1 Protein NMR spectroscopy Jana Sticht AG Freund, FU Berlin 11.09.2014
2
Applications of protein NMR
• dynamic information
NMR of proteins in solution
protein-ligand interactions
structural information
3
What is special about proteins?
• they are big – often too big for standard NMR approaches
atoms in proteins
stable Isotope composition
(natural abundance)
Hydrogen (H)
1H (99.9885%) 2H (0.0115%)
Carbon (C)
12C (98.93%) 13C (1.07%)
Nitrogen (N)
14N (99.632%) 15N (0.368%)
12 12 10 8 6 4 2 0 -2
d1H [ppm]
methyl groups
aliphatic
Hα
aromatic
amide region
10kD protein
• problem: signal overlap overcome by isotope labeling of proteins expressed in bacteria
in combination with more dimensional spectra (nobel prize for Kurt Wüthrich 2002)
15NH4Cl
13C-glucose
4
the HSQC (heteronuclear single quantum coherence) spectrum correlates covalently linked spins by magnetization transfer via J-coupling (JNH=92Hz)
magnetization transfer 15N to 1H
15N chemical
shift evolution
acquisition
N H
C
C
O
N
C
H
R2
H
R1
i i-1
O
C
H
1H-15N-HSQC spectrum – the “protein fingerprint”
magnetization transfer 1H to 15N
5
1H-15N-HSQC spectrum – the “protein fingerprint”
• less peaks / better resolution • one peak per amide-group roughly one peak per residue (fingerprint!!)
N H
C
C
O
N C
H
R2
H
R1
i i-1
O
C
H
d(1H) [ppm]
d(1
5N
) [p
pm
]
1H-15N-HSQC of a 10kD protein
peak position (chemical shift) contains structural information
peak line-width delivers information on protein dynamics
which peak belongs to which residue? assignment needed!
6
HNCO / HN(CA)CO - sequential assignment
N H
C
C
O
N C
H
R2
H
R1
HNCO signal for C‘ of NH(i-1)
i i-1
O
C
H
i
w H(i)
w NH(i)
w C‘(i-1)
w NH(i)
w H(i)
w C‘(i-1)
N H
C
C
O
N
C
H
R2
H
R1
HN(CA)CO signal for C‘ of NH(i) and NH(i-1)
i i-1
O
C
H
i
w H(i)
w NH(i)
w C‘(i-1)
w C‘(i)
w NH(i)
w H(i)
w C‘(i-1)
w C‘(1)
7
HNCO / HN(CA)CO sequential assignment
w NH(i)
w H(i)
w C‘(i-1)
w C‘(i)
w NH(i-1)
w C‘(i-1) w C‘(i-2)
HNCO / HN(CA)CO
w H(i-1)
N H
C
C
O
N
C
H
R2
H
R1
HNCO signal for C‘ of NH(i-1)
i i-1
O
C
H
N H
C
C
O
N C
H
R2
H
R1
HN(CA)CO signal for C‘ of NH(i) and NH(i-1)
i i-1
O
C
H
8
HNCA / HN(CO)CA sequential assignment
N H
C
C
O
N C
H
R2
H
R1
HNCA signal for Ca of NH(i) and NH(i-1)
i i-1
O
C
H
HN(CO)CA signal for Ca of NH(i-1)
N H
C
C
O
N C
H
R2
H
R1
i i-1
O
C
H w NH(i)
w H(i)
w Ca (i-1)
w Ca(i)
HNCA / HN(CO)CA
w H(i-1)
w Ca (i-1)
w Ca (i-2)
w NH(i-1)
very likely a glycine!
-x-Gly-
-x-Val/Ile/Thr-x-x-Gly-
-14Leu-15Thr-16Ala-17Ser-18Gly-
9
assigned HSQC spectrum
available now: • assignment of NH signals in HSQC • backbone Ca and CO assignments
d(1H) [ppm]
d(1
5N
) [p
pm
]
10
how to access this information
• dynamic information
protein-ligand interactions
structural information
11
d1H [ppm]
d15N
[p
pm
]
1H-15N-HSQC spectra:
one signal per NH-group
Gly16
Val12
Ile13
Gly4
Phe6
Mapping protein-ligand interactions
Val12
Ile13
d1H [ppm]
d15N
[p
pm
]
“fast exchange”
one sharp average resonance
d1H [ppm]
d15N
[p
pm
] “slow exchange”
two distinct resonances
12
protein-ligand interactions
HSQC titrations allow to define binding sites and KD values of protein-ligand interactions
0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
0 1 2 3 4
ratio: [P]/[L]
com
bin
ed c
hem
ical
sh
ift
9GluH
20TyrH
29TrpH
53SerH
60TyrH
70TrpH
KD = 20 µM Dd(1
H,1
5N
) [p
pm
]
[P]/[L] d(1H) [ppm]
d(1
5N
) [p
pm
]
protein : ligand 1:0 1:0.7 1:1.3 1:6.7
13
how to access this information
• dynamic information
solution NMR
protein-ligand interactions
structural information
14
collect structural information: secondary structure from backbone shifts
• Ca, Cb, CO assignments already contain secondary structure information
• Ca chemical shifts of a-helical or b-sheet regions differ from random coil values ∆𝛿𝐶𝛼 = 𝛿𝐶𝛼 𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 − 𝛿𝐶𝛼 𝑟𝑎𝑛𝑑𝑜𝑚 𝑐𝑜𝑖𝑙
Dd(1
3C
a)[
pp
m]
protein sequence
protein + ligand A
protein + ligand B
secondary structure derived from crystal structure
15
two published crystal structures – which is present in solution?
