(Cβ)m−(Cγ)p

f(T,ζ) = cosm(2π 1JCα,Cβ T) cosm(2π 1JCα,Cβ ζ)

f(T,ζ) = sin(2π 1JCα,Cβ T) cosp(2π 1JCβ,Cγ T) sin(2π 1JCα,Cβ ζ)

Cα(i-1):

Cβ(i-1):

2 4 6 8

2T [ms]10 12 140

C (i-1)α

C (i-1)β

2 [ms]ζ2 4 60 8 1210 14 16 18

C (i-1)β

C (i-1)α

Basic triple resonance experiments: HBHA(CO)NHBasic triple resonance experiments: HBHA(CO)NH

CBCANH transfer amplitudesCBCANH transfer amplitudes

cosm(2π 1JCα,Cβ T)

cosm(2π 1JCα,Cβ τ1) sin(2π 1JCα,N τ1) cos(2π 2JCα,N τ1)

sin(2π 1JCα,N τ) cos(2π 2JCα,N τ)

cosm(2π 1JCα,Cβ T)

cosm(2π 1JCα,Cβ τ1) sin(2π 2JCα,N τ1) cos(2π 1JCα,N τ1)


sin(2π 1JCα,Cβ T) cosp(2π 1JCβ,Cγ T)

sin(2π 1JCα,Cβ τ1) sin(2π 1JCα,N τ1) cos(2π 2JCα,N τ1)


sin(2π 1JCα,Cβ T) cosp(2π 1JCβ,Cγ T)

sin(2π 1JCα,Cβ τ1) sin(2π 2JCα,N τ1) cos(2π 1JCα,N τ1)


f(T, τ1, τ) in spin systemCα−(Cβ)m−(Cγ)p

Cα(i):

Cα(i-1):

Cβ(i):

Cβ(i-1):

2T 2τ1 2τ

-0.015

-0.01

-0.005

0

0.005

0.01

0.015

2 [ms]τ1

5 10 150 20 25 30 35 40

C (i-1)β

C (i)β

C (i-1)α

C (i)α

Basic triple resonance experiments: CBCANHBasic triple resonance experiments: CBCANH

H2O: z -y y y x -y yy -y -z zz

x y”

Water flipback without water-selective shaped pulse

X

Basic triple resonance experiments: HNCACBBasic triple resonance experiments: HNCACB

Basic triple resonance experiments: (H)CCCONH Basic triple resonance experiments: (H)CCCONH TOCSYTOCSY

Basic triple resonance experiments: (H)CCCONH Basic triple resonance experiments: (H)CCCONH TOCSYTOCSY

Basic triple resonance experiments: H(CC)CONH Basic triple resonance experiments: H(CC)CONH TOCSYTOCSY

• using homo- and heteronuclear TOCSY transfer

• time-saving by simultaneous 1H-13C and 13C-13C transfer

HCCHHCCH--TOCSYTOCSY

H(t1) C(t2) C C H(t3)

• usually recorded in D2O for Hα detection

SemiSemi--constantconstant--time chemical shift evolutiontime chemical shift evolution

t12

t12 t1a t1b t1c2

Δ2Δ

I

S

I

Sreal time semi-constant- time

δ: t1a + t1b − t1c = t1

J: t1a − t1b + t1c= Δ (1)

With the requirement that t1(0) = 0, Eq. (1) yields:

t1a(0) = t1c(0) = Δ/2 and t1b(0) = 0. (2)

For a FID along t1 that is to be digitized by TD data points, two relations can be written for the

increments Δt1a, Δt1b and Δt1c (Δt1=1/SWH, where SWH is the spectral width in Hz):

δ: Δt1a + Δt1b − Δt1c = Δt1

J: Δt1a − Δt1b + Δt1c= 0 (3)

Since it is required that t1c(TD) ≥ 0;. this yields (for t1c(TD) = 0):

Δt1a = Δt1/2; Δt1b = Δt1/2 + Δt1c and Δt1c = −t1c(0)/TD (4)

Fulfilling Eqs. (2) and (4) assures chemical shift evolution with t1(0) = 0 < t1 < t1max and

evolution of the coupling during Δ for all increments. Note, that Δt1c is negative, reflecting the fact t1c is decremented.

