Outline of NMR Spectroscopy

1
Outline of NMR Spectroscopy
Lecture one

• NMR Basics
• NMR Phenomenon-Nuclear Spin
• Relaxation
• J coupling
• Dipolar coupling
• The nuclear Overhauser effect (NOE)
• Chemical shift

• Molecular Motion and kinetics
• I. The NMR timescale
• II. Exchange
• III. Molecular motion

2
NMR
N
uclear
M
agnetic
R
esonance
3
NMR takes advantage of an intrinsic dipole moment
some nuclei have.
Why do some nuclei have an intrinsic dipole
moment and other do not?
If a nucleus has an uneven distribution of
charge, this induces a magnetic field, called
spin. This nucleus behaves like a little magnet.
In general if the number of protons or neutrons
is odd, then the spin is a multiple of 1/2 (i.e.
1/2, 3/2, etc.). If both are odd, then the spin
is 1. If both are even, then the nuclei has no
spin-it has an even distribution of charge and is
not magnetic.
4
Atoms are little magnets
Induced magnetic field
Nuclear cloud
These little magnets affect each other as a
function of DISTANCE to each other. The longest
distance that can be seen from one atom to
another atom is 5 A.
5
Classic representation of a spin
Bo
1/2
The spin precesses in the induced magnetic field,
Bo
-1/2
6
Spin motion
7
Spins in field
Bo
8
Table of NMR active Nuclei
Nuclei Spin Natural Abundance,
1H 1/2 99.98
2H 1 0.015
12C 0 98.9
13C 1/2 1.1
14N 1 99.62
15N 1/2 0.4
16O 0 99.76
17O 5/2 0.039
19F 1/2 100
23Na 3/2 100
31P 1/2 100
59Co 7/2 100
9
What a spectrometer does
10
Energy and wavelength
M
mm
mm
nm
100 10 1
100 10 1
100 10 1
100 10 1
g, wavelength
n, energy
NMR
Microwave
VIS
X-Ray
IR
UV
g-Ray
Radio waves
11
Frequency of Nuclei
The frequency at which a given nucleus precesses
about the induced magnetic field, Bo , is given
by the Larmor equation n g/2p Bo Where g is
the gyromagentic ratio for the given nucleus and
Bo is in hertz. This is the Larmor frequency (of
that nucleus).
For example, when someone tells you that they
have a 600 MHz NMR instrument, they are telling
you the Larmor frequency of the proton nucleus
for that magnet.
12
Chemical Shift

• Chemical Shift is the most basic parameter of NMR
  and is defined as d in parts per million (ppm).
• d (w-w0)/w0 x 106
• Where w0 is the Larmor frequency in Hz of the
  reference line and w is the resonant frequency of
  the line.
• The chemical shift of a nucleus originates in the
  electron cloud around a nucleus, which induces a
  local magnetic field which opposes the applied
  field. This leads to the term shielding and
  deshielding. The more shielded a nucleus is, the
  larger the chemical shift and conversely, the
  more deshielded a nucleus is, the smaller the
  chemical shift.

13
Shielding
K. Wuthrich, NMR of Proteins and Nucleic Acids.
14
Typical Chemical Shifts
Protein chemical shifts
DNA and RNA chemical shifts
15
Relaxation
Relaxation is the precession of the spin from the
perturbed state (x or y) to the relaxed state (z
).

• There are two ways nuclei do this
• Spin-lattice or longitudinal relaxation, T1,
  which returns magnetization to equilibrium.
• Spin-spin or transverse relaxation, T2, which
  leads to loss of phase coherence and signal.

