Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Photocontrol of protein conformation through the use of photoresponsive surfactants, investigated by small angle neutron scattering
(USC Thesis Other)
Photocontrol of protein conformation through the use of photoresponsive surfactants, investigated by small angle neutron scattering
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
PHOTOCONTROL OF PROTEIN CONFORMATION THROUGH THE USE OF
PHOTORESPONSIVE SURFACTANTS, INVESTIGATED BY SMALL ANGLE
NEUTRON SCATTERING
by
Andrea C. Hamill
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMICAL ENGINEERING)
May 2008
Copyright 2008 Andrea C. Hamill
ii
DEDICAITON
This thesis is dedicated to my family: my parents Barbara and John, my brothers
Raymond and Joseph, my sister Margaret and my boyfriend Marc. My deepest and
heartfelt thanks to all of you for your unconditional love and support. Words cannot
express how grateful I am to have you all in my life. I would not be the person I am
today without you.
iii
TABLE OF CONTENTS
Dedication
ii
List of Tables vi
List of Figures vii
Abstract xii
CHAPTER I: Introduction 1
1.1 Protein Structure 1
1.2 Photoresponsive Surfactants 7
1.3 Proteins 9
1.3.1 Lysozyme, controlling secondary and tertiary structure 9
1.3.2 RNase A, controlling secondary and tertiary structure 12
1.3.3 -chymotrypsin, controlling quaternary structure 15
1.3.4 -amyloid peptide (1-40), controlling quaternary structure 18
1.4 Experimental Methods 21
1.4.1 Dynamic Light Scattering (DLS) 21
1.4.2 Small Angle Scattering (SAS) 23
1.4.3 Transmission Electron Microscopy (TEM) 28
1.4.4 Atomic Force Microscopy (AFM) 30
1.5 References 31
CHAPTER II: Probing Lysozyme Conformation with Light
Reveals a New Folding Intermediate
38
2.1 Abstract 38
2.2 Introduction 39
2.3 Experimental Methods 44
2.3.1 Materials 44
2.3.2 Small Angle Neutron Scattering 46
2.3.3 Dynamic Light Scattering 48
2.3.4 FT-IR Spectroscopy 49
2.3.5 UV-visible Spectroscopy 49
2.3.6 Fluorescence Spectroscopy 50
2.4 Results and Discussion 50
2.5 Conclusion 71
2.6 Acknowledgments 72
2.7 References 72
CHAPTER III: Control of Ribonuclease A Structure through the
use of Photoresponsive Surfactants
78
3.1 Abstract 78
iv
3.2 Introduction 79
3.3 Experimental Methods 82
3.3.1 Materials 82
3.3.2 Small Angle Neutron Scattering 83
3.3.3 Dynamic Light Scattering 84
3.3.4 Fluorescence Spectroscopy 84
3.3.5 UV-visible Spectroscopy 85
3.4 Results and Discussion 85
3.4 Conclusion 101
3.6 References 102
CHAPTER IV: Solution Structure of an Amyloid-Forming Protein
During Photinitiated Hexamer-Dodecamer
transitions Revealed through Small-Angle Neutron
Scattering
105
4.1 Abstract 105
4.2 Introduction 106
4.3 Experimental Procedures 110
4.3.1 Materials 110
4.3.2 Small Angle Neutron Scattering 111
4.3.3 Small Angle X-ray Scattering 113
4.3.4 Fourier Tranform Infrared Spectroscopy 113
4.3.5 Microscopy 114
4.3.6 UV-visible Spectroscopy 115
4.4 Results and Discussion 115
4.4.1 Pure Protein Solutions 118
4.4.2 -Chymotrypsin/azoTAB Solutions 124
4.4.3 Photoinduced -Chymotrypsin Oligomers Are Amyloid
Precursors
132
4.4.4 Photocontrol of Protein Association 135
4.4.5 Photoreversibilty 138
4.4.6 Spherulite Formation 141
4.5 Acknowledgments 142
4.6 References 142
CHAPTER V: -amyloid Peptide (1-40) Fibril Growth Influenced
by Photosurfactant
148
5.1 Abstract 148
5.2 Introduction 149
5.3 Experimental Methods 152
5.3.1 Materials 152
5.3.2 Small Angle Neutron Scattering 153
5.3.3 Atomic Force Micrscopy 155
5.3.4 Light Scattering 156
v
5.4 Results and Discussion 156
5.5 Conclusion 173
5.6 References 173
CHAPTER VI: Future Studies 181
6.1 Introduction 181
6.2 Controlling -chymotrypsin Activity with Light 181
6.2.1 Preliminary Data 181
6.2.2 Future Experiments 184
6.3 Using Ligh Sensitive Surfactants to Treat Burn Victims 185
6.3.1 Background 185
6.3.2 Future Experiments 185
6.4 Unpublished Small Angle Neutron Scattering Data 186
6.4.1 Background 186
6.4.2 Preliminary Data 188
6.4.3 Future Experiments 196
6.5 References 196
BIBLIOGRAPHY 199
vi
LIST OF TABLES
Table 2.1: Values of the Radius of Gyration and I(0) determined from
Guinier Analysis of the SANS data in Figure 2.1, as well as the
protein maximum dimension obtained from the Pair Distance
Distribution Function.
53
Table 3.1: Values of Radius of Gyration (R
g
) and I(0) as determined from
Guinier Analysis and Pair Distance Distribution Function
calculations (shown in Figure 3.2) of the SANS data shown in
Figure 3.1. Also shown are the Hydrodynamic Radii (R
H
) values
calculated form the shape reconstructions shown in Figures 3.4
and 3.5 using Kirkwood’s theory.
92
Table 4.1: Values of Radius of Gyration (R
g
), I(0), Resulting n-mer (n
eff
),
Fraction of Oligomer (x
n
), and Radius of Gyration of Oligomer-
Only Data (R
g
n
) Determined from Guinier or PDDF Analysis of
SANS Data in Figure 4.1.
120
Table 5.1: Parameters from the model dependent fitting of A SANS data. 158
vii
LIST OF FIGURES
Figure 1.1: An illustration of the different degrees of protein structure. 2
Figure 1.2: Generalized scheme of the multipathway fibrillization of insulin
as proposed by Jansen et al.
5
Figure 1.3: Photoisomerization of the Azobenzene-trimethylammonium
bromide (azoTAB) surfactant.
7
Figure 1.4: azoTAB UV-vis absorbance spectra. 9
Figure 1.5: Schematic free-energy surface representing features of the folding
of hen lysozyme as present by Dinner et al.
11
Figure 1.6: The Van der Waals space fill representation and the ribbon
diagram of RNase A in its native state.
13
Figure 1.7: The Van der Waals space fill representations and the ribbon
diagrams of -chymotrypsin and Chymotrypsinogen-A.
17
Figure 1.8: Schematic of a dynamic light scattering apparatus. 21
Figure 1.9: Cartoon representation of diffusing protein molecules. 22
Figure 1.10: Example of the NNLS (Non Negative Least Squares) correlation
function of a protein solution.
23
Figure 1.11: Small Angle Scattering setup. 23
Figure 1.12: Scattering Intensity extrapolated to zero angle (I(0)) scaled by
protein concentration as a function of molecular weight.
25
Figure 1.13: Representation of the -chymotrypsin dimer, illustrating the
distances between two scattering centers, r, and the maximum
dimension D
max
and examples of pair distance distribution
functions for different dimer orientations.
27
Figure 1.14: An example of a the GA_STRUCT fitting procedure. 28
Figure 1.15: Diagram of a Transmission Electron Microscope. 29
Figure 1.16: Schematic of an Atomic Force Microscope. 31
viii
Figure 2.1: SANS data of lysozyme-azoTAB solutions as a function of
surfactant concentration and light illumination.
51
Figure 2.2: Conformation of lysozyme in solution determined from shape
reconstruction of the SANS data, compared with the X-ray
crystallographic structure.
55
Figure 2.3: PDDFs of lysozyme-azoTAB solutions as a function of the
surfactant concentration and light illumination.
58
Figure 2.4: Fourier self-deconvoluted FT-IR spectra of lysozyme in the
presence of 5.0mM azoTAB as a function of the light
illumination.
61
Figure 2.5: Fluorescence of Nile red as a function of the azoTAB surfactant
concentration as measured under both visible and UV light
illumination.
