Deuxième école RMN du GERM Cargèse 2008

Dunker et al., Journal of Molecular Graphics and Modelling 19 .... catalytic processes. Regulation ... NMR timescales -2- Spectral or “chemical shift” timescale.
9MB taille 1 téléchargements 26 vues
Deuxième école RMN du GERM Cargèse 2008 Analysis of micro-millisecond timescale motions of macromolecules in liquid state by Nuclear Magnetic Resonance

Carine van Heijenoort ICSN laboratoire de chimie et biologie structurale [email protected] 0169823794

A view of proteins multiple states « lock & key » induced fit

3D structure α↔β

conformational switch

order → disorder

virus/pathogen penetration membrane insertion Nucléosome activation

Folding

Sequence

“Non folding” flexible linkers display of sites entropic bristles, springs and clocks

flexible ensemble disorder → order

Deuxième école RMN du GERM, Cargèse 2008

molecular recognition virus/phages assembly stepping motors Dunker et al., Journal of Molecular Graphics and Modelling 19, 26–59, 2001 Dobson, C., Nature 426, 18-25, 2003

Techniques to monitor protein dynamics time scales static disorder, crystal contacts, ...

X-ray cristallography

B factors

X-Ray, neutron scattering

size/shape modifications

Doniach, Chem. Rev. 2001, 101 ; Zacai, timescales (ps-ns) for 1H positions science 2000, 288.

Fluorescence

ensemble / single molecule

Weiss, Nat. Struct. Biol. 2000, 7 ; Yang, Science 2003, 302 ; Haustein, Curr. Opin. cellular context Struct. Biol. 2004, 14.

Mass Spectroscopy (HX MS) radical footprinting Wales, Mass. Spectrom. Rev. 2006, 25 ; Busenlehner, Arch. Biochem. Biophys. 2005, 433. Guan, Trends. Biochem. Sci. 2005, 30.

probes

large moldecular assemblies

Mössbauer, Raman, 2D infrared spectroscopy Forcefields ... Short timescales ...

Molecular dynamics ➫ 10-12↔ 105 s ➫ Site-specific information ➫ multiple atomic probes 1H, 2H, 15

NMR Boehr, Chem. Rev. 2006, 106,3055. Palmer, Chem. Rev. 2004, 104, 3623.

N,

13

C,

31

P, ...

➫ Simultaneous monitoring of

probes ➫ kinetic & termodynamic profile of dynamic processes

➫ ➫ ➫ ➫

isotope labeling quantities size limitation complexity of the method ?

Deuxième école RMN du GERM, Cargèse 2008

Bibliography Theoretical NMR basis : A. Carrington et A. Mc Lachlan Introduction to Magnetic Resonance with appplications to chemistry and chemical physics, Harper International , 1967 C. Slichter Principle of Magnetic Resonance , 3 ème ed, 1990, Springer Verlag A. Abragam Principles of Magnetic Resonance, Oxford University Press, 1961 Ernst, Bodenhausen, Wokaun Principle of Nuclear Magnetic Resonance in one and two dimensions, Oxford Science 1987 M. Goldman* Quantum description of high resolution NMR in liquids. Oxford 1990 D. Canet* La RMN : concepts, méthodes et applications (2ème ed.), UniverSciences, Dunod 2002 M. H. Levitt*** Spin dynamics. Basic of Nuclear Magnetic Resonance (686 pages) J.Wiley 2001 John Cavanagh, Wayne J. Fairbrother, Arthur G., III Palmer*** Protein NMR Spectroscopy: Principles And Practice (912 pages) Academic Press; Édition : 2nd (novembre 2006) Experimental aspects, applications J. K. M. Sanders, B. K. Hunter* Modern NMR Spectroscopy: A Guide for Chemists 2nd Edition, Oxford 2003 D. Shaw Fourier transform NMR spectroscopy Elsevier 1984 K. Wüthrich* NMR of proteins and nucleic acids. Wiley interscience 1986 Deuxième école RMN du GERM, Cargèse 2008

How motions are visible in NMR ?

