Nuclear Magnetic Resonance

Tai-huang Huang

Inst. Biomedical Sciences, Academia Sinica April 12, 2016 (NTU/IAMS)

Nuclear Magnetic Resonance

– From Basic Physics to Biomedical Applications

3. Manipulation of nuclear spins -

1. The Dawn of NMR –

.

Outline

2. Exploiting the power of NMR –

Chemistry, biology, material science, and medicine.

5. Look back on a wonderful journey.

4. Biomedical applications –

- Packaging of SARS CoV nucleocapsid.

- Mechanism of SUMO mediated signal transduction.

- Macromolecular dynamics in solid and solution.

1. The Dawn of NMR – A fertile ground for physicists

1924 Pauli proposed the presence of nuclear magnetic moment to explain the presence of hyperfine shift in atomic spectra.

1930 Nuclear magnetic moment was detected using the refined Stern-Gerlach experiment by Estermann.

1939 Rabi et al first detected nuclear magnetic resonance by applying rf energy to a beam of hydrogen molecules.

1946 Purcell et al at Harvard reported nuclear magnetic absorption in parafilm wax.

Bloch et al at Stanford reported nuclear magnetic resonance phenomenom in liquid water.

1940s-60s NMR theories were developed by physicists.

2. Exploiting the power of NMR –

1960s

- Ernst and Anderson intrlioduced Fourier Transform technique into NMR that increased NMR sensitivity by orders of magnitude.

- Solid state NMR was revived due to efforts of Waugh at MIT.

Application to material and polymer science insoluble proteins etc.

- Biological application became possible due to the introduction of superconducting magnet and high power computers.

- NMR imaging was demonstrated (Lauterbur at Stony Brook).

1970s

- Development of multi-dimensional NMR (Jeneer, Ernest, Bax ..) - Development of methodologies for determining macromolecular

structure (Wϋthrich).

1980s and beyond – Exploding applications.

- Methods for characterizing macromolecular structure/dynamics in solution matured.

- Macromolecular structures in solid and gel states become feasible.

- Material science: Zeolites, polymers,

fuel cells

(Clare Grey in Cambridge on Li-Air battery 5x more compact)

- MRI become a powerful clinical imaging modality.

- Functional MRI come to stage.

- Development of several fast NMR methodologies.

- NMR-based Metabolomics.

- ……

Non-trivial applications.

- Each become a sub-discipline by itself.

Felix Bloch Physics, 1952 Edward M. Purcell

Physics, 1952

Kurt Wὕthrich Chemistry, 2002 Isador I Rabi,

Physics 1944

Paul C. Lauterbur

Physiol. Medicine, 2003 Peter Mansfield Physiol. Medicine, 2003 Richard R. Ernst

Chemistry, 1991

Nobel Laureates in NMR

NMR Spectroscopy

B

Radio Wave h

n

Energy

B_o= 0 B_o

E = hn

Biologically interested nuclei:

1H, ¹³C, ¹⁵N, ¹⁹F, ³¹P (S=½),²D (S=1)

Larmor Equation (I = ½):

n =  B

/ 2

n = Larmor frequency

 = nuclear gyric ratio

B_o = magnetic field strength

-1/2

+1/2

Basic Nuclear Spin Interactions

Nuclear Spin i Nuclear Spin j Electrons

Phonons 3

4 4

3

5

1 2 1

6

7

H

H

Dominant Interactions: H = H_Z + H_CSA+ H_D + H_Q + H_J + …

H_z : Zeeman Int.; H_CSA : Chemical Shielding Anisotropic Int.;

H_D= : Dipolar Int. H_Q : Quadrupolar Int. H_J : J-Coupling

Zeeman Interaction (H_z.) (Field depend);

Interaction of nuclear spin with external magnetic field .

H_Q = -γI_Z • Bo

Chemical Shielding Anisotropic Interaction (H_CSA) (Field dep.);

The nuclear shielding effect of an applied magnetic field, caused by an induced magnetic field resulting from circulation of surrounding electrons

H_CSA = -γI • σ • B_o

Dipolar Interaction (H_D) (Thru space) (Field indep):

Interaction between adjacent nuclear spins through magnetic dipolar field.