64
Figure 2.6: Lysozyme swelling as measured by (a) DLS and (b) max
of
crystal violet measured by absorption spectroscopy under both
visible and UV light illumination.
66
Figure 3.1: SANS data of RNase A-azoTAB solutions as a function of
surfactant concentration and light illumination.
87
Figure 3.2: SANS experimental data collected for proteins of varying
molecular weight.
91
Figure 3.3: Pair Distance Distibution Functions of RNase A-azoTAB
solutions as a function of azoTAB concentration and light
illumination.
93
Figure 3.4: Conformations of RNase A determined form shape reconstruction
of the SANS data at 0.9 mg/mL compared with the X-ray
crystallographic structure of RNase A.
95
Figure 3.5: Conformations of RNase A determined form shape reconstruction
of the SANS data at 10.0 mg/mL compared with the X-ray
crystallographic structure of RNase A.
97
Figure 3.6: RNase A unfolding as a function of azoTAB concentration and
light illumination; as measured by (a) DLS and (b) Nile Red
Fluoresence and Crystal Violet max
.
98
ix
Figure 4.1: SANS data of -chymotrypsin/azoTAB solutions as a function of
surfactant concentration and light illumination.
117
Figure 4.2: SANS data and pair distance distribution functions of pure -
chymotrypsin solutions.
119
Figure 4.3: Shape reconstructions of the oligomer-only SANS data for pure
-chymotrypsin compared to the X-ray crystallographic
structures of the -chymotrypsin dimer (PDB code 6CHA) and
chymotrypsinogen-A dimer (PDB code 2CGA) at pH 3.
123
Figure 4.4: Guinier analysis of the raw SANS data. 126
Figure 4.5: Pair distance distribution functions of the oligomer only data
scaled by the respective oligomer weight fractions.
129
Figure 4.6: Shape reconstructions of the oligomer-only SANS data at pH 3
and pH 7 as a function of azoTAB concentration and light
illumination.
132
Figure 4.7: (a) FT-IR absorbance spectra at pH 3 for pure -Ch and mixtures
of -Ch with 9.04 mM azoTAB under visible light and UV light.
(b) FT-IR difference spectra (UV - visible) demonstrating the
effect of light illumination. (c) Congo red fluorescence and (d)
apple green birefringence obtained under cross-polarizers. (e) and
(f) TEM images of a fresh solution and (g) of an original SANS
solution after an elapsed time of ~1yr.
134
Figure 4.8: SAXS data of chymotrypsinogen-A/azoTAB solutions as a
function of surfactant concentration and light illumination. Insets
show Guinier fits.
136
Figure 4.9: (a) UV-vis absorbance spectra of 10uM BPB in -
chymotrypsin/azoTAB solutions as a function of surfactant
concentration and light illumination. (b) reversibility of -
chymotrypsin association.
139
Figure 4.10: -chymotrypsin azoTAB solutions exhibiting maltese-crosses
under cross polarized light.
142
Figure 5.1: SANS data of A in the presence and absence of 8 mM azoTAB
as a function of time and light illumination.
157
x
Figure 5.2: Analysis of the SANS data (a) guinier analysis (b) modified
guinier analysis and (c) pair distance distribution functions.
161
Figure 5.3: Shape reconstruction of the A small oligomers with 8 mM
azoTAB and (a-d) their corresponding AFM images.
164
Figure 5.4: Scattering from small oligomers and fibrils combined to model
scattering from the (a) 2 hr and (b) 3 hr samples with azoTAB
under UV light. Insets show shape reconstruction of the small
oligomer fractions.
165
Figure 5.5: AFM images of intermediates and full fibrils. 167
Figure 5.6: (a) Light scattering intensity of pure A and A in the presence of
8 mM azoTAB under both visible and UV light as a function of
time. (b) Light scattering intensity as a function of surfactant light
illumination; visible after exposure to 1 hr UV light and UV after
exposure to 1 hr visible light.
171
Figure 6.1: Fluorescence of -chymotrypsin in the presence of various
azoTAB derivatives, see Figure 6.3.
182
Figure 6.2: Principle of protease detection used in Molecular Probes’
EnzChek Protease Assay Kit.
183
Figure 6.3: Chemical structures of azoTAB derivatives; S1, S4, and S10. 184
Figure 6.4: SANS data of -Lactalbumin/azoTAB solutions as a function of
surfactant concentration and light illumination.
189
Figure 6.5: SANS data of MBP/azoTAB solutions as a function of surfactant
concentration and light illumination.
191
Figure 6.6: SANS data of guanylate kinase in the presence of azoTAB (S1)
under both visible and UV light.
192
Figure 6.7: SANS data of fibrinogen in the presence of azoTAB (S1) under
both visible and UV light.
193
Figure 6.8: SANS data of the SPC and PI3 SH3 domains as a function of time
and in the presence of azoTAB (S1).
194
xi
Figure 6.9: Guinier Analysis of the SANS data from (a) the SPC-SH3 domain
with 8 mM azTAB and (b) the PI3-SH3 domain with 8 mM
azoTAB.
195
xii
ABSTRACT
A photoresponsive surfactant, azoTAB, is used to control protein structure.
When azoTAB is combined with a protein, illumination with visible and UV light can be
used to induce different protein conformational changes. A means to reversibly control
the secondary and tertiary structure of two proteins, lysozyme and RNase A, has been
developed. In the presence of azoTAB and under visible light illumination, the -
domain of lysozyme unfolds, forming a new folding intermediate. Lysozyme transitions
from this intermediate back to its native state under UV light illumination. Similarly, in
the presence of azoTAB and under visible light illumination, a swollen form of RNase A
is observed, which also transitions back to its native state under UV light illumination.
Additionally, a means to control the quaternary structure of -chymotrypsin has been
developed. The degree of self-association of -chymotrypsin, which readily associates
in aqueous solution, is controlled when combined with azoTAB and UV or visible light
illumination. Under visible light, the associated form is a corkscrew hexamer. Under
UV light, these corkscrews self-associate in a slightly offset manner, forming ropelike
dodecamers. The dodecamers were found to be preamyloidal. A means to control the
association of a well-known amyloid protein linked to Alzheimer’s disease, amyloid-
peptide, has been developed. When combined with azoTAB and illuminated with
visible light, the association of amyloid- into fibrils is significantly delayed relative to
the association of pure amyloid- . UV light can be used to trigger the fibril formation
process. In all cases, Small Angle Neutron Scattering was the main tool used to
investigate the various changes in protein structure.
1
CHAPTER I: Introduction
1.1 Protein Structure
Proteins constitute both the building blocks and machinery of all cells,
carrying out an enormous variety of functions. A protein is a string of amino acids,
the sequence of which is considered the protein’s primary structure. Typically a
protein is made up of a portion of amino acids that are hydrophobic (do not like
water) and a portion that are hydrophilic (do like water). When the protein is in
solution this causes the chain of amino acids to fold, protecting the hydrophobic
amino acids from water. The way the amino acids along the chain interact with other,
non adjacent, amino acids via hydrogen bonding is considered the protein’s
secondary structure. There are two kinds of secondary structure, -helices and -
sheets. The -helix is characterized by hydrogen bonds along a chain while the -
sheet is characterized by hydrogen bonds crossing between chains.
Interactions between -helices and -sheets create an overall folding of the
amino acid chain known as a protein’s tertiary structure. Hydrophobic interactions
are largely responsible for tertiary structure. Typically hydrophobic amino acids
make up a protein’s core, while hydrophilic amino acids make up its shell. The
tertiary structure is stabilized by hydrogen and, in many proteins, disulfide bonds
(many but not all proteins contain disulfide bonds).
When two or more amino acid chains (proteins) link together (associate) this
is considered the quaternary structure of a protein. Quaternary interactions can be
2
between two of the same proteins (homogeneous association) or two different
proteins (heterogeneous association). Such associations often result in an “active”
complex (i.e. enzymes, hormones, receptors, chaperones…etc), that can carry out
any number of functions. An illustration of the different degrees of protein structure
is shown below (Figure 1.1).
Figure 1.1: An illustration of the different degrees of protein structure. Primary (amino acid
sequence), secondary ( -helix or -sheet), tertiary (folding of the amino acid chain), and
quarternary (interaction of more than one chain of amino acids) structure.