Molecular motions influence NMR parameters Time scale of motions compared to NMR characteristic timescales ➫ Three characteristic NMR timescales Return of the spin system to equilibrium : T1 / NMR signal lifetime : T2 NMR experiment spins system perturbation Liquid state: T1 ~ 100ms s ; T2 ~ 10ms s ➫ Minimal frequencies of motions that can be characterized during a single NMR experiment.

Spectral timescale : τ=1/Δν spectral features : chemical shifts range, couplings, … Hz

kHz

➫ Averaging of interactions by motions faster than the spectral dispersion due to these interactions. ➫ Perturbation of spectral features by motions in the same range than the spectral timescale.

Larmor timescale : ω0 precession frequency of spins in a magnetic field (B0=ω0/γ) ➫ transitions efficiency between spin states is determined by molecular fluctuations (spectral densities fo motions) at these frequencies.

Deuxième école RMN du GERM, Cargèse 2008

How motions are visible in NMR ? Molecular motions influence NMR parameters Time scale of motions compared to NMR characteristic timescales ➫ Three characteristic NMR timescales Return of the spin system to equilibrium : T1 / NMR signal lifetime : T2 Spectral timescale : τ=1/Δν Larmor timescale : ω0

s

ms

T1, T2 Signal lifetime

µs

1/2pΔω

frequencies differences

ns

1/ω0

nutation frequencies

Deuxième école RMN du GERM, Cargèse 2008

ps

fs

Motions & function & NMR disorder Interactions catalytic processes Regulation, signalization

Functions

Folding ; order

Effect of motions on spins interaction

Types of motions

s

NMR parameters

ms

µs

Internal motions

ns

Macroscopic diffusion

ps

fs

Molecular vibrations

Conformational exchange

Molecular rotations

Diffusion experiments Magnetization exchange Lineshape modifications

Spin relaxation

Averaging of spectral components

T1, T2 Signal lifetime

1/Δω frequencies differences

1/ω0 nutation frequencies

Deuxième école RMN du GERM, Cargèse 2008

Secular interactions averaging

NMR timescales -1- Longitudinal relaxation time T1

Dynamical processes slower than T1

✓T1 characterizes the time for a spin

Real time folding Transient / out of equilibrium experiments Protein - ligand interactions H-N D-N exchange rate constants

system to reach equilibrium.

✓Defines interscan delay ✓Motions slower than T1 cannot be

characterize by a single spectra.

✓NB. T1 depends on magnetic field

B0, nature of spins, molecule size, local flexibility, temperature, etc.

Wüthrich, « NMR of proteins and nucleic acids », Wiley Interscience, 1986

Deuxième école RMN du GERM, Cargèse 2008

NMR timescales -1- Longitudinal relaxation time T1

Dynamical processes slower than T1

✓T1 characterizes the time for a spin

Real time folding kinetic / out of equilibrium experiments Protein - ligand interactions H-N D-N exchange rate constants

system to reach equilibrium.

✓Defines interscan delay

holoP 1 + apoP 2

✓Motions slower than T1 cannot be

characterize by a single spectra.

✓NB. T1 depends on magnetic field

B0, nature of spins, molecule size, local flexibility, temperature, etc.

Deuxième école RMN du GERM, Cargèse 2008

! apoP 1 + holoP 2

NMR timescales -1- Longitudinal relaxation time T1

magnetic field, molecular size and correlation time dependance of T1 and T2

✓T1 characterizes the time for a spin system to reach equilibrium.

✓Defines interscan delay

Relaxation des azotes 15 amides à 500MHz (−) et 800MHz (−−)

✓Motions slower than T1 cannot be

2

10

characterize by a single spectra. B0, nature of spins, molecule size, local flexibility, temperature, etc.

T1,2 (s)

✓NB. T1 depends on magnetic field

1

T1

10

0

10

!1

T2

10

!2

10

!3

10

!2

10

Deuxième école RMN du GERM, Cargèse 2008

!1

10

!c (ns)

0

10

1

10

NMR timescales -2- Spectral or “chemical shift” timescale

✓Defined

by the observed spectral width

✓More

precisely by the difference of resonance frequency between two spins from two different nuclei or from the same nuclei in two different states.