Basic Nuclear Spin Interactions

Quadrupolar Interaction (H_Q) : (Field indep)

Nuclei with spin > 1/2 have a asymmetric distribution of nucleons (non spherical distribution of positive electric charge)

H_Q = I · V · I

J-Couplings (Thru bond connection) : ( Field indep)

Resonance splitting mediated through chemical bonds

connecting two spins. It is an indirect interaction between two

nuclear spins which arises from hyperfine interactions between the nuclei and local electrons.

1H ¹H ¹H

β₁ β₂ The resonance frequency of a nuclear spin in single crystal depends on the orientation of the tensorial interaction w.r.t. the magnet field.

Single crystal

Interaction Magnitude (Hz) (¹H at 2.1T)

Zeeman 10⁸

Quadrupole 10⁶

Chemical shift 10³

Dipole 10³

J-Coupling 10

NMR spectrum of samples in solid states

Powder patterns

NMR spectra of samples in different states

Small molecules in solution

Gel state (Slow motion)

<H_D> = <H_Q> = 0

<H_CSA> = σ_iso; <H_J> = J_iso Well-resolved sharp lines

Macromolecules

(Slow tumbling) Broad overlapping

Gel state

(Featureless humps)

1. NMR spectra contains rich information derived from the presence of multiple interactions.

2. Each interaction provide insights into the structure/dynamics of the spin system.

3. It is difficult to quantify the interaction when there are more than one present.

Question:

How to extract the inter-twined interactions ?

 Design special pulse sequences to selectively observe/

suppress certain interaction(s)

 Spin gymnastics

Features:

1. Dramatically increased spectral resolution !

2. Dramatically increased sensitivity of insensitive nuclei ! Enhancement factor ∝ (γ_H/γ_I)³

3. Opened a door for thru-bond sequential resonance assignment (Thru J-coupling).

4. The idea can be extended to higher dimension to include multiple nuclei and field gradients etc

Example: (HSQC)

(2D Heteronuclear Single Quantum Correlation Spectroscopy)

B

Radio Wave h

n

NMR Spectroscopy Classical view

Y

X Z

M_X M_Y M_Z

RF field (B_1Y)



B_o

Magnetization will be flipped around Y-axis toward X-Y plane by an angle , determined by the RF field strength and the pulse duration.

Net magnetization

M_

 = B

τ

 = 90^o it is call a 90^o pulse or /2 pulse (maximum signal)

 = 180^o it is call a 180^o pulse or  pulse (No signal)

Protein peptide chain

Efficiency  sin(2J); Maximum transfer when 2J = /2.

Pulse sequence for

N-HSQC expt

15N-HSQC of RC-RNase

1H

15N

Ser135 RC-RNase (12 kDa)

Each spot is a ¹H-¹⁵N pair of a residue

Biomedical Applications

Molecules  Cell  tissue  Organ  Whole body

1. Chemical Identification:

A. Identification of metabolites (Metabonomics) B. Drug discovery.

2. Macromolecular structure:

3. Macromolecular Dynamics:

4. Magnetic Resonance Imaging (MRI):

2. Macromolecular structure:

3. Macromolecular Dynamics:

1. Chemical Identification:

 Organic synthesis, natural product identification etc.

NMR spectrum is the finger print of a chemical

Proton spectrum of ethyl acetate

2. Metabonomics

Metabonomics aims to measure the global, dynamic metabolic response of living systems to biological stimuli or genetic manipulation. It seeks an

analytical description of complex biological samples and to characterize and quantify all the small molecules in such a sample (Urine, blood, plasma etc).

(Nicholson and Lindon, Nature 455, 1054, 2008)

Pattern recognition

Identify metabolites

Statistical analysis Raw data (Urine, blood etc)

NMR spectrum of human urine

Very complex !

Population studies show:

Metabolic variation is much larger than genetic variation !

•

Japanese N = 1000

Americans N = 900 (Urinary Metabotypes)

Chinese N = 900

The World Phenome Center network

中研院台灣人體生物資料庫

(Taiwan Biobank)

 Collect and sequencing 300k samples (200K healthy, 100K patients of various diseases).

(Already Collected over 60k samples now.)

