1

Bio 5325, Protein Structure & Function, Prof. David P. Cistola

NMR VI: LIGAND-PROTEIN INTERACTIONS

A. Chemical shift timescales for ligand-protein interactionsfast, intermediate and slow exchange regimes

B. Probing ligand-protein contacts by NMR1. chemical shift mapping

application: PH domain from transcription factor TFIIH; interactions with PI(5)P and VP16

2. inter-molecular NOEsisotope-editing vs. -filteringapplication: Human ileal bile acid binding protein

site-selective steroid-protein interactions

2

Exchange on the Chemical Shift Time Scale

e.g. ligand on/off binding site

Lunbound ↔ Lboundkex (Hz)1.0

20

50

100125200

500

10,000

kex, exchange rate in Hz: (k1 + k-1)/2

∆ν, chemical shift difference in Hz

slow exchange: kex > ∆ν

k1

k-1

from Evans, Ch. 1, p. 45

3

CHEMICAL SHIFT TIME SCALE VARIES WITH NUCLEUS AND BO

nucleus Bo(T)

νo(MHz)

timescale in Hz:(e.g. for

∆ν = 0.5 ppm)1H 11.75 T

14.10 T16.45 T

500600700

250300350

13C 11.75 T14.10 T16.45 T

125150175

62.575

87.515N 11.75 T

14.10 T16.45 T

506070

253035

4

CHEMICAL SHIFT MAPPINGGoal:

Map the location of ligand binding site(s) in a protein

Basis:Ligand binding perturbs the local environment (chemical shifts)

of protein atoms in the binding site

Approach:1. Establish resonance assignments for apo-protein (no ligand)

2. If possible, establish assignments for protein with bound ligand OR

titrate protein with ligand and monitor changes3. Calculate and map the chemical shift differences onto apo structure

Caveats:works best with “rigid-body interactions”

if ligand binding is coupled to conformational changes,mapping may not be possible

5

PLECKSTRIN HOMOLOGY (PH) DOMAINS

adaptor modules of 100-120 residues function in signal-dependent membrane recruitment

bind to PIPs and proteins

model of a PH-PIP interaction at

membrane surface

Lemmon (2001) Biochem. Soc. Trans. 29, 377

6

PIP SIGNALING NETWORK IS COMPLEX

Individual PH domains recognize distinct PIP species

Lemmon(2003)

Traffic 4: 201–213

7

TRANSCRIPTION FACTOR TFIIH CONTAINS A PH DOMAIN IN ITS Tfb1 SUBUNIT

Tfb1 specifically recognizes PI(3)P and PI(5)P(levels of individual PIP species vary

in nucleus during cell cycle)

also interacts with transcriptional activation domainsfrom VP16, p53, E2F-1, HIV-1 Tat protein

Di Lello et al. (2005)solved NMR structure of Tfb1 PH domain (apo)

mapped interactions with PI(5)P and VP16 using chemical shifts

8

Chemical Shift Mapping:PH Domain from TFIIHInteractions with PI(5)P

overlayed HSQC spectra at different ligand/protein ratios

weighted amide chemical shift differences

mapped interaction surfaces (red)

DiLello et al. (2005) Biochemistry 44, 7678

9

Chemical Shift Mapping:PH Domain from Tf2bInteractions with VP16

overlayed HSQC spectra at different ligand/protein ratios

weighted amide chemical shift differences

mapped interaction surfaces (yellow)

DiLello et al. (2005) Biochemistry 44, 7678

10

CHEMICAL SHIFT MAPPING SHOWS OVERLAPOF INTERACTION SURFACES

Phospholipid and protein ligands appear to interact with transcription factor at the same location

Competitive binding?

11

INTER-MOLECULAR NOEs FOR PROBING LIGAND-PROTEIN INTERACTIONS

Isotope-editing:follows 1H’s attached to X nuclei (13C or 15N)

suppresses all other 1H’s

e.g., 15N- or 13C-edited NOESY

Isotope-filtering:suppresses 1H’s attached to X nucleus (13C or 15N)

follows all other 1H’s

e.g., 15N- or 13C-filtered NOESY

12

ISOTOPE-EDITED NOESY

e.g., 13C-enriched ligand bound to unenriched protein

Protein(unenriched)

13C-1H1H13C-ligand

3D 13C-edited NOESY (NOESY-HSQC):

NOE 1JCH1H (protein) ·········· 1H (ligand) - 13C (ligand)

F1 F3 F2

13

Ileal Bile Acid Binding Protein (I-BABP)

• member of intracellular lipid binding protein family

• adopts β-clam topology

• functions in bile acid recyclingbetween ileum and liver

• abundant in absorptive cells of distal small intestine (ileum)

14

• steroid ring structure

Bile Acids

• variable OH groups at 3,7 and 12 positions

• conjugated to glycineor taurine in vivo

• synthesized in the liverfrom cholesterol

OH (OH)

(OH)OH

O

23 4 5 6 7

8910

11121314 15

1617

18

19

20

2122

23 24

1

Glycocholic Acid

A B

C D

• amphipathic:“the detergents of the GI tract”

15

Site 1:GCDA

Site 2:GCA

NMR STRUCTURE OF HUMAN I-BABP COMPLEX(DeKoster et al.)

