Transcript Document

Major Problem in Modern Biochemistry
“The Folding Problem”
Background
Information about the three dimensional structure of
a protein is carried in the amino acid sequence- i.e.
the gene. (Important concept)
Early experiments
Anfinsen - thermally denature ribonuclease (no cleavage)
-refold the protein to be a functional enzyme
How is the information encoded ?
1
Ribonuclease :Reversible folding & unfolding
S-S bonds in brown
Denaturation-studied
by various methods
is reversible
2
Problem:-The Lavinthal paradox
Imagine a polypeptide chain of ~100 residues.
Assume each amino acid can exist in 10
conformations.
Therefore 10100 conformational states are
possible.
Each has a different set of thermodynamic
properties
How can the protein sample each state ?
The number of states is 10 21 times greater than
one estimate of the number of atoms in the
universe. A protein hasn’t time to sample each
conformation. Thus, folding must not be a
completely random phenomenon.
So-how can proteins fold ?
3
4
Second problem:
The existence of a single conformation of a system that
can exist in 10100 states is unlikely. Why ?
Because in selecting a single state the conformational entropy
S (conformational) = RlnN is lost.
The unfavorable contribution
to G is +RTlnN = 8.314 x 300 x ln 10100 = +574 kJ/mole
So there must be stabilizing influences on H to bring the overall
G to a minimum
Thermodynamic parameters for the folding of some
globular proteins at 25 C in aqueous solution
First Issue:
What features of protein folding produce large , negative H
or large positive S changes, to compensate for the
conformational entropy loss ?
5
Internal Interactions - energetically favorable
interactions between groups within the folded
molecule.
1. Charge-Charge Interactions - occur between
positively and negatively charged side chain
groups.
2. Internal Hydrogen Bonds - interactions between
amino acid side chains that are either good
hydrogen bond donors (such as the hydroxyls of
serine or threonine) or good acceptors (such as the
carbonyl oxygens of asparagine or glutamine) .
Though hydrogen bonds are relatively weak, the
large number of them can make a considerable
contribution to stability.
3. van der Waals Interactions - weak interactions
between uncharged molecular groups in the tightly
packed environment of a folded protein. The
contributions of these interactions to the negative
enthalpy of folding is diminished by giving up
favorable interactions with water via folding
.
6
Common Errors - One of the most common folding errors occurs via
cis-trans isomerization of the amide bond adjacent to a proline residue
(see here). Proline is the only amino acid in proteins that forms peptide
bonds in which the trans isomer is only slightly favored (4 to 1 versus
1000 to 1 for other residues). Thus, during folding, there is a significant
chance that the wrong proline isomer will form first. It appears that cells
have enzymes to catalyze the cis-trans isomerization necessary to speed
correct folding.
7
DSC of lysozyme at 3 pH values.
Differential Scanning calorimetry (DSC) is used to
measure the heat required to raise the temperature
of a sample at a constant rate, compared to a
reference buffer. If the sample is undergoing a
physical phase transition, some heat goes into
causing the structural change so that more heat is
needed to raise the temperature.
T2
Δ H = Δ C p dt
T1
where ΔC p is the difference in heat capacity
between the sample and the buffer .
8
9
Dynamics of Protein Folding
How do chains fold into their native structures in spite of
astronomically large numbers of alternatives ?
Clue: Existence of partially unfolded states
High concentrations of Urea (8M) or guanidine-HCl (6M)
leads to completely unfolded states of proteins.
But- Other denaturing conditions lead to small changes in
hydrodynamic and optical parameters.
1967:-John Brandts showed that acid and thermally denatures
proteins can undergo another transition in guanidine-HCl
Example-Bovine or human -lactalbumins undergo two
different conformational transitions when guanidine is
added.
Characterization of different states of protein folding depends
on the availability of a spectroscopic method.
10
NMR Spectroscopy simplified
Nuclei of some elements have spin and behave like small
magnets. The spin states are quantized (take on discrete
values)
Protons = 1/2, deuterons = 1, 13C = 1/2 (I is the number
usually used for nuclear spin)
There are 2I + 1 states for each nucleus. In an applied
magnetic field, these substates have different energies and
can be distinguished.