 Perform genome sequence data of all samples for researchers performing other analyses (Data mining).

 Already identified diabetes markers from genome analysis.

 Hope to include NMR- and Mass-based metabonomics data.

2. Macromolecular structure/function

NMR Sample (1 mM, 0.4 ml)

2H, ¹³C, ¹⁵N-label

Obtain NMR spectra Assign resonances

Obtain restrains (Distances, angles,

Orientations etc) Calculate structures

Determine Protein Structure by NMR

NMR structures

(Ensemble of 20 structures)

Sequential resonance assignments

M transfer pathway for HNCA:

1H  ¹⁵N  ¹³C_α  ¹⁵N

 ¹H for Detection

 Detect ¹H, ¹³C, ¹⁵N resonances

Permit sequential correlation of backbone ¹H-¹³C-¹⁵N resonances !!!

Heteronuclear multidimensional NMR experiments thru J-coupling

1. Build a random structure of the given sequence.

2. Energy minimization with least violation by molecular dynamics and simulated annealing to generate many structures.

E_total = E_bond + E_angle + E_improper + E_VDW + E_cdih + E_NOE + E_RDC +….

E_bond = k_b(b-b₀)²; E_φ = k_φ(φ-φ₀)²; E_VDW = k_ij[(σ_ij/r_ij)¹²-σij/r_ij)⁶] E_improper = k_impr(ω-ω₀)²; E_cdih = k_cdih(Ψ-Ψ₀)²;

E_NOE = k_NOE(γ-γ₀)²; E_RDC = k_RDC(θ-θ₀)²;

3. Select 20 structures of least NOE violation (> 0.5 Å).

4. Criteria for good structures:

a) No NOE violation b) RMSD < 0.5 Å

c) No dihedral angle violation (Ramachandran diagram)

Structure Calculation

NMR structure of RC-Rnase

Ensemble of a set of lowest energy structures

1

H –

H NOESY spectrum of RC-Rnase

Identify short ¹H – ¹H distances

1 H chemical shift (ppm)

1H chemical shift (ppm)

Tedious !

Gallery of structures determined

RC-RNase Onconase E. Coli Thioesterase

Dynamics-Fast Motion

Slow Motion

BCKD - LBD BCKD - SBD

Blo t 5 Allergen KP CoA Binding Protein

KP Feo A protein

SUMO-3

Dynamics of onconase

N248-365 of SARS CoV Telomere binding protein

Gallery of structures determined

HDGF dimer of HDGF

N248-365 of SARS CoV octamer

Model of N248-365 of SARS CoV/RNA complex PWWP-domain of HDGF

AtTRP/DNA complex HDGF/heparin complex

2.1. Packaging of SARS Coronavirus Ribonucleocapsid

Four Structural proteins:

EM Schematic

E: Envelope protein (76 a.a.) S: Spike protein (1255 a.a.);

M: Membrane protein (221)

Causative agent – SARS Coronavirus

1. A single stranded plus-sense enveloped RNA virus.

2. Genome of 29,751 nt, containing 14 ORF encoding 28 proteins

N: Nucleocapsid protein (422 a.a.)

Nucleocapsid Protein (NP)

 Binds to RNA to form a helical ribonucleoprotein (RNP):

 Important in virion assembly, packaging and release.

 Interacts with various host proteins and implicated in functions such as replication and apoptosis etc:

 The most abundant viral protein and a major antigenic determinant:

 Target for detection and vaccine developments.

- Interacts with AP-1 signal transduction pathway ?

- Interacts with Smad3 and Modulates transforming Growth Factor- Signaling

- Inhibits Cell Cytokinesis and Proliferation by Interacting with Translation Elongation Factor 1

Unravel the packaging mechanism of helical ribonucleocapsid (RNP) :

1. Dissect N protein domain architecture 2. Probe N protein interaction with RNA.

3. Determine the tertiary structure of N protein.

4. Understand how RNA packs with N protein to form the helical RNP.

Goal

Dissecting Domain architecture of N protein

N181-246 N248-365 N248-422

 Divide and conquer – Construct many sub-fragments and characterize their structures.