( )

Δν Hz =

( ) ∗ 10

Δω ppm 2π

τ spect = €



−6

B0=14.1T

Δν=7800Hz τspect~100µs

Δω=13ppm ∗ γB 0

B0=18.8T

Δν=10400Hz

B0=14.1T

Δν=120Hz

1 Δν(Hz )

Motions slower than Δν have no effect on spectral€feature

✓Δν depends on magnetic field and on spins nature

τspect~5-10ms

Δω=0.2ppm B0=18.8T

Deuxième école RMN du GERM, Cargèse 2008

Δν=160Hz

NMR timescales -2- Spectral or “chemical shift” timescale

✓Defined

by the observed spectral width

✓More

precisely by the difference of resonance frequency between two spins from two different nuclei or from the same nuclei in two different states.

( )

Δν Hz =

( ) ∗ 10

Δω ppm 2π

τ spect = €



−6

∗ γB 0

1 Δν(Hz )

15

Δω( N)=25ppm

B0=14.1T

τspect~0.7ms

Motions slower than Δν have no effect on spectral€feature

✓Δν depends on magnetic field and on spins nature

Δω(1H)=3.5ppm

Deuxième école RMN du GERM, Cargèse 2008

Δν=1500Hz

B0=14.1T

Δν=2000Hz τspect~0.5ms

NMR timescales -2- Spectral or “chemical shift” timescale Same nucleus ; two different conformations

✓Defined

by the observed spectral width

✓More

precisely by the difference of resonance frequency between two spins from two different nuclei or from the same nuclei in two different states.

( )

Δν Hz =

( ) ∗ 10

Δω ppm 2π

τ spect = €



−6

∗ γB 0

1 Δν(Hz )

Motions slower than Δν have no effect on spectral€feature

Δω(15N)=1.5ppm

B0=14.1T

τspect~10ms

✓Δν depends on magnetic field and on spins nature Δω(1H)=0.5ppm

Deuxième école RMN du GERM, Cargèse 2008

Δν=100Hz

B0=14.1T

Δν=300Hz τspect~3ms

NMR timescales -2- Spectral or “chemical shift” timescale Averaging of secular interactions by motions

✓Defined

✓ Dipolar interaction between spins magnetic moments.

by the observed spectral width

µ γ γ h  3cos 2 θ IS −1 ˆ ˆ ˆ ˆ DD Hˆ IS = − 0 I 3S   3 I z S z − I.S 2 4 πrIS  

✓More

precisely by the difference of resonance frequency between two spins from two different nuclei or from the same nuclei in two different states.

( )

Δν Hz =

( ) ∗ 10

Δω ppm





−6

✓ Chemical shift anisotropy

c −c Hˆ ICSA = −γ S || ⊥ B0 3cos 2 θ −1 Iˆz 3

∗ γB 0

(

1 Δν(Hz )

)

E h < 104 −10 5 Hz €

Motions slower than Δν have no effect on spectral€feature

depends on the nature of the interactions between spins

✓ Interaction spin quadrupole/electric field

(

)

Hˆ IQ ≅ ω IQ 3 Iˆz2 − Iˆ.Iˆ ; ω IQ =

✓Δν depends on magnetic field and on spins nature ✓Δν

)

E h < 104 −10 5 Hz €



τ spect =

(

I>

1 2

3eQI VzzI (θ ) 4 I (2I −1)

E h ≈ 2.10 5 Hz 2H and 3.10 6 14 N

€ ✓ Unpaired electron

Deuxième école RMN du GERM, Cargèse 2008

( )

( )

NMR timescales -2- Spectral or “chemical shift” timescale Averaging of secular interactions by motions

✓Defined

by the observed spectral width

soluble protein

✓More

precisely by the difference of resonance frequency between two spins from two different nuclei or from the same nuclei in two different states.

( )

Δν Hz =

( ) ∗ 10

Δω ppm 2π

τ spect = €



−6

∗ γB 0

1 Δν(Hz )

membrane protein

Motions slower than Δν have no effect on spectral€feature

✓Δν depends on magnetic field and on spins nature ✓Δν

depends on the nature of the interactions between spins

Deuxième école RMN du GERM, Cargèse 2008

NMR timescales -2- Spectral or “chemical shift” timescale Averaging of secular interactions by motions

✓Defined

by the observed spectral width

✓More

precisely by the difference of resonance frequency between two spins from two different nuclei or from the same nuclei in two different states.