16
90o pulse
17
T1
Mz Mo(1-e -t/T1)
T1 is the time constant at which the
magnetization returns to equilibrium (when all of
the magnetization is aligned along z). T1 is
experimental determined many ways, but the most
common is the inversion recovery method. In
structural work, T1 is estimated based on the
molecules molecular weight (discussed in
Molecular motions). 5xT1 is the time for return
of gt99 of magnetization to equilibrium. This
the amount of waiting time between experiments.
18
T2
T2 is the spin-spin relaxation. T2 represents
the difference in magnetization between otherwise
identical nuclei. T2 has to due with sample
inhomogenity and this contributes to line
broadening. A simple approximation of T2,
T2 n1/2 1/pT2, where n1/2 is the linewidth
at half height.
Many different conformations for identical spins,
causing slight variations in chemical shift.
19
Peak width at half height
n
n1/2
d, Hz
20
Exchange
When a nucleus is exchanging between two or more
different environments, this will give rise to
complicated spectra. Exchange is important in
structural work when you are looking at a dynamic
system. For example, if you have a side chain
that is moving slowly on the NMR timescale,
between two different environments, it can give
rise to multiple peaks that can be described by
T2.
21
J or scalar coupling
Most common coupling between two or more adjacent
spin ½ nuclei. The general formula for splitting
is of lines in the multiplet (N) 2nI 1,
where I is the spin of the neighboring nuclei and
n is the number of identical nuclei. When I1/2,
this simplies to Nn1. In structure work, J
coupling is critical to measuring the bond angles
and dihedral angles between nuclei. The Karplus
equation is used to calculate the dihedral angles
for 3 bond or vicinal coupling J A B cosq C
cos2q Where A, B, C are coefficients that depend
on the nuclei electronegativity.
22
Ethanol A Classic NMR 1 D experiment
TMS
0
d, ppm
23
Dipolar coupling
Dipolar coupling is the interaction between two
dipoles (of two different nuclei) who are not
connected through a bond.
This is one of the most important ways nuclei
relax. It contributes the most to T1 and is
heavily dependent on molecular motion of the
molecule.
24
Nuclear Overhauser Effect
The NOE is the transfer of magnetization between
nuclei. This is the most important effect in
structural work.
It is important to note that it is heavily
dependent on molecular motion (tc) and distance
(rIS). A general equation that describes the
NOE R1 K g2I g2S rIS-6 tc, where I and S are
the interacting nuclei.
25
NMR timescale
NMR has a broad time scale in terms of what it
can see. The useful range for structural
analysis are from seconds to microseconds.
Anything much faster than microseconds, the event
tends to be averaged out and a single event is
seen. Much slower than seconds leads to line
broadening (a T2 effect) and decreasing peaks
(NOE effect), which lead to the reduction and
sometimes disappearance of the signal.
26
Some useful timescales
Average chemical reaction ns (ps to ms) Average
enzymatic reaction ms (s to ns) Average tumbling
time for a globular protein 107 Hz or 100
ns Average tumbling time for a small molecule
1011 Hz or 10 ps
27
Molecular motion
NMR is very dependent on how molecules rotate in
solution. It can be calculated using the Debye
equation (assuming a spherical shape) tc
4pha3/3kT hviscosity of the solvent and aradius
of the molecule. An approximate tc, assuming
typical values for organic solvents, is tc
10-12 Mw, where Mw is in Daltons.
Since this isnt always the case, relaxation
rates (T1 and T2) and NOE measurements can be
used to calculate the correlation time.
Normally, this is done through other means (like
fluorescence anisotropy or DLS).
28
Reference Signal
All NMR spectra must be referenced. This means
that there has to be a compound that has a known
chemical shift in the sample. For example, in
organic chemistry NMR, TMS (tetramethylsilane) is
used to reference spectra-it is set to zero
ppm. For structural work, the reference is set to
the most common molecule/frequency.
29
Reference Table
Isotope Detected Reference Compound Chemical Shift, ppm
1H H2O 4.7
15N NH4Cl 125.7
13C Glucose 50.5
17O H2O 0.0
31P H3PO4 0.0
30
Data Processing
Raw Data
t
Fourier transformation
Processed data
31
(No Transcript)
32
Lecture 2 Outline

• Multidimensional NMR
• Why do we need Multidimensional NMR?
• 2 D NMR
• Nomenclature
• Basic experiments
• COSY
• NOESY
• TOCSY/HOHAHA
• Multiple Quantum Filtration

• 2. How to make a multidimensional
  experiment
• 3. How to read a 3 and 4 D
  experiment
• Structural Information and Structure Generation
• A. Distance Measurements
• B. Coupling Constants and Angle
  measurements
• C. Energy Minimization and
  Restrained Molecular Dynamics
• D. Quality of NMR structures

3. Solvent supression C. 3
and 4 D NMR 1. Why we need 3 and 4
D NMR
33
How to get a protein structure by NMR
Clone and express lots of soluble protein
Preliminary Spectra (1D 1H, NOESY, COSY)
Sequential assignment
Collection of conformational contraints
Calculation of 3D structure
34
Protein Solutions