1H
substates can have mI = -1/2 or +1/2
2H substates can have m = -1, 0 or +1
I
Energy
The energy of these substates in an applied magnetic (H)
field is shown below.
I= -1/2
E = difference in spin state energies
I = +1/2
Magnetic field
The spin of a nucleus can be flipped according to the
equation: E = h. This occurs in the microwave region
of the spectrum. Experimentally, either the microwave
energy can be fixed and the magnetic field changed or
vice versa. Normally the magnetic (H) field
is swept.
11
The main use of NMR is derived from the fact that the energy
levels of a spinning nucleus in a magnetic field depend on
the atomic (electronic) environment.
Different protons for example, absorb at different frequencies
with respect to a reference material:
 = (Href -H)/ Href x 106 (ppm units)
H = nucleus of interest
Href = reference nucleus
Spectacular advances in NMR result from the fact that nuclear
spins also interact through space. Protons closer than 5 perturb
each other’s spins (This is called the Nuclear Overhauser
effect). This interaction allows determination of 3D structure
in solution.
> 6000 structures in the data base. Usually a “family” of structures
is the solution for any single protein
12
13
14
Urea-induced unfolding of bovine carbonic anhydrase B (pH 3).
At this pH the protein is in a molten globule state and unfolds by
a one-step process.
The fraction of unfolding is given by fu =(x-xo)/(xu- xo)
where x is the value of the measured parameter, xo is its value
in the absence of urea and xu is its value at high levels of urea.
Symbols:
, Relative intensity of trp fluorescence at 320/360 nm
, increase in 1/P (P = trp fluorescence polarization)
,
decrease in the negative ellipticity at 220 nm
 increase in intrinsic viscosity
Intrinsic viscosity and the spectrum and polarization of trp
fluorescence reflect the compactness of the molecule, while the
220 nm ellipticity reflects secondary structure.
15
Unfolding of bovine carbonic anhydrase B at pH= 7.5
The fraction of unfolding is given by fu =(x-xo)/(xu- xo)
where x is the value of the measured parameter, xo is its value
in the absence of urea and xu is its value at high levels of urea.
Symbols:
 Relative intensity increase in 1/P
(P = trp fluorescence polarization)
 Increase in I(320)/I(360)
x,   , increase in signal intensities of aliphatic protons
at 3.17, 2.97, 2.00, 0.86 and 1.38 ppm
decrease in negative ellipticity at 270 nm
 Decrease in area under high field NMR resonance
Fluorescence parameters reflect compactness of the molecule
ellipticity at 220 reflects secondary structure
area under high field resonance reflects tertiary structure
16
Spectroscopy shows the following:
The first transition (pH 7.5) shows destruction of the rigid
environment of the aromatic groups,
The second transition shows destruction of the secondary
structure
The absence of rigid tertiary structure and the presence of
secondary structure lead to a model of the intermediate
state as possessing unfolded, non-compact molecules
with local secondary structure.
The state is termed the “molten globule”
Properties of the molten globule
I-Compactness . For -lactalbumin, the hydrodynamic
radius is ~15% greater than the native state and the volume
is 50% greater than the native state.
The fully unfolded molecule (with S-S bonds) has a
hydrodynamic radius increased by 49% from the native state
and a volume increased by a factor of 3.3 from the native state.
II. Presence of core. Non-polar groups are in contact but not
as tightly as in the native protein
III. Secondary structure. Similar to native
17
NMR can be used to determine the mobility of protein structure
HD Exchange can be monitored in D2O solution
Some internal H-bonded atoms can exchange quickly.
Why ?
Local unfolding or “breathing”.
Either a normally buried group must surface occasionally to appear
at the surface or the reagent must permeate to the interior.
k1
Folded

k2
open

exchange
k-1
k1 <<< k-1
18
Five hundred MHz proton NMR spectra of guinea pig
-lactalbumin in the native (pH 5.4), acid (pH 2.0) and
unfolded (in 9M urea) states recorded at 52oC.
The NMR spectrum of the molten globule state is much
simpler than that of the native protein;the number of perturbed
resonances is much smaller, and the overall picture is much
more similar to the unfolded state.