JJ--coupling coupling =: =: ΔΔ

Chemical shift Chemical shift δδ =: t=: t11

Double sensitivity enhanced HCCHDouble sensitivity enhanced HCCH--TOCSYTOCSY

J. Biomol. NMR (1995) 6, 11-22.

Can be recorded in HCan be recorded in H22O due to excellent water suppression by O due to excellent water suppression by heteronuclearheteronuclear gradient echogradient echo

HBHDHE (aromatic side chain assignments)HBHDHE (aromatic side chain assignments)

JACS (1995) 115, 11054

Cβ

H

γδ ε

δ ε

H H

H H

H

H

Cα

•• Delay Delay ξξ for for 1,1,--1 1801 180o o 1313C pulse to decouple C pulse to decouple 11J(CJ(Cαα,C,Cββ) )

•• 1313C pulses are selective with excitation nulls at C pulses are selective with excitation nulls at 1313CCββ and and 1313CCarar, respectively, respectively

Relative sensitivity of triple resonance experimentsRelative sensitivity of triple resonance experiments

Experiment Assignment Comment RelativeS/N [%]

HNCO H(i), N(i), C’(i-1) <20 kD, above use 2H labeling 100

HNCA H(i), N(i), Cα(i),Cα(i-1) <20 kD, above use 2H labeling 50/15

HN(CO)CA H(i), N(i), Cα(i-1) <20 kD, above use 2H labeling 71

HN(CA)CO H(i), N(i), C’(i) <20 kD, above use 2H labeling 13/4

CBCA(CO)NH H(i), N(i), Cα(i-1), Cβ(i-1) <20 kD, above use 2H labeling 13/9 α/β

HBHA(CO)NH H(i), N(i), Hα(i-1), Hβ(i-1) <20 kD, above use 2H labeling 13/9 α/β

CBCANH,HNCACB

H(i), N(i), Cα(i), Cβ(i),

Cα(i-1),Cβ(i-1)<15 kD, above use 2H labeling 4/1.7 α/β(i)

1.3/0.5α/β(i-1)

(H)CC(CO)NH-TOCSY

H(i), N(i), Caliph.(i-1) <15-20 kD, above use 2H labeling

H(CC)(CO)NH-TOCSY

H(i), N(i), Haliph.(i-1) <15-20 kD, above use 2H labeling

HCCH-TOCSY Haliph., Caliph. <25 kD, - sensitive, but tedious to analyze,combine with HCCONH type experiments

Assignment based on JAssignment based on J--correlationscorrelations

In a uniformly 13C/15N-labeled protein numerous chemical shifts can be measured and correlated via 1J and 2J-couplings

Backbone assignment

Side chain assignment

Sattler et al. Prog. NMR Spectrosc. (1999) 34, 93-158.

Assignment strategy for Assignment strategy for 1313C/C/1515NN--labeled proteinslabeled proteins

Cβ

Cα

CBCANHintra-/interresidue correlation

(strong/weak)

CBCA(CO)NHinterresidue correlation

HN/N of residue

Cβ (i)

Cβ (i-1)

Cα (i)

Cα (i-1)

i (Gly)i-1i-2 i+1 i+2

HN

HN

NN

O

CαCα

Cβ

O Cβ

HN

N

O

Cα

Cβ HN

HN

NN

O

CαCα

Cβ

O

HeteronuclearHeteronuclear NOESY experimentsNOESY experiments

• water-flip-back during NOE mixing time due to radiation damping

• φ1 = 45o to enhance water flipback independent of TPPI on φ1

HeteronuclearHeteronuclear NOESY experimentsNOESY experiments

•• HMQC rather than HSQC for HMQC rather than HSQC for 1313C C edited NOESY to minimize offedited NOESY to minimize off--resonance effects for resonance effects for 1313C C rfrf pulsespulses

•• Alternatively use adiabatic Alternatively use adiabatic inversion pulses, i.e. at higher fieldsinversion pulses, i.e. at higher fields

Isotope edited/filtered experiments: single filterIsotope edited/filtered experiments: single filter

1H-12C and 1H-13C can be separated into subspectra

add ψ = +/- x → 1H-12C filteredsubtract ψ = +/- x → 1H-13C edited

Single J-filter: Δ = 1/(2J)

filter efficiency:residual magnetization = cos(πJΔ)

add a + b → 1H-12C filteredsubtract a − b → 1H-13C edited

13C,15N

12C,14N

13C,15N

12C,14N

13C,15N

12C,14Nediting filtering

editing/filtering

Double isotope filterDouble isotope filter

Double JDouble J--filter:filter:ΔΔ’’ = 1/(2J= 1/(2J’’))ΔΔ’’’’ = 1/(2J= 1/(2J””))