• Not just any protein structure can be solved by
  NMR
• If the protein meets these conditions, then its
  a good chance its structure can and will be
  solved by NMR
• Mw range up to 30 kD (for most practical
  purposes)
• Soluble at high concentrations (0.5- 1.0 mM) with
  no aggregation (need to rotate freely)
• Must be able to purify to 95 or greater
• Stable at room temperature for days/weeks.
  Better if relatively temperature insensitive.
• Must be structured on the NMR timescale

35
Protein Considerations
Up to 15-18 kD, it is possible to do the
structure without isotopic labeling From 15- 30
kD, need 13C/15N labeling in the protein From 30
kD and up, not only need 13C, 15N labeling, also
need 2H labeling Above 40 kD, need either solid
state NMR or special NMR techniques to remove tc
requirement in NOE (and other parameters)
36
Isotopic labeling
In order for proteins to be isotopically labeled,
they must be grown in the presence of
isotopically labeled nutrients. This requires
some knowledge of metabolism of E. coli or the
organism used to produce the labeled protein In
the case of E. coli, a strain is used that can
grow on minimal media. (i.e. BL21, DH5a) The
cells are grown in minimal media made with
uniformly labeled 15N ammonium chloride or 13C
glucose. Protein purification proceeds as without
labeling
37
Why Multidimensional NMR?
Multidimensional NMR helps resolve resonances
that would otherwise be overlapped and not
identifiable. The multidimensional component
refers to another frequency (not time, space,
etc.) We need to further resolve spectra in
biological NMR because our molecules are not
simple.
38
Typical Chemical Shifts
Protein chemical shifts
DNA and RNA chemical shifts
39
1 D NMR Spectra
1H 1D NMR of Lysozyme
1H 1D NMR of DNA
40
Interesting Facts
In a protein/peptide There are at least 20 spin
systems (1 for each amino acid) The average
number of protons per amino acid is 8 There are
a maximum of 20 inter-residue contacts per
amino acid (in a structured protein/peptide) In
DNA/RNA There are 4-5 spin systems (1 for each
base) The average number of protons per base is
11 There are a maximum of 80 inter-residue
contacts per base
41
Experimential design flow chart
42
2 D NMR- Nomenclature
NMR experiments are named in acronyms Some
examples COSY COrrelation SpectroscopY NOESY-
Nuclear Overhauser Effect SpectroscopY TOCSY-
TOtal COrrelation SpectroscopY HOHAHA-HOmonuclear
HArtman HAhn QF- Quantum Filtered
43
Protein Connectivities
Short range
Longer range
44
2 D NMR- Basic Experiments
Staple experiments for structural work 1. COSY
(1H, 1H) 2. NOESY (1H, 1H) 3. TOCSY/HOHAHA (1H,
1H) These are the basic experiments that form the
framework of more specific experiments
45
COrrelation SpectroscopY
COSY experiments show who is coupled to who
(dipolar coupling and J coupling) This
experiment Confirms the sequence of a
protein Sees residues that are coupled together
(very close together) This basic experiment can
be used in many different ways to look at
different couplings (i.e. 1H-13C coupling) and to
filter out couplings (quantum filtering)
46
COSY in the rotating frame
t1
This is for a one spin system
47
COSY-spectrum
R Ca H
O C
N H
1H
1H
48
COSY of Nucleic Acids
K. Wuthrich, NMR of Proteins and Nucleic Acids
49
COSY of Amino Acids
K. Wuthrich, NMR of Proteins and Nucleic Acids
50
Quantum Filtering
A quantum is a single resonance (also known as
Single Quantum, SQ). Double/Triple/Multiple
Quantum (DQ, TQ and MQ) are multi-coupled
resonances. With filtering, you are selecting for
the quantum number so named and is useful when 2
or more step connectivities are required. i.e.
DQF-COSY selects for resonances that are coupled
twice (hence, Double Quantum).
51
TOtal Correlated SpectroscopY
Otherwise known as HOmonuclear HArtman HAhn
TOCSY experiments detect indirect J coupling
though cross-polarization rather than by relayed
magnetization transfer (relayed magnetization
transfer experiments are called
RELAY-COSY) Cross-polarization (which is caused
by spin-locking) causes the spins to become
equivalent and give rise to a single resonance.
The time it takes for this to happen is equal to
1/2J This experiment is repeated at various
spin-lock times in order to measure this coupling