The biological function of a protein is largely determined by it’s structure. A
great deal of work has been aimed at investigating this form-function relationship.
However, this relationship is not static in nature. Often the protein must adopt a non-
native (partially-folded or associated) conformation to perform its function, whether
beneficial (e.g., enzymatic catalysis, molecular chaperones, transport across cellular
membranes, immune protection, etc.) or harmful. Ailments such as amyloid diseases
(Alzheimer’s, Parkinson’s, type II diabetes), cystic fibrosis, sickle cell disease, prion
diseases (Mad Cow disease, Creutzfeldt-Jacob disease), and even many cancers are
believed to result from improperly folded and/or associated proteins(1, 2).
secondary
structure
primary
structur
amino acids
tertiary
structure
quaternary
structure
-helix
-sheet
H-bonds
-helix
-sheet
3
Partially-folded proteins typically aggregate over the time required for
crystallization, a result of the exposure of the hydrophobic regions of the protein to
solvent(3), limiting the ability to perform x-ray crystallography to determine high-
resolution structural information on unfolded or partially folded protein states.
However these intermediately-folded states are the key to understanding protein
folding. Thus, a novel method of investigating intermediate states in solution, as
opposed to in the crystalline form, is necessary. Nevertheless, direct high resolution
structural studies of these intermediates are usually precluded because of their
transient nature (lifetimes < 100 ms), thereby consigning the investigation of these
kinetic intermediates primarily to the field of simulations(4, 5). Novel experimental
strategies have been developed to study the folding process, such as time-resolved
small angle X-ray scattering, nuclear magnetic resonance (NMR), and various
spectroscopic techniques combined with stoppedflow methods(6-9). An alternate
approach is to study equilibrium intermediates, partially folded states stabilized by
selected solvents, denaturants, or surfactants, because a variety of evidence suggests
that equilibrium intermediates are similar in conformation to the corresponding
kinetic intermediates(10-17). In this manner, “snapshots” along the protein-folding
pathway can be stabilized as equilibrium intermediates, allowing an investigation of
the mechanisms of protein folding.
The method in which a protein folds/associates depends upon the interplay
between electrostatic, hydrogen-bonding, van der Waals, and hydrophobic
interactions among the amino acid residues making up the protein. In the native/non-
4
associated state, the charged and polar amino acid groups typically reside on the
exterior of the protein exposed to water, while the non-polar amino acid groups are
largely found inside the folded structure of the protein, protected from unfavorable
solvent interactions. Protein unfolding/association can then be induced to different
extents by changing variables such as pH, temperature, pressure(18-20), or through
the addition of a chemical denaturant (i.e., urea or GdmCl)(11, 15). Surfactants have
also been shown to change the structure of proteins(13, 14, 21-27), including the
secondary and tertiary structures (swelling/unfolding) of lysozyme and RNase A, and
the quaternary structure (association) of -chymotrypsin and -amyloid (1-40). This
swelling or association can result in the loss of enzymatic activity. However some
surfactants can actually enhance the activity and/or stability of some enzymes as is
the case with -chymotrypsin(25, 26) and lysozyme(28). Because surfactant
hydrophobicity increases as the length of the hydrocarbon tail increases, it has been
found that a greater degree of protein unfolding occurs with increased surfactant
molecular weight for a given headgroup.(29)
Many enzymes as well as virtually every biological process are regulated by
protein-protein interactions. Therefore to fully comprehend biological functions it is
vital to first understand protein-protein interactions. When it comes to protein-
protein interactions the most widely studied are those of neurodegenerative diseases
(Parkinson’s Disease (PD), Huntington’s Disease (HD), and Alzheimer’s Disease
(AD)). Each disease is linked to a specific protein; aggregates of this protein are
believed to be the origin of the pathological conditions associated with each disease.
5
These human neurodegenerative syndromes are viewed as one of the most
perplexing and obstinate problems in medicine. The proteins linked to the disease
misfold and then oligomerize into long insoluble amyloid fibrils. These fibrils then
accumulate forming microscopic deposits, resulting in plaques in the brain.
The fibril growth process is generally understood (see Figure 1.2). Nuclei,
unstructured aggregates form from the association of monomer units; this nucleus is
believed to be the seed for fibril growth. The nuclei associate further into
protofibrils (aggregates exhibiting -sheet structure)) which then associate into
protofilaments (elongated aggregates ~2-5 nm in diameter) and then into fibrils (2-6
intertwined protofibrils).
Figure 1.2: Generalized scheme of the multipathway fibrillization of insulin as proposed by
Jansen et al(30). The hierarchical intertwining of protofilaments and the lateral interaction of
early prefibrillar forms followed by the lateral association of protofilaments are illustrated.
6
Despite the variety of proteins involved in these diseases, the amyloid fibrils
found in the diseased states are extremely similar in their overall appearance. They
are long, straight and unbranched fibrils exhibiting cross- -structure (inter molecular
-sheets) and they bind to dyes such as Congo Red and Thioflavin-T. No sequence
or structural similarities are apparent between any of the proteins that display the
ability to form amyloids.(31)
In this dissertation the effects of photoresponsive surfactants on protein
structure, along with the techniques used to characterize these structures, will be
discussed. The secondary and tertiary structures of lysozyme and RNase A are
reversibly controlled with similar surfactants. A new lysozyme folding intermediate
(a swollen form of lysozyme, with an unfolded -domain), that can be switched back
to the native form with light, is revealed(23). The reversible control of a very similar
swollen form of RNase A is also observed(22). In both cases the hydrophobic
surfactant tails interact with the hydrophobic -helices in the protein causing these
regions of the protein to swell. Having demonstrated the ability to reversibly control
secondary and tertiary structure, it is desired to control quaternary structure (protein
association). Thus -chymotrypsin, which readily associates in aqueous solution, is
studied. When combined with surfactant, preamlyoid oligomer associates are
formed, these associates can be broken up and reformed with light(21). Studies
revealed the oligomers to be corkscrew hexamers that when exposed to UV light
wrap around one another in a slightly offset manner into ropelike dodecamers.
These dodecamer where determined to be preamyloid, thus the next protein studied
7
is -amyloid (1-40) (A 40), another well known amyloid forming protein. A 40 is
the peptide associates with Alzheimer’s Disease (AD). It is shown that azoTAB has a
dramatic effect on the A 40 amyloid growth process(24). Under visible light the
surfactant inhibits amyloid growth, growth can then be accelerated upon exposure to
UV light.
1.2 Photoresponsive Surfactants
Azobenzene-trimethylammonium bromide surfactants (azoTAB) of the form
shown in Figure 1.3 (below), similar to the surfactant used in a previous study(14,
21, 23, 27), was synthesized according to published procedures(32, 33) via an azo-
coupling reaction of 4-ethlyalaniline with phenol followed by reactions with 1,2-
dibromoethane and trimethylamine..
Figure 1.3: Photoisomerization of the Azobenzene-trimethylammonium bromide (azoTAB)
surfactant. Left: the trans, more hydrophobic, planar form under visible light; Right: the cis,
more hydrophilic, bent form under UV light.
N O(CH
2
)
2
N+(CH
3
)
3
Br-
N
CH
3
CH
2
CH
3
CH
2
N
N O(CH
2
)
2
N+(CH
3
)
3
Br-
Visible light
UV light
8
The surfactant undergoes a photoisomerization when illuminated with UV
(350nm) light that is fully reversible with exposure to visible (434nm) light.(14, 21,
23, 34) The visible-light form of the surfactant is primarily in the trans state (75:25,
trans:cis), while the UV-light form is almost entirely in the cis state (>90% cis), as
determined by UV-vis spectroscopy(35). The UV-vis absorbance spectra of each sate
is shown in Figure 1.4 below. The energy associated with light having a wavelength
of 350nm corresponds to the energy need to transition form a to a * bond thus
allowing for rotation about the -N=N- into the cis form. While the energy associated
with light having a wavelength of 434nm corresponds to the energy need to
transition form an n to a * bond again allowing for rotation about the -N=N- back
into the trans form. This photoisomerization changes the dipole moment across the -
N=N- bond in the surfactant.(33) The visible-light (trans, planar) form of the
surfactant has a lower dipole moment and is therefore more hydrophobic than the
UV-light (cis, bent) form, resulting in the UV-light form having a higher affinity for
water and the visible-light form a higher affinity for protein binding.
