Investigations of Macromolecular Structure and Function

1
Investigations of Macromolecular Structure and
Function

• VII-XI Protein folding, unfolding, and
  prediction of protein structure. Membrane protein
  folding, structure and dynamics.

2
Lecture Plans

• 7-8. Membrane protein folding and dynamics
• EPR of bacteriorhodopsin identifies TM segments
• Structure of K-channels
• EPR provides evidence to back up hinge motion
• MD provides evidence to back up selectivity and
  conduction

3
Learning Outcomes and Biophysical Techniques

• Understand complexities in studying membrane
  proteins
• Understand how folding and dynamic motions of
  membrane proteins are investigated using
• EPR spectroscopy
• Crystallography

4
Why are we interested in membrane proteins?

• 25 of all genes encode membrane-spanning
  proteins
• Either integral or peripheral
• Responsible for signalling, transport, receptor,
  motor processes
• Linked to many inherited diseases
• Cystic fibrosis
• Drug targets
• G-protein coupled receptors
• Ion channels
• Membrane transporters

5
The PDB is Skewed Membrane proteins are poorly
represented

• 25 of all genes encode membrane-spanning
  proteins lt 0.1 of proteins in the PDB

Subset of the protein databank (PDB)
6
Why are membrane proteins hard to study?

• Solubility
• not water soluble
• hard to study in their native environment
• to purify them youve got to strip them from
  their cosy membrane environment
• You may have to reconstitute them to retain
  activity
• Low yields in most expression systems
• you make lots of a membrane protein..you are
  going to disrupt the membrane substantially.

7
The light driven proton pump, bacteriorhodopsin,
provides a nice exception

• bR is the most abundant protein in the
  light-harvesting bacteria, Halobacterium halobium

8
bR is a 7 TM protein

• Seven membrane spanning a-helices
• predicted by hydrophobicity profiling

9
Hydrophobicity scales

• Each amino acid has a given likelihood of being
  in a membrane spanning region of a protein
• Octanolwater partitioning of the individual
  amino acid
• Statistical observation of occurrence in known
  membrane spanning regions

Trp
Gly (normalized middle of the scale)
Asp
Ser
Leu
Ala
Increasing hydrophobicity
10
Hydrophobicity profiles

• Move a 7-11 amino acid window along a proteins
  primary sequence.
• 7 amino acids is two a-helical turns, 11 is three
• Calculate the mean hydrophobicity at each
  position
• Plot mean hydrophobicity as a function of amino
  acid position in the primary sequence.
• Look for peaks of ca. 20 amino acids.
• Each residue in an a-helix is 1.5Å from the
  previous
• 20 x 1.5 30Å thickness of the hydrocarbon
  core of a membrane is about 30Å

MSEGVEAIVANDNGTDQVLGNITGWFNVVLNDGSTGSNLSNFLWHGGSVW
11
bR is a 7 TM protein

• Seven membrane spanning a-helices
• predicted by hydrophobicity profiling
• But this prediction needs confirming by
  structural methods

12
Electron microscopy revisited membrane proteins
are a special case.
Recall that in electron microscopy having all the
particles in solution means that there is a
problem in identifying common orientations.
13
Electron microscopy revisited membrane proteins
are a special case.

• Recall that in electron microscopy having all the
  particles in solution means that there is a
  problem in identifying common orientations.
• Not if you have a protein in a membrane
  environment.
• All the particles will have the same orientation
  2D crystal

Hundreds/thousands of protein molecules all
viewed from exactly the same orientation
14
Electron microscopy revisited membrane proteins
are a special case.

• All the particles will have the same orientation
  2D crystal
• To get 3D data you tilt the specimen in the beam
  of the EM

electron beam

• tilting of 2D crystal in plane of electrons
• missing cone, so 3D data is incomplete

15
Bacteriorhodopsin structure by EM to 7Å
resolution

• At this resolution you can show there are 7
  regions of protein that are all about the same
  length and which all span the membrane.
• Dimensions are compatible with those of an
  a-helix
• Cannot see which helix is which
• Cannot see individual amino acids

16
bR is a 7 TM protein

• Seven membrane spanning a-helices
• predicted by hydrophobicity profiling
• This basic prediction is confirmed by structural
  methods
• But exactly where are the ends of the
  transmembrane regions?