constant. The difference between TOCSY and
HOHAHA is in how the spin-locking is applied, but
these experiments yield the same information
52
TOCSY-spectrum
H Ca H
O C
N H
1H
1H
53
Heteronuclear Correlation Spectroscopy
These experiments are also known as HMQC and HSQC
(Heteronuclear Multiple/Single Quantum
Coherence) This experiment is essentially the
same as a COSY, except the magnetization transfer
is to another coupled nucleus at another
frequency This is done simultaneously with the
(usual) proton pulses Homonuclear coupling can
interfere with the effect you are trying to
observe, so 1H-1H coupling is irradiated
(decoupled) The delay in this experiment measures
¼ JIS Heteronuclear couplings are usually very
weak effects
54
Nuclear Overhauser Effect SpectroscopY
NOESY experiments detect NOEs (magnetization
transfer between 2 or more nuclei) This
experiment can see other nuclei up to 5 A away
through space NOESY will give the same
information as a COSY, plus the through space
information If a standard distance is known and
NOEs can be measured, the NOE equation can be
simplified to rIM/rIS (aIS/aIM)1/6 Where r is
the distance between IM and IS and a is the
crosspeak intensity Also for this to be true, NOE
buildup curves must be done in order to be
assured that you are at maximum NOE intensity
55
NOESY in the rotating frame
tm
t1
This is for a one spin system
56
NOEs for DNA
57
Solvent Suppression
Very necessary in biological NMR since our
samples are in 1H2O. Solvent suppression involves
irradiation of the water signal with high power
irradiation. This saturates the signal by
equalizing the populations of and spin
states. This can be avoided by putting the
sample in 2H2O, but for protein samples, this is
not easily accomplished.
58
Pulse sequences
59
3 D and 4 D NMR
Why do we need 3D and 4D experiments? 2D
experiments can get very crowded as the number of
residues increases 3 and 4D experiments can help
resolve spectral overlap by modulating 2 D
experiments in various time domains which lets
different magnetization buildup (and selecting
against other magnetization)
60
3 and 4 D experiments
61
How to read a 3 or 4 D experiment
2 D to 3 D to 4 D
2 D to 3 D
62
Structural information
Experiments are finished, how do we compile the
data?
63
How data is fitted
64
Energy Minimization and Restrained Molecular
Dynamics
This subject is a whole class unto itself. These
computational techniques are used to evaluate the
model generated by the data, then help correct
errors via minimization and dynamics. Modeling is
frequently used for sidechains, even residues,
that cannot be visualized by NMR (motion is too
rapid/slow). Because the data typically gives a
range for distances and angles, multiple
structures are generated. The more structures
generated the better.
65
Quality of NMR Structures
Since a lot of structures are generated, one
quick quality assessment is the RMSD between
them. The RMSD should not be more than 2 A The
RSMD equation (for any structure
comparison) RMSD (1/N S (ri ri)2)1/2 Where
N is the number of atoms being compared, r is the
atomic coordinates for the structures in
question Another quality assessment for NMR
structures is the number of NOEs per residue
(both intra/inter residue NOEs)
66
NOEs per Residue
A general rule for resolution of NMR structures
(to keep nomenclature consistent with
Crystallography Less than 5 NOEs per residue 8
10 A structure 6-12 NOEs per residue 5 A
structure 12-15 NOEs per residue 3 A
structure 16-20 NOEs per residue 2.3-2.5 A
structure Over 21 NOEs per residue 2.0 A
structure
Clore GM and AM Groenborn (1991) Science
67
Dynamics
This is NMRs real strength. Can tell what atoms
are mobile and how fast they vibrate. In order
the evaluate dynamics in NMR, one needs to
measure T1, T2 and tc. After these values are
obtained, an order parameter is calculated
(S2). An order parameter is calculated for all
atoms (got the information, why not?) and
plotted. In an average molecule there is motion,
but mobile regions are visible by large shifts in
the order parameter compared to the backbone
atoms.
68
(No Transcript)
69
Homework Handout
Question 1.
70
Question 2
2A. Without isotopic enrichment, which of the
nuclei above would be found in significant levels
in biological systems? 1H, 12C, 14N, 16O, 23Na,