61
68
56
68
collect structural information: secondary structure from backbone shifts
Dd(1
3C
a)[
pp
m]
protein sequence
a-helix
56 68
16
collect structural restraints: torsion angle
• quantitative J-correlation methods allow to determine 3J(HN-Ha) coupling constants
• cross-peak intensity ratio is correlated with coupling constant
Karplus correlation: 𝐽 𝜑 = 𝐴 𝑐𝑜𝑠2 𝜑 − 60 − 𝐵 cos 𝜑 − 60 +3 𝐶
17
collect structural information: side-chain assignments
2D TOCSY (total correlation spectroscopy) correlation of entire spin system
3D TOCSY-HSQC
for bigger proteins
d(1H) [ppm]
d(1
H)
[pp
m]
HN
Ha
NH strip
Hb
Hg
t1 mixing pulse 1H
18
collect structural restraints:
NOE secondary structure and distance information NOE (nuclear Overhauser effect) is due to through-space dipolar interaction between spins in
spatial vicinity (<5Å; NOE intensity ∝1
𝑟6)
2D NOESY
t1 tm 1H
for bigger proteins
3D 1H-15N-NOESY-HSQC intra- or intermolecular
cross-peaks
d(1H) [ppm]
d(1
H)
[pp
m]
NOEs to HN
NOEs to aliphatic hydrogens
NH strip
Ha
Hb
Hg
19
NOE cross-peak intensity correlates with distance between atoms
strong 1.8-2.7 Å medium 1.8-3.3 Å weak 1.8-5.0 Å
collect structural restraints:
NOE secondary structure and distance information
20
F22
V34 M36
G20 F22
V34
d(1
H)[
pp
m]
d(1
H)[
pp
m]
d(1H) [ppm]
X
G20 M36
collect structural restraints:
NOE secondary structure and distance information
22
how to access this information
• dynamic information
solution NMR
protein-ligand interactions
structural information
23
protein dynamics: peak line-width and relaxation
relaxation is dominated by rotational motion of the molecule:
small protein:
long FID sharp signal
big protein:
short FID broad signal
the peak line-width carries information on rotational motion of the molecule and internal dynamics
I(t) = I0 ei(wt) e-Rt
signal intensity (I) oscillating with Larmor frequency
exponential decay of the signal dominated by relaxation rate R2
24
dynamics: quantitative information
qualitative information: broad signals indicate oligomerization, protein-protein interaction, conformational exchange
N H
quantitative information: regions of the protein displaying additional fast internal motion (ti, ps-ns) or conformational exchange (Rex, µs-ms) can be identified from relaxation rate measurements
ti
N
H
25
relaxation time measurements
t
ma
gn
etization
0
equilibrium z magnetization Mz,eq
Mz(t) = Mz,eq - (Mz,eq - Mz(0)) e-t/T1
Inversion recovery to measure T1
1H t
Mxy(t) = Mxy(0) e-n(t-180-t) /T2
CPMG spin-echo to measure T2
1H t t
n
n(t-180-t)
ma
gn
etization
0
Mxy
longitudinal (T1) relaxation time transverse (T2) relaxation time
26
dynamics: quantitative information
d(1H) [ppm]
d(1
5N
) [p
pm
]
1H-15N T1, T2, and hetNOE measurements allow to deduce protein backbone flexibility
27
labeling strategies in big proteins
d(1H) [ppm]
d(1
5N
) [p
pm
]
uniform labeling
selective labeling
problems: • too many signals • broad lines due to fast relaxation
A125G B135G B54G A49G
B47Y
A104V
B90T
A45L
B41D
A55E
B157T
b-c
hain
a-c
hain
1H15N-TROSY-HSQC of deuterated 46kD protein
d(1H) [ppm]
d(1
5N
) [p
pm
]
28
labeling strategies in big proteins d(1
5N
) [p
pm
]
uniform
amino-acid type specific 15N labeling 1H-15N-HSQC
d(1H) [ppm]
incorporation of
protonated amino acids
in deuterated background 1H-15N-NOESY-HSQC
Ala56Hb
d(1H) [ppm]
d(1
H)
[pp
m]
29
labeling strategies in big proteins
d(1H) [ppm]
d(1
3C
) [p
pm
]
I,L,V-methyl group specific 1H-13C labeling 1H-13C-HSQC
105Leu
72Ile
6Val
30
how to access this information
• dynamic information
protein-ligand interactions
structural information