 The full length protein (422 a.a.) cannot be crystallized and the NMR spectrum is bad

NTD

CTD

Linker NTD+N-term

CTD+C-term Di-domain

Characterization of protein order by 2D ¹⁵N-HSQC

1H-Chemical shift

1H-Chemical shift

15N-Chemical shift15 N-Chemical shift

Folded protein

Disordered protein

NTD NTD

NTD

CTD CTD

+ CTD

45-365

45- 181 Overlay

248-365

1 45 181 248 365 422

NTD

CTD

Domain architecture of SARS-CoV NP

Structured

(136 a.a.) Structured

(117 a.a.)

Disordered N-terminus

(44 a.a.)

Disordered Linker (67 a.a.)

Disordered C-terminus

(57 a.a.)

CTD CTD

NTD NTD

 Light scattering

 Analytical Ultra- Centrifugation

 Size exclusion chromatography

 Chemical cross linking

 NMR relaxation

CTD forms a dimer

 ~ 50% of SARS-CoV residues exist in intrinsically disordered state.

 Nucleocapsid proteins belong to a class of proteins with the most disordered residues.

Why ?

What are the advantages ?

1 45 181 248 365 422

NTD

CTD

Structured

(136 a.a.) Structured

(117 a.a.)

Disordered N-terminus

(44 a.a.)

Disordered Linker (67 a.a.)

Disordered C-terminus

(57 a.a.)

Domain architecture of SARS-CoV NP

1D 2D 1. Increase collision cross section.

2. Adapt to different shapes.

3. Coupled allosteric effect (Multi-valency effect).

Advantage of intrinsic disorder

NMR Structure Of SARS-CoV NP CTD

 28 kDa homo-dimer solved by Stereo-Array Isotope Labeling (SAIL) method (M. Kainosho of Nagoya U)

A flatten rectangular domain-swapped dimer

Primary RNA binding site.

(ppm) 

R320

H335

A337 Q304

Residue Number

Identification of RNA binding site in CTD

Black: Free

Red: RNA-bound

N protein binds to nucleic acid at multiple sites

cooperatively, much like an octopus clinching onto it prey.

N – Nucleic Acid Interaction

 Modular nature and intrinsic disorder are keys to binding cooperativity and RNP packaging

Top view Side view

X-ray crystallography

- Structure similar to that determined by NMR.

- CTD packs as an octamer in an unit cell.

Crystal packing

 Stacking of 3 octamers forms a complete turn of a left-handed twin helix.

210 Å

30 Å90 Å

DNA binding site NMR (magenta)

 We propose that RNA binds to the Left-handed helix grooves.

DNA binding sites are located in the positively charged grooves

Surface Charge Potential RNA binding model

 A modular protein: It consist of two structured domains and three disordered segments.

 It is highly flexible: ~50% of the residues are intrinsically disordered (ID).

 A sticky protein: It binds to RNA at multiple sites cooperatively.

 The CTD forms a dimer and packs in helical structure in crystal.

Key features of SARS CoV N protein

Proposed model of the N/RNA complex

 CTD forms the core of the left-handed twin-helix .

 NTD covers the exterior and interacts with the bases.

NTD

 Backbone of RNA wraps around CTD core and with bases facing outward.

Side view Top view

N/RNA complex (RNP)

This is just a model !

RNA

Helical RNP

Acknowledgements

Huang’s lab

Dr. Chungke Chang Dr. Chi-fon Chang Dr. Shih-Che Su Dr. Wen-Jing Wu

Yen-lan Hsu Yuan-hsiang Chang Fa-an Chao Tsan-Hung Yu

Hsin-I Bai Liliarty Riang

Hsin-hao Hsiao Yen-Chieh Chiang

X-ray crystallography

Dr. Chwan-Deng Hsiao Chun-Yuan Chang

Yi-Wei Chang

SAIL NMR (Nagoya U) Prof. M. Kainosho

Mitsuhiro Takeda Dr. Chungke Chang

SAXS (NSRRC, Taiwan)

Dr. Yu-shan Huang (SAXS)

NMR Structure

Prof. Peter Guetert (RIKEN)

3. Dynamics

Protein Dynamics

- Energy landscape of protein conformations

Ref. 1. Henzler-Wildman & Kern (2007) Nature 450 :964-72 2. Boehr and Wright (2006) Chem Rev. 106(8):3055-79

Measurement of Macromolecular Dynamics by NMR

NMR experiments

Biological processes

Time scale

NMR can measure a wide range of dynamic processes

N-palmitoylglactosylceramide

Characterize the restricted rotational isomerization of polymethylene chains by deuterium NMR lineshape simulation

Huang et al J. Am. Chem. Soc. 102, 7377-7379 (1980)

Deuterium quadrupole spectra were simulated with two site flipping model similar to that of the crankshaft motion.