9
0
0.5
1
1.5
2
2.5
200 250 300 350 400 450 500 550 600
UV
visible
A
b
s
o
r
b
a
n
c
e
Wavelength (nm)
Figure 1.4: azoTAB UV-vis absorbance spectra. The trans (visible light) form shown in red
and the cis (UV light form) shown in blue, [azoTAB] = 1mM, pathlength = 0.1 cm.
Photoresponsive surfactants have been used in a variety of systems to control
different phenomena. Applications include solution and interfacial properties
(surface tension and wettability), aggregation properties, oil-water interface and
microemulsions, gels, and biological systems.(35) azoTAB surfactants have also
been shown to bind and effect protein form as well as function.(14, 21-24, 27, 28)
1.3 Proteins
1.3.1 Lysozyme, controlling secondary and tertiary structure(23)
The c-type (chicken type) lysozymes have been
well studied and shown to
have a wide distribution, including
mammals, birds, reptiles, fishes, and insects. The
main biological function of c-type Lysozymes is to protect the host from bacterial
infections. They do this by catalyzing the hydrolysis of -1,4 glycosidic bonds of the
Absorbance (A.U)
10
peptidoglycan of bacterial cell walls. Hen Egg White Lysozyme has 129 amino acid
residues and has a molecular weight of ~14 kDa. Lysozme is positively charged at
neutral pH (pI = 11.0).
In the specific case of lysozyme, mutations of the protein have been found to
produce structural transitions that lead to amyloidosis(36), with large quantities,
sometimes kilograms, of aggregated protein accumulating in organs such as the liver,
kidney, and spleen(37, 38). It is generally observed that c-type lysozymes do not
exhibit intermediately-folded states during equilibrium folding experiments(39),
except for extreme conditions such as low pH or when mixed with a solvent. In
kinetic folding experiments, however, a populated intermediate has been
observed(40-43). As shown in Figure 1.5 upon initiating refolding from the
denatured state, lysozyme rapidly (within a few milliseconds) forms an ensemble of
different hydrophobically collapsed states(6). After this rapid collapse, the majority
of the molecules (~70%) accumulate into a stable kinetic intermediate, often termed
the -intermediate (I
), with a structured -domain and disordered -domain. The
rate limiting step for the formation of the native state is the refolding of the -
domain, allowing formation of the hydrophobic interface between the and -
domains. Molecular dynamics simulations(44) and mutation studies(45-47) suggest
that the final step in the folding process may be insertion of the -domain residues
Leu-55 and Ile-56 into a hydrophobic pocket of the -domain. Of the remaining 30%
of molecules that do not fold via the formation of the I
intermediate, ~2/3 fold very
11
rapidly as the and -domains form simultaneously (I
/ intermediate), while 1/3
fold very slowly, potentially limited by a cis-trans isomerization of proline.
Figure 1.5: Schematic free-energy surface representing features of the folding of hen
lysozyme as present by Dinner et al.(40) The yellow trajectory represents a “fast track”
where both domains of the protein form concurrently. The red trajectory represents a “slow
track” where the protein becomes trapped in a long-lived intermediate with a folded -
domain and an unfolded -domain.
It is demonstrated that intermediately-folded states of lysozyme can be
stabilized in solution with azoTAB while light illumination can be used to induce
changes between these conformations. Small angle neutron scattering (SANS),
dynamic light scattering (DLS), FT-IR, UV-visible, and fluorescence spectroscopic
measurements are performed on the lysozyme-azoTAB system to investigate the
folding of the protein. At low surfactant concentrations, native-like structures of
lysozyme are observed. Increasing the azoTAB concentration under visible light
12
causes the -domain of the protein to swell, resulting in a partially unfolded structure
at intermediate surfactant concentrations, and to eventually form a swollen/unfolded
form of lysozyme at high concentrations. However, under UV light, at all azoTAB
surfactant concentrations, the protein appears to be in a native-like state; thus, light
can be used to manipulate the protein, allowing for reversible control of protein
folding.
The reversible control of lysozyme’s structure allows for the reversible
control of its function. It was found that azoTAB under visible light causes an 8 fold
increase in lysozyme’s activy. The swelling of the -domain allows for greater
enzymatic flexibility resulting in this SUPERactivity. When the system is then
exposed to UV-light the protein returns to its native form and thus the native activity
of lysozyme.(28)
1.3.2 RNase A, controlling secondary and tertiary structure (23)
RNA depolymerases, “ribonucleases”, catalyze the degradation of RNA.
Specifically, Ribonuclease A (RNase A) is a 124 residue monomeric enzyme that
catalyses the cleavage of the P-O
5
bond in single-stranded RNA. RNase A from
bovine pancreas is one of the model systems of protein science; it was the most
studied enzyme of the 20
th
century. The “A” refers to the predominant form of the
enzyme in the pancreas of Bos Taurus (domestic cow).(48) RNase A has an overall
shape resembling that of a kidney bean with the active site located in the cleft, as
shown in Figure 1.6. It is composed of three -helices ( 1, the N-terminal helix,
13
residues 3-13; 2, residues 24-34; and 3, residues 50-60) and seven -strands ( 1,
the central -strand, residues 41-45; 2- 3, the 65-72 hairpin; 4- 5, the major -
hairpin, residues 79-104; and 6- 7, the C-terminal hairpin, residues 105-124). (49)
The enzyme has four disulphide bonds, which are critical to the proteins stability and
significantly limit conformational changes.
Figure 1.6: (a) The Van der Waals space fill representation: carbon (gray), oxygen (red),
nitrogen (blue), sulpher (yellow); (b) front view; and (c) top view of the ribbon diagram -
sheet (blue), -helix (green), -turn (turquoise), random coil (pink) of RNase A in its native
sate, PDB code 1RBX.
Like lysozyme, RNase A (pI = 9.3) is positively charged at neutral pH and it
has both a fast and slow track when folding. RNase A contains all but one of the 20
naturally occurring amino acids (tryptophan). The three residues believed to be most
important for catalysis are His12, His 119 and Lys 41. A cluster of hydrophobic
residues makes up the major hydrophobic core (residues 58-110, large shaded circle
in Figure 1.6b) of RNase A, composed primarily of the residues form the C-terminal
-hairpin as well as the first and third -helix. A minor hydrophobic core (small
shaded circle in Figure 1.6b) also exists which contains residues from the second -
helix, the major -hairpin and adjoining residues. If one looks at the kidney bean
model of RNase A from the top (Figure 1.6c) the enzyme is divided into two halves,
(a) (b) (c)
Major hydrophobic core
Minor hydrophobic core
14
a front and a back half (as if the kidney bean where sliced lengthwise in two); an N-
terminal, predominantly -helical half (residues 1-60), and a C-terminal
predominantly -sheet half (residues 65-124). Upon thermal unfolding -2 becomes
destabilized at moderate temperatures this loosens the packing of -1 in the major
hydrophobic core.(49) The thermally denatured state has a significantly smaller
radius of gyration compared to that of the chemically denatured state.(50) The
radius of gyration for the denaturant induced unfolded state is smaller than that
expected for a random coil, seemingly because of the four disulfide bonds.(51)
Surfactant denaturants have been shown to have similar effects on proteins to those
of chemical denaturants.
Similar to the lysozyme study, azoTAB is used to stabilize intermediately-
folded states of RNase A in solution. The changes induced by the introduction of
azoTAB into the system can be reversibly controlled with light illumination. The
same techniques are used to investigate the folding of RNase A as in the lysozyme
study (SANS, DLS, FT-IR, UV-vis, and fluorescence). As RNase A and lysozyme
are very similar proteins one would expect similar effects. At low surfactant
concentrations, native-like structures are observed. Increasing surfactant
concentration under visible light causes the protein to swell, again resulting in
partially unfolded structures at intermediate surfactant concentrations. Eventually,
higher surfactant concentrations result in a swollen/unfolded form of RNase A. As
expected this swelling occurs predominantly in the -helical regions. As is the case
with the lysozyme system, under UV light the protein remains in a native-like
15
conformation. This is true for RNase A at all but the highest azoTAB concentration
(8.33 mM). At this concentration, under UV light, the protein is slightly swollen and
not in its native form. As expected, because some of the surfactant is still binding
under UV light (a much smaller fraction than under visible light), at high enough
concentrations the UV form of the surfactant will begin to swell the protein.