17
Can Biophysical Methods Add to Knowledge of
Membrane Protein Folding?

• Electron spin resonance (ESR) or electron
  paramagnetic resonance spectroscopy (EPR) can.
• but.

18
maths
physics
chemistry
19
Can Biophysical Methods Add to Knowledge of
Membrane Protein Folding?

• Electron spin resonance (ESR) or electron
  paramagnetic resonance spectroscopy (EPR) can
• Electrons are paramagnetic i.e. they spin and
  this confers a magnetic dipole (magnetic moment)
  upon the electron.
• Most electrons are paired, so we cant utilise
  this.
• Unpaired electrons are another kettle of fish
  altogether

20
The Ugly
21
Physics Unpaired electrons spin

• Spin of an electron has two quantum mechanical
  states
• Convention defines these as ½ and -½.
• Normally degenerate

22
EPR Spectroscopy

• In a magnetic field these two spin-states
  (dipoles) will line up approximately either
  parallel or anti-parallel with the applied field.
• Small energy differences between the parallel and
  anti-parallel aligned states, i.e. DE
• Absorb electromagnetic radiation of the right
  wavelength and electron makes transitions between
  the two states.
• This absorbance is the basis of ESR.

23
The Bad
24
Maths(what wavelengths can sense the spin of
unpaired electrons?)

• DE can be estimated using a version of the Bohr
  equation
• DE gbB0,
• where g gyromagnetic ratio (magnetic
  moment/angular momentum) ca. 2
• b is the Bohr magneton constant 9 x 10-24JT-1
• B0 is the applied magnetic field strength in
  Tesla (T)
• (0.1 to 0.7 T magnetic field is used in ESR
  machines)
• DE 2 x 9 x 10-24 x 0.25
• 4.5 x 10-24 J
• I said it was a small difference between the
  parallel and anti-parallel aligned states

25
What frequency of radiation does EPR use?
clue

• DE 4 x 10-24 J
• DE hv
• where h Plancks constant (6.6 x 10-34)
• v frequency
• DE/h v 0.7 x 1010 Hz gigahertz
• microwave radiation, ESR is like a small
  microwave for unpaired electrons

26
The Good
27
The chemistry(what do spin labels look like?)

• Spin labelled nitrogen atom, N (nitroxide group)
• Reactive group to enable covalent attachment to
  protein
• sidechain group most easily targeted by reagents
  is?

28
EPR Spectra what do they look like in an ideal
world?
Data obtained as an absorbance spectrum
For practical purposes, data is always presented
as a 1st derivative
29
EPR Spectra what do they look like in the real
(complicated) world?

• The unpaired electron is on the nitrogen.
• Nitrogen atoms also have a nuclear spin
• The electron spin is affected by interactions
  with the nuclear spin known as splitting
• One line becomes three due to this splitting

30
But I still dont see what that can possibly tell
me!

• Freedom of movement affects the line-width
• Chemical quenchers (other molecules with
  unpaired electrons) also diminish the signal

increasing quenching
increasing immobilization
i.e measure residue accessibility and freedom of
movement dynamic information
31
Hmmm, still seems like a lot of trouble to go to.
There must be some massive advantage of EPR
spectroscopy

• Advantages
• Very few (often there are none) unpaired
  electrons in proteins
• metallo-enzymes (transition metals)
• easy to introduce unique or pairs of unpaired
  electrons by chemical labelling of proteins
• Interactions of spin-labels can report on
• accessibility to chemical quenchers (relaxation
  EPR)
• inter-atomic distances

32
Bacteriorhodopsin Mutants containing Single
Cysteines
single cysteine introduced at 18 positions
from residue 125 to 142
A
B
D
F
C
E
G

• Label each mutant protein with nitroxide spin
  label.
• Incorporate/reconstitute into lipid vesicles
  (real membrane) or detergent (lipid mimetic)
• Probe spectra with spin relaxation/quenching
  chemicals

33
bR and Spin-relaxation EPR

• Three possible states for each probe
• buried in the interior of the protein
• exposed on the aqueous surface
• exposed on the lipid surface

c
a
b
34
The DataChromium oxalate quenching
Chromium is a water-soluble quencher
reduction of the EPR signal
Residues 129-133 are exposed to the aqueous phase
35
The DataOxygen quenching
Molecular oxygen can permeate in water AND
membranes
reduction of the EPR signal
Residues 131-142 are a-helical
36
The Interpretation