Expt Simulated

Crankshaft motion

Huang et al J. Am. Chem. Soc. 102, 7377-7379 (1980)

Tetrahedral two site flipping model

D1

D1

Liquid state - NMR Relaxation

R₁ =1/T₁ = (d²/4)[J(_H - _N) + 3J(_N) + 6J(_H + _N)] + c²J(_N) --- (1) R₂ =1/T₂ = (d²/8)[4J(0) + J(_H - _N) + 3J(_N) + 6J(_H) + 6J(_H + _N)]

+ (c²/6)[4J(0) + 3J(_N)] + R_ex --- (2)

Relaxation Mechanism

Dominated by dipolar and chemical shift anisotropic interactions, and are related to the spectral density functions, J(), by the following

equations:

where d = (_oh_N  _H/8²)(r_NH^-3), c = _N(σ_‖- σ_)/3.

_o : permeability constant of free space; h: Planck constant;

_i : magnetogyric ratio of spin i; _i: Larmor frequency of spin i;

r_NH = 1.02 Å: length of the NH bond vector; R_ex: exchange rate;

σ_‖- σ_ = -170 ppm (size of the CSA tensor of the backbone amide nitrogen).

(Dipolar term) (Chemical shift term)

XNOE = 1 + (d²/4)(_H/ _N)[6J(_H + _N) - J(_H - _N)] T₁ --- (3)

W

Modelfree analysis

J() = ]

)2 ( '

1

) ' 2 ( 2

)2 ( '

1

) ' 1 2

( )2

( 1 [ 2 52

s S s S f

f f S f

m S m















 



 



For a rigid macromolecule undergoing Brownian motion with a rotational

correlation time _m and local internal motion with rotational correlation time _s the spectral density function, J() is given by:

S²: Order parameters (Magnitude of motion)

R _ex : Chemical exchange rate (Slow motion in ms or s regime)

 : Correlation times (Speed of motion)

Fitting T₁, T₂ and NOE data to determine

S

,  and R

Relaxation Data

Obtained in two fields:

: 500 MHz : 600 MHz

Order parameter

S²_av= 0.85

S = 1 rigid S = 0 random

Mostly rigid Flexible

region

Exchange rate – Residues with low motion

Dynamics of E. coli Thioesterase I

Order parameter Exchange term

Huang, et al. (2001) J. Mol. Biol. 307, 1075-1090.

Order parameter Slow motion

Ref. Loria, Rance, and Palmer III (JACS. 1999, 121, 2331-2332)

Carr-Purcell-Meiboom-Gill (CPMG) Sequence

 _cp

Measuring millisecond time scale motion

In which _ex = (₁- ₂)²p₁p₂; p_i and _i are the populations and Larmor frequencies for the nuclear spin in site i, respectively; and

_ex is the lifetime of the exchanging sites.

Solve for _ex for different _cp (measure 0.5 – 5 ms range)

Onconase

800 MHz 600 MHz

69

Energy

Reaction coordinate ES

EI

ΔEa ~ 22 kcal/mol

pA = 99.2% pB = 0.8%

ΔG ~ 2.9 kcal/mol

𝐸 + 𝑆 ⇌ 𝐸𝑆 ⇌ 𝐸𝐼 → 𝐸 + 𝑃

Catalytic scheme:

Reflection of a Wonderful Journal

1. NMR is a prime example of the importance of basic research. The impact of basic research often takes long time to realize.

2. Science is full of surprises. It is only limited by your imagination.

Griffin “John Waugh basically invented the field of solid- state NMR when everyone else had left the field because they thought it was never going to work,"

3. Many areas of today’s science is inter-disciplinary

in nature and a broad knowledge is essential.

Thank you !