Although this is the case there is still a significant change in the size of RNase A at
8.33 mM azoTAB when switching from visible to UV light and visa versa, allowing
for photoreversible control of protein structure.
1.3.3 -chymotrypsin , controlling quaternary structure(21)
-chymotrypsin has been studied extensively; both the structure and
mechanism of action are well known (52-54). The self-association of -
chymotrypsin has been the subject of many studies(20, 55-58). -chymotrypsin
consists of two antiparallel -barrel domains each containing six -strands and a C-
terminal -helix (59) as seen in Figure 1.7). -chymotrypsin is capable of forming
amyloid fibrils. 2,2,2-trifluoroethanol (TFE), a solvent known to stabilize partially
folded proteins promoting amyloid formation (60, 61), has been shown to induce
amyloid-like features in -chymotrypsin. Resembling what happens to proteins
involved in conformational disease, -chymotrypsin aggregation into amyloid-like
structures is triggered by accumulation of a “sticky”, partially folded intermediate
(62). -chymotrypsin activity has also been studied in the presence of cationic
surfactants.
Enzyme activity was found to depend on both the charge and head group
16
of the surfactant; it also depends on the surfactant concentration and the nature of the
buffer used.(26) -chymotrypsin activity can either be depressed or significantly
promoted by cetyltriethylammonium bromide (CTABr) and cetyltributylammonium
bromide (CTBABr), respectively(25). These surfactants were tested in the
enzymatic hydrolysis of N-glutaryl-L-phenylalanine p-nitroanilide (GPNA). In
comparison with the value in pure buffer the GPNA hydrolysis rate was halved in the
presence of CTABr while it was 5.9 times higher with CTBABr.(25) The role of
surfactant cationic head group appears to be particularly important in determining the
positive interactions that induce the enzyme superactivity. The increase of the alkyl
head group hydrophobicity in the series methyl < ethyl < n-propyl < n-butyl leads to
a marked enhancement of the observed enzyme activity. It is believed that this result
is attributed to the progressive increase of microinterface net charge (63, 64) since
affinity of surfactant monomers and aggregates for counterions decreases with
increasing bulk hydrophobicity of alkyl head groups in the series methyl < ethyl < n-
propyl < n-butyl. Also, it is well known that head group enlargement from CTABr to
CTBABr dramatically changes the degree of ionization ( ) in micellar aggregates,
i.e., of CTABr is approximately 0.2 and for CTBABr is ~0.5.(25, 65)
17
Figure 1.7: The Van der Waals space fill representations (left), carbon (gray), oxygen (red),
nitrogen (blue), sulpher (yellow) and the of the ribbon diagrams (right) -sheet (blue), -
helix (green), -turn (turquoise), random coil (pink) of -chymotrypsin/ (top) and
Chymotrypsinogen-A (bottom), corresponding PDB codes shown.
The ability to reversibly control protein quaternary structure, specifically that
of -chymotrypsin with light, will be shown. In addition to the techniques used to
study the lysozyme and RNase A systems, SAXS and several microscopy techniques
(polarized light, fluorescence and transmission electron) will be employed.
Illumination with UV light results in a higher degree of association, while the visible
form of the surfactant appears to inhibit such association. Thus, protein-surfactant
interactions are replacing protein-protein interactions when the surfactant is in the
more hydrophobic, visible-light (trans) form. Specifically a monomer oligomer
equilibrium is observed under both visible and UV light. As surfactant concentration
is increased this equilibrium shifts toward the oligomer, a hexamer under visible
light and a dodecamer under UV light. The monomer structure remains unchanged,
independent of both surfactant concentration and light condition.
Chymotrypsinogen-A, the inactive form of -chymotrypsin, association is
also studied to determine the role of the active site in association (see Figure 1.7).
-chymotrypsin
Chymotrypsinogen-A
dimer
2CGA
6CHA
monomer
1CHG
2CHA
18
Chymotrypsinogen-A is activated to -chymotrypsin by the removal of two
dipeptides, Ser
14
-Arg
15
and Thr
147
-Asn
148
. The full proteolytic activity is caused by
small polypeptide chain rearrangements near the active site, resulting in the complete
formation of the side-chain specificity pocket and the oxyanioin hole for catalytic
activity.(66) A significant difference between the two proteins is the ability of -
chymotrypsin, but not chymotrypsinogen-A, to readily dimerize in aqueous
solution.(67-69) Although the monomer structures of Chymotrypsinogen-A and -
chymotrypsin are nearly identical their dimer structures are quite different (see
Figure 1.7) . Chymotrypsinogen-A associates “back-to-side” while -chymotrypsin
associates “active site-to-active site” (“front-to-front”).
1.3.4 -amyloid peptide (1-40), controlling quaternary structure(21)
The -amyloid peptide (A ) implicated in Alzheimer’s Disease (AD) is a 39-
43 amino acid peptide(70), a proteolytic fragment of the amyloid- precursor protein
(A PP). The assembly of A is believed to be the key to AD. The most abundant
forms are A 40 and A 42, 40 and 42 amino acids long, respectively. Recently,
focus on what is believed to be the primary pathogenic species has shifted to the
prefibrillar intermediates (early oligomers, nuclei and protofilaments). Once fibrils
have fully developed the process is considered irreversible thus the development of
strategies for treating Alzheimer’s disease could be greatly improved if the early
stages of the A 40 assembly process were well understood.
19
Very little is know about the solution structure of early amyloid species, as
X-ray crystallography and solution NMR, the most common techniques used to
determine protein structure, are generally limited to the study of native proteins in
the solid state. Thus a technique capable of investigating proteins in solution,
allowing one to study non-native protein states is warranted.
SANS is such a technique; it is capable of investigating structures ranging
from 10 to 1000 Å in solution. Thus, in the case of protein aggregation one can
follow the process from the single monomer units through early assembly (nuclei
and protofibrils) all the way up to fully aggregated states (protofilaments and fibrils).
SANS experiments have contributed to a better understanding of the mechanism of
different types of protein aggregation in solution.(71) Specifically, in the case of
A 40, SANS was used in the low-resolution structural determination of protofibrils
found to be cylindrical with radii and length of 24 and 110 Å, respectively.(72)
Atomic Force Microscopy (AFM) has been used extensively in the study of
protein and peptide self assembly systems, specifically those involved in amyloid
formation.(73) AFM is a very powerful tool for the structural analysis of biological
samples because of the ability to image nonconductive specimens under a variety of
different conditions(74, 75) (samples do not have to be stained by heavy metals or
imaged under high-vacuum conditions as in Transmission Electron Microscopy
(TEM)). The ability to image in an ambient environment is a huge advantage over
TEM. AFM is capable of characterizing the wide range of structures present in the
A growth process from the nanometer sized nucleus up to micron sized amyloid
20
fibrils (as is SANS). However, the major advantage of AFM is that even if several
different species exist in one given sample, under the right conditions these species
can be individually imaged, while SANS gives a weighted average of all species in
solution. However major disadvantage is the drying of samples required for AFM,
this can cause “drying effects” in the images obtained where the structures are
directly influenced by the drying method, this is where SANS has the major
advantage as the structures obtain are solution structures. Therefore when SANS and
AFM are combined as in the present study a tremendous amount of detailed and
complementary structural information can be obtained.
The present study sets out to investigate the effect of azoTAB on the A 40
peptide. SANS, AFM and light scattering (LS) are used to fully investigate the A 40
fibril growth process over time, as a function of surfactant and light illumination.
The ability to inhibit and then trigger the growth process of amyloid fibrils is
demonstrated. The presence of azoTAB inhibits initial fibril formation. Under
visible light this delay in the fibril growth process is more severe than under UV
light thus UV light can be used to trigger fibrillization. Shape reconstruction applied
to the SANS data from samples containing azoTAB allowed for the structural
determination of the A 40 nucleus, which is consistent with the A nucleus from the
literature.(21, 72, 76) Under visible light these nuclei are the only significant
scatterers at an age of 0, 2, and 3 hrs while at an age of 5 hrs fibrils have formed.
Under UV light, the 0 hr sample shows only scattering from the nuclei the 2 and 3 hr
samples have scattering consistent with a mixture of nuclei and fibrils. The 5 hr
21
sample under UV light has scattering consistent with that of a 3D network of fibrils.