• Helix D ends at residue 128
• Helix E begins at residue 131
• Residues 129-132 are solvent exposed (i.e. not
  buried in the membrane)
• The helix exposes residues 133, 136 and 139 to
  the lipid phase
• Residues 134, 138 and 141 buried within protein
• i.e. topology AND orientation are determined

37
Subsequent Confirmation bR crystal structure

• Seven TM a-helices confirmed
• Labelled residues in red
• Dotted lines are the membrane plane

38
Subsequent Confirmation bR crystal structure

• Green residues are those that EPR said would be
  exposed to the solvent
• Chromium oxalate affected their ESR signal

39
Subsequent Confirmation bR crystal structure

• White residues are those that EPR said would be
  exposed to the lipid
• Oxygen affected their ESR signal

40
Subsequent Confirmation bR crystal structure

• Inverted view
• Yellow residues are those that EPR said were
  buried in the protein interior
• Neither oxygen nor chromium affected the signal

41
Summary
42
Investigations of Macromolecular Structure and
Function

• X XI Membrane protein structure and dynamics
  a study of a membrane ion channel

43
Membrane Bound Ion Channels

• May be gated (opened/closed) by different stimuli
• voltage (e.g. K channel)
• pH
• force (i.e. mechanical stress on the membrane)
• ligand (e.g. acetylcholine)
• Closed state non-conducting
• Open state conducting

44
Fundamental biology of K-channels

• Gated by voltage changes across the membrane
  (action potential)
• Open state lifetime milliseconds
• Conductance rate of ion flow several million
  ions per second
• Selectivity K/Na gt 1000
• Yet, Na is smaller than Na
• How are high selectivity and high conductivity
  reconciled?

45
Domain Structure of K-channels

• Homo-tetrameric
• Multiple membrane-spanning
• Common core containing
• inner selectivity region (filter)
• outer conduction region
• Variable regions
• controlling exact method of gating
• Ca2 dependence, pH dependence

pH dependent K channel
voltage dependent K channel
46
Schematic Core Structure of K-channels
M1
SF
M2
single subunit
tetrameric assembly
M1 outer helix M2 outer helix SF
selectivity filter
47
Atomic structure of prokaryotic K-channels

• KcsA pH gated potassium channel from
  Streptomyces lividans
• pH7 channel closed
• pH4 - channel opened
• crystal structure to 3.2 and then 2.0 Angstrom
• (Rod MacKinnon, 2003 Nobel Laureate.)

48
KcsA Structure I pH 7.0 - closed stateA
4-fold symmetric multi-ion pore
The selectivity filter forms a short helix, then
an ordered loop
90 deg
Intracellular
M2 outer helix
pH 7.0 - closed state
49
KcsA Structure II Ion desolvationhow wide is
the channel as you move through it?

• Must gain water upon exit

• Ions go through the selectivity filter
  de-solvated

Direction of ion flow
intracellular
pH 7.0 - closed state
50
KcsA Structure - IIK ion selectivity
Extracellular

• Selectivity filter ordered loop region
• Backbone oxygen atoms interact with the yellow
  K-ions
• Ions do not interact with side-chains in the
  filter
• Possible only with an GYGx sequence

Intracellular
pH 7.0 - closed state
51
KcsA Structure - IIK ion selectivity
Looking down pore
Section along pore axis

• Ions in the pore are still solvated by rigid
  backbone carbonyl atoms
• Just not by water

pH 7.0 - closed state
52
What have structures given us?

• The closed state of the channel
• The positions where K ions may interact
• What could supplement this?
• Experimental evidence of gating
• Computational analysis of proteinion
  interactions

53
Proposed Gating Mechanism

• Selectivity filter relatively unchanged
• M2 helix undergoes large conformational change
• (bloack arrow)
• Opens up the intracellular mouth entrance for
  K ions.

Pore helix
Selectivity filter
Ordered loop
M1 helix
M2 helix
54
Can biophysical methods supplement this data?

• Recall ESR is particularly suited to membrane
  proteins
• Recall that spin labels can be added selectively
  to sites in a protein
• Spin labels can interact with each other ? causes
  splitting of the ESR signal
• This can be used to map the dynamics of a
  membrane protein.