( )

Δν Hz =

( ) ∗ 10

Δω ppm 2π

τ spect = €



−6

∗ γB 0

1 Δν(Hz )

Motions slower than Δν have no effect on spectral€feature

✓Δν depends on magnetic field and on spins nature ✓Δν

depends on the nature of the interactions between spins

Deuxième école RMN du GERM, Cargèse 2008

NMR timescales -2- Spectral or “chemical shift” timescale Averaging of secular interactions by motions

✓Defined

by the observed spectral width

✓More

precisely by the difference of resonance frequency between two spins from two different nuclei or from the same nuclei in two different states.

( )

Δν Hz =

( ) ∗ 10

Δω ppm 2π

τ spect = €



−6

Averaging of chemical shift anisotropy (31P)

∗ γB 0

1 Δν(Hz )

Motions slower than Δν have no effect on spectral€feature

✓Δν depends on magnetic field and on spins nature ✓Δν

depends on the nature of the interactions between spins

Burnell et al., Biochim. Biophys. Acta 603, 63 (1980)

Deuxième école RMN du GERM, Cargèse 2008

NMR timescales -2- Spectral or “chemical shift” timescale motions generated to average interactions by the observed spectral width

precisely by the difference of resonance frequency between two spins from two different nuclei or from the same nuclei in two different states.

( )

Δν Hz =

( ) ∗ 10

Δω ppm 2π

τ spect = €



−6

∗ γB 0

1 Δν(Hz )

Motions slower than Δν have no effect on spectral€feature

Poudre de Glycine

Lipides + eau

Statique

✓More

Rotation à l’angle magique ωr=5000Hz

✓Defined

8

4

0

8

4

0

1H

(kHz)

2

1

0

2

1

0

1H

(kHz)

✓Δν depends on magnetic field and on spins nature ✓Δν

depends on the nature of the interactions between spins Davis, Auger & Hodges Biophysical Journal 69:1917-1932 (1995) Gross et al., J. Magn. Res. 106, 187-190 (1995) Carlotti, Aussenac & Dufourc Biochim. Biophys. Acta. 1564:156-164 (2002)

Deuxième école RMN du GERM, Cargèse 2008

NMR timescales -3- The “Larmor” timescale

✓Defined

by the resonance frequencies of spins

✓i.e.

the energy difference between spins states levels

ω0 τLarmor ✓

= −γB0 1 1 = = |ω0 | 2πν0

ββ ωI

in these timescales are responsible for the efficiency of spins relaxation processes

βα

ωI -ωS

αβ

ωI

ωS αα

Motions in these timescale have no direct effect on spectra.

✓Motions

ωI+ωS

ωS

B0=14,1T

✓ The

relationship between motions and relaxation rate/time constants is not simple

Deuxième école RMN du GERM, Cargèse 2008

|ωI|=2π.600MHz ; τL(I)=265ps |ωS|=2π.60MHz ; τL(S)=2,65ns

the energy difference between spins states levels

ω0 τLarmor ✓

= −γB0 1 1 = = |ω0 | 2πν0

Motions in these timescale have no direct effect on spectra.

✓Motions

in these timescales are responsible for the efficiency of spins relaxation processes

✓ The

relationship between motions and relaxation rate/time constants is not simple

1H

✓i.e.

relaxation times (s)

by the resonance frequencies of spins

15N

✓Defined

relaxation times (s)

NMR timescales -3- The “Larmor” timescale

100 10

T1

1 0,1 0,01 10-12

400,600,800MHz

T2

400,600,800MHz

10-11

10-10

τc(s)

10-9

10-8

B0=14.1Teslas

100 10

T1

1 0,1 0,01 10-12

T2 10-11

Deuxième école RMN du GERM, Cargèse 2008

10-10

τc(s)

10-9

10-8

Motions & function & NMR disorder Interactions catalytic processes Regulation, signalization

Functions

Folding ; order

Effect of motions on spins interaction

Types of motions

s

NMR parameters

ms

µs

Internal motions

ns

Macroscopic diffusion

ps

fs

Molecular vibrations

Conformational exchange

Molecular rotations

Diffusion experiments Magnetization exchange Lineshape modifications

Spin relaxation

Averaging of spectral components

T1, T2 Signal lifetime

1/Δω frequencies differences

1/ω0 nutation frequencies

Deuxième école RMN du GERM, Cargèse 2008

Secular interactions averaging

Analysis of µs-ms motions by NMR Conformational exchange ✓ ✓

Motions in the µs-ms range induce spectral modifications. These motions usually correspond conformational exchange conformationnel, chemical exchange.

to or

✓ Their effect on spectra depends of relative values

of kex (τex) and Δω (1/Δω). They are called either slow, intermediate or fast exchange processes at the chemical shift timescale.