This demonstrates that UV light illumination directly results in a high degree of
association. Light scattering is used to demonstrate the ability to triggre fibril growth
by simple illumination with UV light.
1.4 Experimental Methods
1.4.1 Dynamic Light Scattering (DLS)
When light comes in contact with a protein solution, the light will be
scattered by the protein molecules. The scattered light intensity is determined by the
size, shape, and molecular interactions of the protein molecules in solution. The set
up of a dynamic light scattering experiment is shown in Figure 1.8
Figure 1.8: Schematic of a dynamic light scattering apparatus.
Proteins in solution are not stagnant, Figure 1.9 shows an example of this. Their
positions are constantly changing with time, therefore the scattered light intensity
also changes with time.
sample
632.8nm HeNe Laser
22
Figure 1.9: Cartoon representation of diffusing protein molecules.
As time passes the existing state changes from the initial state. An
autocorrelation describes how a given measurement relates to itself in a time
dependent manner. As time progresses the autocorrelation diminishes. The decay of
the autocorrelation is described by an exponential decay function,
2
2
) (
Dq
e G
,
which relates the autocorrelation to the diffusion coefficient D and the measurement
vector q, =
2
sin
4 q , where n is the refractive index, is the wavelength of
incident light (632.8nm), and is the scattering angle. By fitting the points of
autocorrelation to the function G( ), the diffusion coefficient, D, can be measured.
The hydrodynamic radius, R
H
, is then calculated assuming a spherical shape,
according to the Stokes Einstein equation, D T k R
B H
6 / = , where k
B
is
Boltzmann’s constant, T is the temperature, is the viscosity of the solvent, and D is
the experimentally-determined diffusion coefficient. Figure 1.10 illustrates the
correlation function collected for a protein solution and its non negative least squares
fit.
t = t
t = 0
1
1
2
2
3
3
4
4
23
Figure 1.10: Example of the NNLS (Non Negative Least Squares) Correlation Function of a
protein solution; correlation (blue), experimental data (red).
1.4.2 Small Angle Scattering (SAS)
Small Angle Scattering is very similar to light scattering. In SAS, the incident
wavelength, , and scattering angle, , determine the length scale probed through the
relationship
d . The set up of a small angle neutron scattering experiment is
shown in Figure 1.11.
Figure 1.11: Small Angle Scattering setup.
24
Through the use of cold neutrons (long wavelength,
neutron
= 6 Å) and tight
beam collimation, the 30m SANS instrument at the National Center for Neutron
Research (NCNR) at the National Institute of Standards and Technology (NIST), in
Gaithersburg, MD, is able to probe structures on a length scale, d, ranging from 60 Å
to 1000 Å. SAXS,
X-ray
= 1.24 Å, is typically used to investigate structural details in
the 0.5 to 50 nm size range. Q is the measurement vector it has units of Å
-1
,
Q =
4 sin
2
where is the scattering angle. It can be difficult to conceptualize
“inverse space” thus we relate Q to the real-space length scale,
Q
L
2
= .
One of the most widely used techniques in the analysis of SAS data is
Guinier Analysis. From the Guinier approximation (77), the scattering intensity I(Q)
is given by
) 3 / exp( ) 0 ( ) (
2 2
g
R Q I Q I =
where I(0) is the extrapolated intensity at Q = 0, and R
g
is the radius of gyration (78).
Thus, the radius of gyration can be determined from the slope of a plot of ln(I(Q)) vs.
Q
2
. The weight-average molecular weight of the protein can be determined using the
equation
I(0) =
M
w
c 2
( p
s
)
2
N
A
where S
and P
are the scattering length densities of the solvent and protein,
respectively, c is the protein concentration, and is the protein-specific volume
(79). A graph of I(0)/c vs MW is shown in Figure 1.12.
25
Figure 1.12: Scattering Intensity extrapolated to zero angle (I(0)) scaled by protein
concentration as a function of molecular weight
Scattering Intensity is additive, for a mixture of A and B, the scattering
intensity of the mixture, I
mix
, is equal to the summation of the scattering intensity
from species A, I
A
, and species B, I
B
, (I
mix
= I
A
+ I
B
). This allows the contribution
from each species to be deconvoluted out individually, or even “contrasted out”
through the use of different solvents.
The weight averaged molecular weight can be used to determine the fraction
of each component
=
=
=
L
i
i
L
i
i i
w
w
c
c M
M
1
1
where M
i
is the molecular weight of component i, c
i
is the concentration of
component i and L is the number of components.
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0 10203040 506070
0
c
MW (kDa)
ovalbumin
BSA
carbonic anhydrase
Ribonucleas A
lysozyme
lactalbumin
y = -0.0024734 + 0.0010399x
R= 0.99873
I(0) / c
26
If the scatterer (protein) is not globular then modified Guinier analysis must
be taken (in this case for rod-like structures)
Q*I(Q) =I
C
(0)exp( Q
2
R
C
2
/2)
where I
C
(0) is the extrapolated intensity at Q = 0, and R
C
is the cross-sectional radius
of gyration. Thus, R
C
can be determined from a plot of ln(Q*I(Q))
versus Q
2
. R
C
can
then be related to a geometric radius, R, 2
2 2
R R
C
= . Just like the original Guinier
analysis the molecular weight can be determine from the intercept, I
C
(0),
L N
C M
I
A
S P W
C
2
2
) (
) 0 (
=
where L is the length of the rod-like aggregate.
Another common method for SAS analysis is the calculation of a pair
distance distribution function (PDDF)(80)
=
max
0
) sin(
) ( 4 ) (
D
dr
Qr
Qr
r P Q I
where P(r) is the probability of finding two scattering centers at a distance r apart,
and D
max
is the maximum dimension within the scatterer (protein), see the left hand
side of Figure 1.13. The right hand side of Figure 1.13 illustrates how the pair
distance distribution functions calculated for a dimer depends on the orientation of
each monomer with respect to one another.
27
r
D
max
r
D
max
Figure 1.13: Left: representation of the alpha-chymotrypsin dimer, PDB code 6CHA,
illustrating the distance between two scattering centers, r, and the maximum dimension,
D
max
; Right: Examples of pair distance distribution functions for different dimer
orientations.(81)
The most novel SANS analysis technique is shape reconstruction, which
solves the “inverse scattering problem”. The inverse scattering problem refers to
determining the location of each scattering center within a protein from the scattering
of the protein as a whole, while the scattering problem does just the opposite (based
on the location of each scattering center the protein’s scattering intensity is
calculated). This involves the fitting of thousands of spherical scattering centers,
given only a few hundred data points. Scattering data files are fit by treating the
protein as a collection of n spherical scattering centers. The position of each
scattering center is changed until the simulated data is in good agreement with the
experimental data. Below (Figure 1.14) is an example of a fit at the beginning (left)
and at the end (right) of the iteration process. In obtaining the end result many
transitional structures (not shown) are encountered. The most commonly used
D
max
28
algorithms for this type of shape reconstruction are GA_STRUCT(82) and
GASBOR(83).
Figure 1.14: An example of a the GA_STRUCT fitting procedure, experimental scattering
data (blue), scattering calculated from the arrangement of scattering centers (red), scattering
centers (magenta). Initial guess (left), final fit (right).
1.4.3 Transmission Electron Microscopy (TEM)
The TEM operates on the same principles as a light microscope (see figure
1.15) but instead of using light it uses electrons, thus a much higher resolution is
achieved. Since the energy of an electron determines the wavelength of that
electron, the energy can be tuned to produce a specific wavelength. An electron
source located at the top of the microscope emits the electrons, which travel through
the column under vacuum. Electromagnetic lenses are used to focus the electrons,
the electron beam then travels through the sample. At the bottom of the microscope
there is a screen that fluoresces when hit by an electron. Depending on the density of
the sample, some electrons are scattered (not transmitted) and do not hit the screen.
Therefore an image of the sample appears on the screen. Different parts of the
sample display varied darkness depending on their density. The image can be viewed
29
directly by an observer or film can be inserted in the beam. Just like the fluorescent
screen, depending on sample density, the film is exposed to different degrees by the
electrons hitting it. In order to image biological samples they must first be dried
(since imaging occurs in a vacuum chamber) and then stained with heavy metals.