31P 2B. How would you enrich a particular
protein with an isotope that isnt at a
biologically significant level? To enrich a
protein with an isotope, it would need to be
produced in the presence of the isotope in
question Cells would need to be grown in the
presence of compounds enriched for the
isotope(s).
71
Question 3
A.
72
Question 3B
B. Because of their differences, they can be
used to compliment one another, however it has to
be certain that the solvent conditions did not
change the overall structure of the molecule.
Small changes can be accommodated, but large ones
will cause contradictions. 1. NMR can confirm
that crystallization doesnt give a structure due
to solvent conditions. 2. NMR can help solve the
phase problem for crystallography, but 1 has to
be true first. 3. If a region appears to be
mobile in the crystal, NMR can confirm this and
get a minimized structure of the mobile region to
complete the picture. 4. If working together, a
relatively large protein (25-40 kD) can be solved
fast.
73
Problems in the book 12.2
Consider spin-spin (through bond) splitting of
proton peaks for CH3-NH2. A. Diagram the
possibilies for the various interactions. B.
Show the spectrum with the expected splittings
and relative intensities.
ppm
74
Question 12.5
Given that Glu has the following peaks NH8.4,
aH4.3, bH2.1 and 1.9, gH2.3 for both at what
2D coordinates do you expect COSY interactions?
Remember The knowing the exact chemical shift is
not necessary, but knowing the pattern and what
each peak represents is!
75
Question 12.6
In the NOESY below, the off-diagonal circles are
crosspeaks that were in the corresponding COSY,
while the crosses are new crosspeaks in the
NOESY. Explain, using proton numbers, what each
crosspeak tell us about the molecule.
Things to remember Diagonal are crosspeaks to
themselves Peaks that are in the COSY, but not
NOESY are correlated to each other New Peaks in
the NOESY tell us about non-coupled, spins that
are close in space.
76
Question 12.6-NOESY
77
Cell paper
Inhibiting HIV-1 Entry Discovery of D-Peptide
Inhibitors that Target the gp41 Coiled-Coil
Pocket Eckert, DM et al. Cell (1999)
99103-115 What did they do Discovered a
pocket on HIV-1 gp41 and exploited it by
designing D-peptides that bind in it. They appear
to make good contact and are therefore good
inhibitors of HIV-1 infection. They used a
variety of structural techniques to confirm
binding and contact of the peptides with the
pocket. What did they use NMR for? They used NMR
to confirm they mode of binding. Specifically,
they used NMR to show a shift in the trp571
resonances, which indicate that trp571 is
involved in binding of the peptides. This is a
good example of the use of shielding!
78
Figures from Cell paper
79
Structural characterization of the complex of the
Rev response element RNA with a selected peptide
Zhang, Q. et al. Chemistry and Biology (2001) 8
511-520
What did they do Did the structure of the
RSG-1.2 peptide in complex with the Rev response
element RNA because its sequence is different
than the natural Rev peptide and therefore its
binding should be different than Rev. NMR
conditions Peptide 19 residues, RNA 85
bases They used both unlabeled samples and
13C/15N samples They also dissolved their samples
into D2O Used 15, 25, 35 and 45 oC for their
experiments
80
Figure 1
81
Figure 2
82
Figure 3
83
Figure 4
84
Figure 5
85
How good is it?
This works out to be 189 restraints/19 residues
10 restraints/residue for the peptide For RNA
620 restraints / 84 bases 7 restraints/base The
se RMSD measurements are for the converged
structures, NOT the structure(s) generated by NMR!
86
Questions to think about

1. Why did they label both the peptide and the RNA?
   The peptide is small (only 19 residues, Mw
   20Kd)
2. Why different temperatures? Why would that help?
3. Do you believe this structure? Why or why not?
4. How would you make this better (theoretically, of
   course)?

87
(No Transcript)
88
The Domain-Swapped Dimer of Cyanovirin-N is in a
Metastable Folded State Reconciliation of X-Ray
and NMR Structures
Solution structure of cyanovirin-N, a potent
HIV-inactivating protein C.A. Bewley, et al.
Nature Structural Biology (1998) 5571-578
L. G. Barrientos, et al. Structure (2002) 10
673-686