A ωA

ωA

k1 ! k−1

B ωB

= k1 /k−1 = pB /pA

Kd

= 1/τex = k1 + k−1 = k1 /pB = k−1 /pA

kex

ωB

ωA ms “slow” exchange

1/Δω c o a l e s c e n c e

μs “fast” exchange

ωB

Ea

ωA

B ωB

A

Δω ωA Deuxième école RMN du GERM, Cargèse 2008

ωB

Analysis of µs-ms motions by NMR Conformational exchange ✓

Conformational exchange is characterized by a probability kex to switch from one state to the other, associated with an instantaneous change of precession frequency.

✓ These sudden frequency changes induce a dephasing of the transverse magnetization, which is added to the natural loss of coherence of spins.

A ωA Kd kex

k1 ! k−1

20 molecules in state A

B ωB

Δν=1000Hz kex=500Hz pA=pB

= k1 /k−1 = pB /pA

= 1/τex = k1 + k−1 = k1 /pB = k−1 /pA

Δν=1000Hz kex=500Hz pA=pB

Δω ωA

ωB

Resulting transverse magnetization for a great number of spins initially in the same state (effect of exchange only)

Deuxième école RMN du GERM, Cargèse 2008

Analysis of µs-ms motions by NMR Conformational exchange ➫ Modification of spectral features

A ωA

ωA

B ωB

= 1/τex = k1 + k−1 = k1 /pB = k−1 /pA

kex

ωB

ωA

20 molecules in state A

Ea

ωA

Δω/2π = 1000Hz kex=500Hz pA=pB

ωB

= k1 /k−1 = pB /pA

Kd ➫ depends of the relative values of kex and Δω

k1 ! k−1

B ωB

A

➫ Additional dephasing of the magnetization ➫ Apparent enhanced transverse relaxation ➫ R2app = R2 + Rex

Δω

➫ Larger linewidths

ωA Deuxième école RMN du GERM, Cargèse 2008

ωB

Analysis of µs-ms motions by NMR Conformational exchange ➫ Modification of spectral features A ωA

k1 ! k−1

Kd

B ωB

kex

= k1 /k−1 = pB /pA

= 1/τex = k1 + k−1 = k1 /pB = k−1 /pA

Δω ➫ depends of the relative values of kex and Δω kex/Δω = 1/10π

ωA

ωB kex(kHz)

Δω/2π = 1000Hz pA=pB

-3kHz

0

kex(kHz)

kex/Δω

0.5

1/4π =0.08

1.0

1/2π =0.16

2.5

5/4π =0.40

3.5

7/4π =0.56

6.3

6.3/2π =1.00

3kHz

Motional broadening

Motional narrowing Deuxième école RMN du GERM, Cargèse 2008

kex/Δω

6.3

6.3/2π =1.0

10

5/π =1.6

20

10/π =3.2

30

15/π =4.8

50

25/π =8,0

Analysis of µs-ms motions by NMR Conformational exchange The case of a two state exchange General equation for a system undergoing conformational exchange r  M (t) r 1  v M (t) M(t) =  2  r…  M t   n ( )

r d r Δ M z (t) = (−R + K )Δ M z (t) dt r d r M + (t) = (iΩ −R +K )M + (t) dt

A ωA

k1 ! k−1

B ωB

ρ − k −k −1  A 1  R −K =   −k ρB − k −1  1

 +   0 d ΔM A t  −iΩA − R 2A − pBk ex €=   dt ΔM+ t   pBk ex B  

() ()

M+ t   A    − p Ak ex M+B t 

p Ak ex −iΩB − R 02B

 +    +  M A t  a AA (t) a AB (t)M A 0   + =  +  MB t   a BA (t) a BB (t)MB 0 

() ()