19
Time-dependence of the spin echo amplitudes for the methyl (A)
and aromatic (B) protons of carbonic anhydrase B in the native
(pH= 7.2) , molten globule (pH = 3.3) and unfolded (8M Urea)
states.
The spin-spin relaxation of the methyl groups in the molten globule
state coincides with the unfolded state, while it is quite different
for the native state. In contrast, the spin echo curves for the aromatic
groups are intermediate between those in the native and in the
unfolded states. Therefore, intramolecular movements of aromatic
side chains are much more hindered in the molten globule than in the
unfolded state, although not as hindered as in the native state.
20
IV: Native-like structural Organization
NMR studies show that at least some  -helices are located
in their native positions along the polypeptide chain.
How is this done:
1.
Allow protein in the molten globule state to exchange for a
given length of time.
2.
Transform to native state
3.
Use 2D-NMR to identify N-H protons protected from
exchange in the molten globule B,C helices in -lactalbumin are
protected in the acid molten globule state.
4.
NMR of the molten globule state is much simpler than the
native molecule. Some pronounced resonance at < 1 ppm in the
native form completely disappear in the molten globule.
5.
Environment of many side chains is much less rigid in the
molten globule vs the normal state
(Spin-spin relaxation time diminishes)
Motions of methyl groups coincide with the unfolded state while
motions of aromatics is intermediate between native and unfolded
Why is denaturation so widely studied ? It is thought that denaturation
pathways are the opposite of folding pathways.
Old idea-the concept of an all or nothing transition to a completely
unfolded state is probably wrong.
The all or nothing transition is probably native  molten globule
(movement of side chains, destruction of tight tacking)
Then the system goes from molten globule random coil
through loss of secondary structure.
21
Facts about Molten Globule States in Proteins
1. Proteins can be transformed into the MG state by low or
high pH, by high temperature, by moderate levels of
urea or guanidine HCl, and by the influence of LiClO4 and
other salts (i.e. under mild denaturing conditions).
2. Proteins can be transformed into the MG state without
a change of the environment , simply by small alterations
of their chemical structure.
Examples:
I. Staphylococcal nuclease with 21 C terminal residues removed
II. Point mutants of lambda repressor
III. BPTI with reduced S-S bonds
IV. Alpha-lactalbumins after Ca2+ removal.
22
Example: Urea induced unfolding of carbonic anhydrase B at pH
3 compared with pH 7.5
Folding pathway. Folding starts with the formation of fluctuating
embryos of regions with secondary structure (stabilized mainly by
h-bonds), followed by collapse of these regions into an
intermediate compact structure that is stabilized mainly by
hydrophobic interactions. The final structure is driven by van der
Waals/ and other specific interactions.
Major Problem:
Does any of the above happen in cells?
1. Enzymes exist that accelerate cis-trans isomerization in proline
residues, and others catalyze S-S bond rearrangement.
2. All cells contains “chaperonins”, which either aid protein
folding or prevent protein folding or prevent proteins from
associating prematurely with other proteins.
These were discovered as heat shock proteins accumulated after
cells were subjected to temperature jumps or other stress.
Some chaperonins prevent improper folding of membrane
proteins-or prevent membrane proteins from aggregating.
GroEL is a beautiful molecular machine. The protein to be
protected is bound within the central cavity as a molten globule.
23
Chaperonins - In addition to the enzymes mentioned
previously that assist with proper folding (e.g., cis-trans
isomerase for proline and disulfide bond making
enzymes), cells have a class of proteins called
chaperonins, which "chaperone" a protein to help keep it
properly folded and non-aggregated. Aggregation is a
problem for unfolded proteins because the hydrophobic
residues, which normally are deep inside of a protein, may
be exposed when the protein is released from the
ribosome.
If they are exposed to hydrophobic residues in other
strands, the two strands may associate with each other
hydrophobically (to aggregate) instead of folding
properly.
These proteins were first identified as heat shock proteins
, induced by elevated temperatures or other stress. The
most thoroughly studied are hsp-70 (70 kD) and hsp-60
(60-kD). The GroEL-ES complex from E. coli is one such
chaperone system (hsp-60). It provides a central cavity in
which new protein chains can be "incubated" until they
have folded properly .