To solve structure of complex:

1. Determined protein structure:[U-13C/15N]-protein,unenriched ligand

2. Obtained ligand-protein NOEs:unenriched protein,13C- or 15N-ligand

3. Docked ligands into bindingsites using NOE restraints

16

I-BABP COMPLEX:DETERMINATION OF PROTEIN STRUCTURE

(ligands present, but not shown)

~3800 NOE restraints~30 restraints/residue

15N15N NOESY15NH NOESY13C-13C Methyl NOESY13CH Aromatic NOESY13CH NOESY

TALOS α & β dihedrals

CSI H-bonds for α-helix

17

Ligand Isotopic Incorporation StrategyTochtrop et al. (2002) J. Org. Chem. 67, 6764-6771

OH OH

C13 C13

NH

O

OH

OH

O

C13C13OH OH

NH

O

OH

OH

O

Via Synthetic Isotopic Incorporation

23, 24 13C Labeled Glycocholic/Glycochenodeoxycholic Acid

3, 4 13C Labeled Glycocholic/Glycochenodeoxycholic Acid

( )

( )

OH OH

NH

O

OHC13

C13

OH

O

OH OH

N15H

O

OH

OH

O

Via Labeled Glycine Conjugation

15N Labeled Glycocholic/Glycochenodeoxycholic Acid

1’, 2’ 13C Labeled Glycocholic/Glycochenodeoxycholic Acid

( )

( )

18

C13C13

NH

O

O

O

OHOH

13C HSQC:3/1 GCDA/I-BABP

⇑

19

13C HSQC:3/1 GCDA/I-BABP

C1 3C1 3

NH

O

O

O

O HOH

HU 3β

U ≡ unbound ligand

20

C13C13

NH

O

O

O

OHOH

H H

13C HSQC:3/1 GCDA/I-BABP

U 4α

U ≡ unbound ligand

21

C13C13

NH

O

O

O

OHOHH H

U 4β

13C HSQC:3/1 GCDA/I-BABP

U ≡ unbound ligand

22

C13C13

NH

O

O

O

OHOHH HH

13C HSQC:3/1 GCDA/I-BABP

Binding site 1

23

C13C13

NH

O

O

O

OHOHH HH

13C HSCQ:3/1 GCDA/I-BABP

Binding site 2

24

ISOTOPE-EDITED NOESY

e.g., 13C-enriched ligand bound to unenriched protein

Protein(unenriched)

13C-1H1H13C-ligand

3D 13C-edited NOESY (3D NOESY-HSQC):

NOE 1JCH1H (protein) ·········· 1H (ligand) - 13C (ligand)

F1 F3 F2

25

13C-Edited NOESY: [3,4-13C]-Bile Salts Bound to I-BABP

26

NOE-detected contacts between Bile Salts and I-BABP1’ Position

Site 1: GCDA Site 2:GCAT73 αH, βH, γH Y53 δH, εHM74 NHG75 NHI23 NH, γH, δH, εH

15N PositionSite 1: GCDA Site 2:GCA

Y53 εH

23 PositionSite 1: GCDA Site 2:GCAT73 αH, γH Y53 εH

3,4 PositionSite 1: Site 2:

GCDA (3) GCDA (4) GCA (3) GCA (4)F63 δH, εH, ζH F63 δH F6 εH F2 εHW49 αH, βH L90 γH, δH W49 5H, 6H L108 βHL90 βH, γH, δH F63 δH L108 βH

L90 γH

27

NOE-observable Ligand-Protein Contacts

28

Residues that may mediate cooperativity and selectivity

29

Sample: [U-13C/15N]-protein, unenriched ligand

Isotope-editing:selects 1H’s attached to X nuclei (13C or 15N)

Isotope-filtering:selects 1H’s attached to 12C or 14N

These two approaches can be combined into one experiment

e.g., 12C,14N-filtered (F1) – 15N-edited (F2) NOESY(ligand) (protein)

DETECTING LIGAND-PROTEIN NOEs WITHOUTISOTOPE ENRICHMENT OF LIGAND

30

X-filter (refocused half-filter)Otting & Wuthrich (1990) Q. Rev. Biophys 23, 39-96

Where:I = 1H attached to X nucleusH = all other 1H (not attached)

And: ∆1 = 1/2 1JIS (X nucleus coupling constant)

Hz + Iz -Hy - Iy

-Hy - IycosπJIS∆1 + 2IxSzsinπJIS∆1 (= -Hy + 2IxSz)

Hy - 2IxSz Hy - Iy A

Hy + 2IxSz Hy + Iy B

∆1

90˚(1H)

180˚(1H,X)

∆1

∆1

180˚(1H)

A + B = Hy (filtered) OR A - B = Iy (edited)

}

31

1st Filter 2nd Filter

2D 13C, 15N Double-half-filtered NOESY(Slijper et al. 1996, J. Magnetic Resonance B 111, 119)

Sample: U-13C,15N-enriched protein, unenriched ligand

Adjustment of 180 pulses leads to selection of:13C-attached protons15N-attached protons

12C- or 14N-attached protons

NOESY

32

I-BABP Complex: 12C-filtered (F1), 13C-edited (F2) NOESY-HSQC

1H from 13C/15N I-BABP

1H fromUnenrichedBile Salts

100 msec

Ligand 1H’s

ProteinAromatic 1H’s

PIP SIGNALING NETWORK IS COMPLEX