To use as a filtered/edited pulse sequence:To use as a filtered/edited pulse sequence:ΔΔ’’ ++ ΔΔ”” = 1/(J) = 1/(J) (single J(single J--filter)filter)ΔΔ’’’’−− ΔΔ’’ = 0= 0

((11HH--1212C and C and 11HH--1313C can be separated into C can be separated into subspectrasubspectra))

Double JDouble J--filter:filter:ΔΔ’’ = 1/(2J= 1/(2J’’))ΔΔ’’’’ = 1/(2J= 1/(2J””))

Double filter:Double filter:cos(cos(ππJJΔΔ’’) ) cos(cos(ππJJΔΔ””))

Single filter:Single filter:cos(cos(ππJ(J(ΔΔ’’++ΔΔ””)/2))/2)

J [Hz]J [Hz]

Res

idu

al m

agn

etiz

atio

nR

esid

ual

mag

net

izat

ion

Isotope filters employing adiabatic frequency Isotope filters employing adiabatic frequency sweepssweeps

• dJCH/dν = 40 Hz/120 ppm (1J(Car-Har) = 160 Hz, 1JCα-Hα = 140 Hz, 1JCH3 = 120 Hz)

• use frequency sweep for adiabatic 180 deg 13C inversion pulse in filter-element

Frequency sweep:dν/dt

Δ Δ

130ppm(J=160Hz)

55ppm(J=140Hz)

10ppm(J=120Hz)

Inversion of 13C spins at:

JACS (1997) 119, 6711

Isotope filtered 2D NOESYIsotope filtered 2D NOESY

1H-[12C]and

1H-[13C]

1H-[12C]

Intermolecular

NOEs

1D (protein + RNA)

1D filter experiment(RNA only)

Isotope filtered NOESYIsotope filtered NOESY

1D(protein + RNA)

1D filter experiment(RNA only)

Isotope edited/filtered Isotope edited/filtered 1313C HMQCC HMQC--NOESYNOESY

Isotope edited/filtered NOESYIsotope edited/filtered NOESY

Isotope filtered NOESYIsotope filtered NOESY

Problems with higher molecular weightsProblems with higher molecular weights

• slower tumbling in solution fast decay of NMR signal poor signal-to-noise

• larger number of signals signal overlap in NMR spectra

τc 4 nsMW 8 kDa

8 ns16 kDa

12 ns24 kDa

25 ns50 kDa

linewidth Δν1/2 = 1/πT2

78910 ppm78910 ppm78910 ppm78910 ppm

Transverse relaxation

optimized spectroscopy

TROSY and TROSY and 22HH--labeling for molecular weights > 50 labeling for molecular weights > 50 kDakDa

2H-labeling

• reduced relaxation (γD / γH ~ 1 / 6.5)

• improved signal-to-noise

• better resolution

• reduced number of cross peaks

• suppression of spin diffusion

•

13C

1H

dipole/dipolerelaxation

N

H

D D

DD

N

H

H H

HH

Pervushin et al. PNAS (1997) 94, 12366-71.

ConstantConstant--time HNCA with time HNCA with 22HH--decouplingdecoupling

H HN

HN

NN

HO

D HN

HN

NN

DO

75 % randomfractional 2H-

labeling

no 2H-labeling

Effect of Effect of deuterationdeuteration in 3D HNCA experimentsin 3D HNCA experiments

Larger proteins: Larger proteins: 22HH--labeling + TROSYlabeling + TROSY

JACS (1999) 121, 844; JACS (1999) 121, 2571.

TROSY HNCA with TROSY HNCA with 22H decoupling and H decoupling and 11H purgeH purge

1H purgeno purge

• CT-HNCA not sensitive enough

• real time HNCA needs 1H purge

Identification of Identification of interresidualinterresidual cross peaks in HNCA TROSYcross peaks in HNCA TROSY

JBN (2001) 20, 127-133

• C’ CSA relaxation increases with B0 field strength (JMR (1999) 141, 180-184), HN(CO)CA is insensitive

Add: Subtract:

Cα(i)

1J(Cα,C’)

Cα(i-1)

• shifted peak position in-phase and antiphase spectra of the Cα(i-1) peak by 1J(Cα,C’)

allows unique identification of the sequential peak in this HNCA experiment:

1J(Cα,C’)However, two signals per residue in HNCA introduces

additional undesirable signal overlap intra HNCA:

JMR (2002) 156, 155-159; JACS (2002) 124, 11199-11207;

JBN (2002), 201-209.

design, principles and building blocks of heteronuclear

Documents