() ()

In a general case, the expression of the matrix coefficients is complicated

Deuxième école RMN du GERM, Cargèse 2008

() ()

Analysis of µs-ms motions by NMR Conformational exchange  +    +  M A t  a AA (t) a AB (t)M A 0   + =  +  MB t   a BA (t) a BB (t)MB 0 

General equation

() ()

for a two state exchange



a AA (t) =

A ωA

k1 ! k−1

B ωB

Δω ωA

1  1− 2 

 

() ()

      −iωA + iωB + ρA − ρB + k1 − k−1  −iωA€+ iωB + ρA − ρB + k1 − k−1   exp −λ − t + 1 +  exp −λ + t     λ + − λ− λ + − λ−    

(

( )

)

(

( )

)



      −iωA + iωB + ρA − ρB + k1 − k−1    −iωA + iωB + ρA − ρB + k1 − k−1  a BB (t) = 1 1 + exp −λ t + 1 − exp −λ t − +     2     λ + − λ− λ + − λ−     



a AB (t) =



a BA (t) =

(

ωB €

k−1

( )

)

(

(

)

  exp −λ − t − exp −λ + t  λ + − λ− 

( )

k1

(

)

  exp −λ − t − exp −λ + t  λ + − λ−

λ± =

( )

1 2

( )

    −iωA − iωB + ρA + ρB + k1 + k−1 ±  −iωA + iωB + ρA − ρB + k1 − k−1   

(



Deuxième école RMN du GERM, Cargèse 2008

) (

)

2

 + 4k1k−1  

1

2

    

)

Analysis of µs-ms motions by NMR Conformational exchange

An easier case The two state symmetrical exchange €

A ωA



k ! B k ωB

   iω  iω  a AA (t) = 1 1 +  exp − ρ + k − Δ t + 1 −  exp − ρ + k + Δ t  2  Δ Δ        iω iω a BB (t) = 1 1 −  exp − ρ + k − Δ t + 1 +  exp − ρ + k + Δ t  2  Δ Δ    k  a AB (t) = a BA (t) = exp − ρ + k − Δ t − exp − ρ + k + Δ t  2Δ 

{(

)}

{(

)}

{(

)}

{(

)}

{(

  Δ = k2 − ω2   

1

)}

{(

)}

2

€ €

Fast exchange kex >> ∆ω

a AB (t) = a BA

1

k/Δν

[1 + exp(−2kt)] exp(−ρt) (t) = [1 − exp(−2kt)] exp(−ρt) 2

a AA (t) = a AA (t) =

2 1

500 50

( )

+ M + (t) = MA (t) + MB+ (t) = M + (0) exp −ρt



0



ω



kex ~ ∆ω

() ()

Slow exchange kex 3.5 Localized conformational exchange ☞ Local Unfolding ?

Rex(Hz)

30

25

pH=5.8 pH=4.5 pH=3.5

20

15

10

5

0 0

10

20

30

40

sequence

Deuxième école RMN du GERM, Cargèse 2008

50

60

70

80

The method : relaxation dispersion NMR experiments First analysis : qualitative analysis of relaxation dispersion curves An example : variation of exchange contribution as a function of pH pH < 3 Global conformational exchange

(A)

(B)

40

pH > 3.5 Localized conformational exchange

30

I34

N42

35 25

30

(F)

pH=2.5

pH=2.7

Reff (Hz)

pH=3.0

pH=2.7

20

20

35

pH=3.5

pH=3.0

30

2

Reff (Hz) 2

pH=2.5 25

15

15

I34 N42 L68 H71

25 10

5 0

(C)

200

400

600

800

!CPMG (Hz)

1000

1200

5 0

1400

200

400

600

!