24
An Amazing Molecular Machine
25
The best studied chaperonin: GroEL-ES complex from E. Coli
GroEL has two rings each with 7 protein molecules. The center
of each ring has the open hole which is accessible to the solvent.
Either cavity can be capped with Gro-ES again a seven membered
ring of smaller subunits.
The insulating property of this molecular machine prevents
aggregation or misfolding. The process is ATP-hydrolysis driven
Hypothetical Model for chaperonin action in Rubisco
folding.
Active dimer (top) can be unfolded (e.g. 8M Urea)
to give an unfolded polypeptide. The dimer can also be
acid-denatured to give a polypeptide that still retains
elements of secondary structure. It is suspected that a
common intermediate forms from either of these two
states on removal of the denaturant. This intermediate
is labeled Rubisco I. In the absence of chaperonins,
dilution of denatured Rubisco leads to precipitation.
If cpn -60 (a chaperonin) is present during dilution, a
binary stable complex is formed between it and Rubisco I.
When cpn10 and MgATP are added to this complex in the
presence of K+ ions, active Rubisco dimers are formed.
27
Rubisco-major protein component of chloroplasts, possibly
the most abundant protein in the world. Its function is the
key step in CO2 fixation in the Calvin cycle of photosynthesis.
26
+ GroE
- GroE
Suppression by groE complex of aggregation during
refolding. Citrate Synthase was denatured, and refolding was
initiated by 100-fold dilution of the unfolded protein to the
indicated protein concentration. The refolding was measured
in the presence and absence of the GroE complex. A 6-fold
molar excess of GroE complex over CS was used.
28
Prediction of Secondary and Tertiary Protein Structure
Investigators have examined the structures found in proteins and
tried to relate them to the individual amino acids.
Secondary Structure - Table 6.6 lists the relative probabilities that
a particular amino acid will form an -helix, -sheet, and a "turn" in
proteins. Note that the top group of amino acids favors -helices,
the middle group favors -sheets, and the last group favors turns.
The Chou-Fasman rules for predicting secondary structure of a
region of a polypeptide sequence are the following:
1. Any segment of 6 residues or more with an -helix
probability of over 1.03, and not including proline or
phenylalanine, is predicted to be -helix.
2. Any segment of 5 residues or more, with beta -sheet
probability greater than 1.05 (except histidine) is
predicted to be -sheet.
3. Tetrapeptides with an helix probability less than 0.9 and a
turn probability greater than a -sheet probability have a good
chance of being turns.
The secondary structures observed in the native protein BPTI as
predicted .by the Chou-Fasman rules provides exceptionally good
agreement between prediction and experiment.
29
Hydrogen exchange of individual backbone amide protons
in BPTI followed by 2D NMR methods. The protein was 30
dissolved in D2O and kept at 36 C for the indicated times
before spectra were acquired.The resonances from the amide
protons that disappear are identified on the last spectrum
on which they were apparent, using the one letter code for the
amino acids.
Prediction of protein structures
Rose and Srinivasan assume that protein folding is both local
and hierarchichal. Local means that each amino acid’s folding
is influenced by other residues nearby in the sequence.
Hierarchichal means that folded structures develop from the
smallest structural units and work up to more and more
complex entities. The program they have developed is
called LINUS
“Local Independently Nucleated Units of Structure”
The program considers groups of three amino acids in a sequence
-for example residues 12,13,1 4 in a group of 50. The initial
assumption is that this group will (randomly) adopt one of 4
possible structures helix, sheet, turn or loop (other). The program
then asks whether this assumed “ministructure” is energetically
suited to the six amino acids on either side. The program then
moves on to the next (overlapped) set of residues 13, 14, 15 in
this case.
This random selection and testing is carried out many times for the
entire protein. Once this is done, the program analyzes all trials,
looking for local groups of amino acids that prefer one of the four
conformations 70% of the time. Such groups are held in these
conformations while the program repeats the entire process using
comparisons of energetic preferences with respect to 12, amino
acids on either side of selected groups of three residues, then 18
amino acids and so on up to 48 residues.
31
The results of LINUS for some proteins are shown below:
Actual and predicted
structures of three
domains of intestinal
fatty acid binding
protein
Actual and predicted
structures of a helical
domain of cytochrome
b-562
32
33