Staining with heavy metals enhances the structural detail of the sample as the dense
nuclei of such atoms (i.e. uranium) scatter electrons very strongly eliminating them
from the optical path. This is often a disadvantage as biological samples are very
sensitive to the surrounding environment. The application of heavy metals and/or
the drying of a sample can cause changes in structure and/or orientation of the
sample. However when the properties of the sample are carefully considered and the
sample preparation optimized TEM imagining can be a very powerful tool.
Figure 1.15: Diagram of a Transmission Electron Microscope.
30
1.4.4 Atomic Force Microscopy (AFM)
AFM uses a microcantilever with a very fine ceramic or semiconductor tip to
scan a surface. A laser beam is focused on the cantilever and as the tip is attracted to
or repelled by the surface this laser beam is deflected. A detector is aligned to
measure these deflections. Figure 1.16 shows the schematic set up of an AFM. A
plot of these deflections versus tip position will results in a topographic image of the
sample surface. In the case of soft samples (i.e. biological samples) direct contact
between the tip and the surface can cause changes in the sample, thus a tapping as
opposed to a contact mode is used. Here the tip is tapped across the surface instead
of being dragged across the surface. Unlike TEM staining is not necessary. AFM
also has the advantage of being performed in solution as well as under a variety of
different environmental conditions, allowing biological samples extremely sensitive
to drying techniques to be imaged. All of the samples imaged in this work are dried
and imaged under ambient temperature and pressure.
31
Figure 1.16: Schematic of an Atomic Force Microscope.
1.5 References
(1) Dobson, C. M. (1999) Protein misfolding, evolution and disease. Trends Biochem.
Sci. 24, 329-332.
(2) Horwich, A. (2002) Protein aggregation in disease: a role for folding intermediates
forming specific multimeric interactions. J. Clin. Invest. 110, 1221-1232.
(3) Bowie, J. U. (2001) Stabilizing membrane proteins. Curr. Opin. Struc. Biol 11, 397-
402.
(4) Brooks, C. L., III. (1998) Simulations of protein folding and unfolding. Curr. Opin.
Struc. Biol. 8, 222-226.
(5) Duan, Y., and Kollman, P. A. (1998) Pathways to a protein folding intermediate
observed in a 1-microsecond simulation in aqueous solution. Science 282, 740-744.
(6) Bachmann, A., Segel, D., and Kiefhaber, T. (2002) Test for cooperativity in the
early kinetic intermediate in lysozyme folding. Biophys. Chem. 96, 141-151.
32
(7) Jones, J. A., Wilkins, D. K., Smith, L. J., and Dobson, C. M. (1997) Characterization
of protein unfolding by NMR diffusion measurements. J. Biomol. NMR 10, 199-203.
(8) Uversky, V. N., Karnoup, A. S., Segel, D. J., Seshadri, S., Doniach, S., and Fink, A.
L. (1998) Anion-induced folding of Staphylococcal nuclease: characterization of
multiple equilibrium partially folded intermediates. J. Mol. Biol. 278, 879-894.
(9) Yokota, A., Hirai, K., Miyauchi, H., Iimura, S., Noda, Y., Inoue, K., Akasaka, K.,
Tachibana, H., and Segawa, S. (2004) NMR characterization of three-disulfide
variants of lysozyme, C64A/C80A, C76A/C94A, and C30A/C115A - a marginally
stable state in folded proteins. Biochemistry 43, 6663-6669.
(10) Chen, L., Hodgson, K. O., and Doniach, S. (1996) A lysozyme folding intermediate
revealed by solution x-ray scattering. J. Mol. Biol. 261, 658-672.
(11) Dobson, C. M. (1994) Solid evidence for molten globules. Curr. Biol. 4, 636-640.
(12) Gimel, J. C., and Brown, W. (1996) A light scattering investigation of the sodium
dodecyl sulfate-lysozyme system. J. Chem. Phys. 104, 8112-8117.
(13) Ibel, K., May, R. P., Kirschner, K., Szadkowski, H., Mascher, E., and Lundahl, P.
(1990) Protein-decorated micelle structure of sodium-dodecyl-sulfate - protein
complexes as determined by neutron scattering. Eur. J. Biochem. 190, 311-318.
(14) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2005) Photocontrol of Protein
Folding: The Interaction of Photosensitive Surfactants with Bovine Serum Albumin.
Biochemistry 44, 524-536.
(15) Matthews, C. R. (1993) Pathways of protein folding. Ann. Rev. Biochem. 62, 653-
683.
(16) Stenstam, A., Montalvo, G., Grillo, I., and Gradzielski, M. (2003) Small Angle
Neutron Scattering Study of Lysozyme-Sodium Dodecyl Sulfate Aggregates. J.
Phys. Chem. B 107, 12331-12338.
(17) Valstar, A., Brown, W., and Almgren, M. (1999) The Lysozyme-Sodium Dodecyl
Sulfate System Studied by Dynamic and Static Light Scattering. Langmuir 15, 2366-
2374.
(18) Ismail, A. A., Mantsch, H. H., and Wong, P. T. (1992) Aggregation of
chymotrypsinogen: portrait by infrared spectroscopy. Biochim. Biophys. Acta 1121,
183-188.
(19) Przybycien, T. M., and Bailey, J. E. (1989) Aggregation kinetics in salt-induced
protein precipitation. AIChE J. 35, 1779-1790.
(20) Ramakrishna, T., and Pandit, M. W. (1992) Involvement of active site in self-
association of proteins: a case of alpha-chymotrypsin. Indian J. Biochem. Bio. 29,
192-200.
33
(21) Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2007) Solution Structure of an
Amyloid-Forming Protein During Photoinitiated Hexamer-Dodecamer Transitions
Revealed Through Small-Angle Neutron Scattering. Biochemistry 46, 7694-7705.
(22) Hamill, A. C., and Lee, C. T., Jr. (in prep) Control of Ribonuclease A structure
through the use of photoresponsive surfactants.
(23) Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2005) Probing lysozyme
conformation with light reveals a new folding intermediate. Biochemistry 44, 15139-
15149.
(24) Hamill, A. C., and Lee, T. C., Jr. (in prep) -amyloid peptide (1-40) fibril growth
influenced by photosurfactant.
(25) Spreti, N., Alfani, F., Cantarella, M., D'Amico, F., Germani, R., and Savelli, G.
(1999) a-Chymotrypsin superactivity in aqueous solutions of cationic surfactants. J.
Mol. Catal. B-Enzym. 6, 99-110.
(26) Spreti, N., Di Profio, P., Marte, L., Bufali, S., Brinchi, L., and Savelli, G. (2001)
Activation and stabilization of a-chymotrypsin by cationic additives. Eur. J.
Biochem. 268, 6491-6497.
(27) Wang, S.-C., and Lee, C. T., Jr. (2006) Protein Secondary Structure Controlled with
Light and Photoresponsive Surfactants. J. Phys. Chem. B 110, 16117-16123.
(28) Wang, S. C., and Lee, C. T., Jr. (2007) Enhanced Enzymatic Activity Through
Photoreversible Conformation Changes Biochemistry 46, 14557-14566.
(29) Lee, P. K. J., and Jirgensons, B. (1971) Conformational transitions of
immunoglobulin fragment Fe(t) incited by alkyl sulfates of various hydrophobic
chain lengths. Biochim. Biophys. Acta 229, 631-641.
(30) Jansen, R., Dzwolak, W., and Winter, R. (2005) Amyloidogenic Self-Assembly of
Insulin Aggregates Probed by High Resolution Atomic Force Microscopy. Biophys.
J. 88, 1344-1353.
(31) Sunde, M., and Blake, C. (1997) The structure of amyloid fibrils by electron
microscopy and X-ray diffraction. Adv. Protein Chem. 50, 123-159.
(32) Hayashita, T., Kurosawas, T., Miyata, T., Tanaka, K., and Igawa, M. (1994) Effect
of structural variation within cationic azo-surfactant upon photoresponsive function
in aqueous solution. Colloid Polym. Sci. 272, 1611-1619.
(33) Shang, T., Smith, K. A., and Hatton, T. A. (2003) Photoresponsive Surfactants
Exhibiting Unusually Large, Reversible Surface Tension Changes under Varying
Illumination Conditions. Langmuir 19, 10764-10773.