CPMG

(D)

35

800

(Hz)

1000

1400

20

15

40

L68

10

H71 35

30

pH=2.5

Reff (Hz) 2

pH=2.7 pH=3.0

20

5

30

pH=2.5

25

Reff (Hz) 2

1200

Reff (Hz) 2

10

pH=2.7

25

0 0

pH=3.0 20

200

400

600

800

!CPMG (Hz)

15

15 10

5 0

10

200

400

600

800

!CPMG (Hz)

1000

1200

1400

5 0

200

400

600

800

!CPMG (Hz)

1000

1200

1400

☞ Physical parameters of these exchange processes ? Deuxième école RMN du GERM, Cargèse 2008

1000

1200

1400

The method : relaxation dispersion NMR experiments Quantitative evaluation of physical parameters of exchange from the relaxation dispersion data

The analysis depends of the timescale of the process



Fast exchange : ∆ω > kex ➫ ➫ ➫ ➫ ➫

#

50

# 2 φ + 2∆ω 2 φ + 2∆ω5.0 D+ = 0.5 1 + " ; D− = 0.5 −1 + " 2 φ2 + ζ 2 φ2 + ζ1.5 $ $ " " 0.5 τCP τ CP η+ = " φ + φ2 + ζ 2 ; η− = " −φ + φ2 + ζ 2 (2) (2)

ζ = 2∆ω(pB − pA )kex

(A)

Rex →kBA=pAkex ; Rex →kAB=pBkex

Rex(B) ≫ Rex(A)



!

!

2 φ = [(pB − pA )kex ]2 − ∆ω 2 + 4pA pB kex

Two peaks Rex independant of B0 Tollinger equation ka, ∆ω, Rinf (B)

R2app = R2inf + 0.5kex − νCP acosh (D+ cosh η+ − D− cos 500 η− )

Detection of minor populations from the

0.05

!

R2app = R2inf + ka 1 − -150

-100

-50

broadenning of the major state resonance : use of Carver-Richards equation Deuxième école RMN du GERM, Cargèse 2008

0

sin ∆ω.τCP ∆ω.τCP 50

ν(Hz) pAνA+pBνB

100

" 150

0

The method : relaxation dispersion NMR experiments Quantitative evaluation of physical parameters of exchange from the relaxation dispersion data The analysis depends of the timescale of the process ☞ Fast exchange : ∆ω > kex ➫ Two peaks ! " sin ∆ω.τCP ➫ Rex independant of B0 R2app = R2inf + ka 1 − ∆ω.τCP ➫ Tollinger equation ➫ ka, ∆ω, Rinf ➫ Rex(B)→kBA=pAkex ; Rex(A)→kAB=pBkex Rex(B) ≫ Rex(A)

0.05

0 -150

-100

-50

0

50

100

ν(Hz) ☞ Detection of minor populations from the broadenning of the major state resonance : use of Carver-Richards equation Deuxième école RMN du GERM, Cargèse 2008

pAνA+pBνB

150

The method : relaxation dispersion NMR experiments Quantitative evaluation of physical parameters of exchange from the relaxation dispersion data pH < 3 Global conformational exchange (A)

(B)

40

pH > 3.5 Localized conformational exchange

30

I34

N42

35 25

30 pH=2.5

pH=2.7

Reff (Hz)

pH=3.0

(F)

pH=2.7

20

pH=3.0

35

pH=3.5

2

Reff (Hz) 2

pH=2.5 25

20

15

30

10

25

I34 N42 L68 H71

15

5 0

(C)

200

400

600

800

!CPMG (Hz)

1000

1200

5 0

1400

200

400

600

!

CPMG

(D)

35

800

(Hz)

1000

1400

40

L68

H71

pH=2.5

Reff (Hz) 2

pH=2.7 pH=3.0

20

15

30

pH=2.5

25

20

10

35

30

Reff (Hz) 2

1200

Reff (Hz) 2

10

5

pH=2.7

25

pH=3.0

0 0

20

15

200

400

600

800

!CPMG (Hz)

1000

1200

1400

15 10

5 0

10

200



400

600

800

!CPMG (Hz)

1000

1200

1400

5 0

200

400

600

800

!CPMG (Hz)

1000

1200

Intermediate/Slow exchange regime pB ≪ pA

1400



Deuxième école RMN du GERM, Cargèse 2008

Fast exchange regime pB ≃ pA

The method : relaxation dispersion NMR experiments relaxation dispersion experiments for the analysis of “invisible” states k1

A B k ωA -1 ωB

ωA

Ea

ωB

B

A Δω

ωA

ωB

pH < 3



Individual fits

➫ Carver statistically favored ➫ τex : few ms ➫ pB