55
KcsA and EPR
Label two of the four subunits with a spin-label
at positions along lower half of M2 helix HOW?
If the M2 helix moves as the channel opens then
the effect that the spin-labels have on each
other will change. Magnitude of that change is a
measure of the distance moved.
56
KcsA and EPR

• Measure EPR at pH7 (closed, thick line) and pH4
  (open, thin line)
• Identical spectra means no change in the
  label-label distance.
• Evidence for a hinge at position 103-107
• Large tilt and twist motion proposed

57
EPR Interpretation
103
107
ESR proposes a 10 tilt and 30 twist
58
Recent X-ray Crystallography Confirms MthK
Open State

• Calcium gated potassium channel
• Crystals grown in presence of high Ca2 and K
  ions
• Interpreted as open channel form.

pore region only shown Ca-sensing domains omitted
59
KcsA (closed) vs. MthK (open)
Extracellular
barrier
Intracellular
60
EPR interpretation and X-ray analysis are in
agreement with each other
103
107
ESR proposes a 10 tilt and 30 twist
X-ray shows a 30 twist of the M2 helix in MthK
61
K-Channel Conduction Hypothesis

• Seven positions for K are found in the X-ray
  structure. Are they all occupied at once?
• How does conduction occur?
• Molecular dynamics sheds light

62
Molecular dynamics

• Computational technique for exploring the dynamic
  behaviour of macromolecules
• follow protein folding
• following ligandprotein interactions
• follow lipidprotein interactions
• exploring conformational space

63
Molecular dynamics theory I(its not easy)

• Know the co-ordinates of all the atoms in a
  protein and the ions and the water molecules at
  time t0
• Also know the velocity of each atom at time t0
  because we assign each atom a velocity such that
  the overall system has a given temperature.Boltz
  mann distribution

But, how do you work out where the atoms will be
at time t t0 dt?
64
Molecular dynamics theory II(still not easy)

• Need to know the force acting on every atom
• F ma (force mass x acceleration)
• We know m (the mass of each atom C 12, O 16
  etc)
• If we knew a then some calculus would give us the
  velocity for each atom and hence its new
  position.

65
Molecular dynamics theory IIIRecall the
Potential Energy Function
Ei Ecov Enon-cov
energy for each atom in the system

• calculus gives us the forces on each atom from
  the energy
• therefore predicting the movement of each and
  every atom
• simulate several million steps
• at each step calculate Fma and integrate

66
Simulations of the transit of potassium through
KcsA

• simulation of ions moving through KcsA
• Just a few ps of time is simulated
• Potassium ion (red) entering from the
  intracellular mouth replaces ions expelled from
  the narrowest region of the pore the
  selectivity filter

Khalili-Araghi  et al., 2006
67
Conduction I the filter is mobile

• Backbone O atoms are not rigid
• But wont that allow any old ion in?
• No, because they maintain an electric field
  strength that gives K selectivity

68
Conduction II the filter is not fully occupied

• In the filter only 2/4 sites are filled. Ions
  swap from one position to another freely.
• Rapid fluctuations from one site to another keeps
  ion flux high

69
Summary

• Membrane proteins are tough but not impossible to
  study
• Combine crystallography, EPR, molecular dynamics.
  electron microscopy and other techniques to fully
  understand mechanism

70
References(Biophysical Methods)

• Sheehan, D. Physical Biochemistry principles
  and applications.
• van Holde, Johnson and Ho. Principles of
  Physical Biochemistry.

71
References(channel structure and dynamics)

• http//blanco.biomol.uci.edu/MemPro_resources.html
  
• (Steve Whites website, numerous useful pages on
  membrane proteins)
• Nobel Prize Lecture of R. MacKinnon
• http//nobelprize.org/chemistry/laureates/2003/mac
  kinnon-lecture.html
• Sansom et al. Potassium channels structures,
  models, simulations. Biochim. Biophys Acta 1565
  (2002) 294-307
• Liu et al. Structure of the KcsA channel
  intracellular gate in the open state. Nature
  Structural Biology 8 (2001) 883-887.
• Doyle (2004). Trends Neuroscience 27 298-302.
• Sansom, Current Biology Curr Biol. 1995 5373-5.
• Khalili-Araghi  et al., Biophys J. 91 L72-L74