34
(34) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2004) Photoreversible Viscosity
Changes and Gelation in Mixtures of Hydrophobically Modified Polyelectrolytes
and Photosensitive Surfactants. Macromol. 37, 5397-5405.
(35) Eastoe, J., and Vesperinas, A. (2005) Self-assembly of light-sensitive surfactants.
Soft Matter 1, 338-347.
(36) Morozova-Roche, L. A., Zurdo, J., Spencer, A., Noppe, W., Receveur, V., Archer,
D. B., Joniau, M., and Dobson, C. M. (2000) Amyloid Fibril Formation and Seeding
by Wild-Type Human Lysozyme and Its Disease-Related Mutational Variants. J.
Struc. Biol. 130, 339-351.
(37) Pepys, M. B., Hawkins, P. N., Booth, D. R., Vigushin, D. M., Tennent, G. A.,
Soutar, A. K., Totty, N., Nguyen, O., Blake, C. C. F., and et al. (1993) Human
lysozyme gene mutations cause hereditary systemic amyloidosis. Nature 362, 553-
557.
(38) Valleix, S., Drunat, S., Philit, J.-B., Adoue, D., Piette, J.-C., Droz, D., MacGregor,
B., Canet, D., Delpech, M., and Grateau, G. (2002) Hereditary renal amyloidosis
caused by a new variant lysozyme W64R in a French family. Kidney Int. 61, 907-
912.
(39) Griko, Y. V., Freire, E., Privalov, G., Van Dael, H., and Privalov, P. L. (1995) The
unfolding thermodynamics of c-type lysozymes: a calorimetric study of the heat
denaturation of equine lysozyme. J. Mol. Biol. 252, 447-459.
(40) Dinner, A. R., Sali, A., Smith, L. J., Dobson, C. M., and Karplus, M. (2000)
Understanding protein folding via free-energy surfaces from theory and experiment.
Trends Biochem. Sci. 25, 331-339.
(41) Kiefhaber, T. (1995) Kinetic traps in lysozyme folding. P. Nat. Acad. Sci. 92, 9029-
9033.
(42) Matagne, A., Radford, S. E., and Dobson, C. M. (1997) Fast and slow tracks in
lysozyme folding: insight into the role of domains in the folding process. J. Mol.
Biol. 267, 1068-1074.
(43) Radford, S. E., Dobson, C. M., and Evans, P. A. (1992) The folding of hen lysozyme
involves partially structured intermediates and multiple pathways. Nature 358, 302-
307.
(44) Gilquin, B., Guilbert, C., and Perahia, D. (2000) Unfolding of hen egg lysozyme by
molecular dynamics simulations at 300K: insight into the role of the interdomain
interface. Proteins 41, 58-74.
(45) Booth, D. R., Sunde, M., Bellotti, V., Robinson, C. V., Hutchinson, W. L., Fraser, P.
E., Hawkins, P. N., Dobson, C. M., Radford, S. E., Blake, C. C. F., and Pepys, M. B.
(1997) Instability, unfolding and aggregation of human lysozyme variants
underlying amyloid fibrillogenesis. Nature 385, 787-793.
35
(46) Canet, D., Sunde, M., Last, A. M., Miranker, A., Spencer, A., Robinson, C. V., and
Dobson, C. M. (1999) Mechanistic Studies of the Folding of Human Lysozyme and
the Origin of Amyloidogenic Behavior in Its Disease-Related Variants. Biochemistry
38, 6419-6427.
(47) Shih, P., Holland, D. R., and Kirsch, J. F. (1995) Thermal stability determinants of
chicken egg-white lysozyme core mutants: hydrophobicity, packing volume, and
conserved buried water molecules. Protein Sci. 4, 2050-2062.
(48) Raines, R. T. (1998) Ribonuclease A. Chemical Reviews (Washington, D. C.) 98,
1045-1065.
(49) Scheraga, H. A., Wedemeyer, W. J., and Welker, E. (2001) Bovine pancreatic
ribonuclease A: oxidative and conformational folding studies. Methods Enzymol.
341, 189-221.
(50) Harrington, W. F., and Schellman, J. A. (1956) Evidence for the instability of
hydrogen-bonded peptide structures in water, based on studies of ribonuclease and
oxidized ribonuclease. Compt. rend. trav. lab. Carlsberg, Ser. chim. 30, 21-43.
(51) Salahuddin, A., and Tanford, C. (1970) Thermodynamics of the denaturation of
ribonuclease by guanidine hydrochloride. Biochemistry 9, 1342-7.
(52) Blow, D. M. (1971) The structure of chymotrypsin, in The Enzymes (Boyer, P. D.,
Ed.) pp 185-212, Academic Press, New York.
(53) Dugas, H. (1996) Bioorganic Chemistry, 3rd. ed., Springer Verlag, New York.
(54) Hess, G. P. (1971) Chymotrypsin-chemical properties and catalysis, in The Enzymes
(Boyer, P. D., Ed.) pp 213-248, Academic Press, New York.
(55) Aune, K. C., and Timasheff, S. N. (1971) Dimerization of alpha-chymotrypsin. I. pH
dependence in the acid region. Biochemistry 10, 1609-1617.
(56) Kitano, H., Maeda, Y., and Okubo, T. (1989) Kinetic study of the effects of
solvation on the dimerization process of a-chymotrypsin. Biophys. Chem. 33, 47-54.
(57) Koren, R., and Hammes, G. G. (1976) A kinetic study of protein-protein
interactions. Biochemistry 15, 1165-1171.
(58) Pallares, I., Vendrell, J., Aviles, F. X., and Ventura, S. (2004) Amyloid Fibril
Formation by a Partially Structured Intermediate State of a-Chymotrypsin. J. Mol.
Biol. 342, 321-331.
(59) Czapinska, H., and Otlewski, J. (1999) Structural and energetic determinants of the
S1-site specificity in serine proteases. Eur. J. Biochem. 260, 571-595.
36
(60) Chiti, F., Taddei, N., Bucciantini, M., White, P., Ramponi, G., and Dobson, C. M.
(2000) Mutational analysis of the propensity for amyloid formation by a globular
protein. EMBO J. 19, 1441-1449.
(61) Srisailam, S., Kumar Thallampuranam Krishnaswamy, S., Rajalingam, D., Kathir
Karuppanan, M., Sheu, H.-S., Jan, F.-J., Chao, P.-C., and Yu, C. (2003) Amyloid-
like fibril formation in an all beta-barrel protein. Partially structured intermediate
state(s) is a precursor for fibril formation. J. Bio. Chem. 278, 17701-9.
(62) Pallares, I., Vendrell, J., Aviles, F. X., and Ventura, S. (2004) Amyloid Fibril
Formation by a Partially Structured Intermediate State of a-Chymotrypsin. J. Mol.
Biol. 342, 321-331.
(63) Bacaloglu, R., Bunton, C. A., Cerichelli, G., and Ortega, F. (1989) NMR study of
the location of bromide ion and methyl naphtalene-2-sulfonate in cationic micelles:
reaction to reactivity. J. Phys. Chem. 93, 1490-1497.
(64) Bonan, C., Germani, R., Ponti, P. P., Savelli, G., Cerichelli, G., and Bunton, C. A.
(1990) Micellar head-group size and anion nucleophilicity in SN2 reactions. J. Phys.
Chem. 94, 5331-5336.
(65) Bacaloglu, R., Bunton, C. A., and Ortega, F. (1989) Micellar enhancements of rate
of SN2 reactions of halide ions: the effect of head group size J. Phys. Chem. 93,
1497-1502.
(66) Wang, D., Bode, W., and Huber, R. (1985) Bovine chymotrypsinogen A. X-ray
crystal structure analysis and refinement of a new crystal form at 1.8 Å resolution. J.
Mol. Biol. 185, 595-624.
(67) Egan, R., Michel, H. O., Schlueter, R., and Jandorf, B. J. (1957) Physicochemical
investigation of the chymotrypsins. II. On the mechanism of dimerization of
chymotrypsin. Arch. Biochem. Biophys. 66, 366-373.
(68) Rao, M. S. N., and Kegeles, G. (1958) An ultracentrifuge study of the
polymerization of a-chymotrypsin. J. Am. Chem. Soc. 80, 5724-5729.
(69) Winzor, D. J., and Scheraga, H. A. (1964) Chemically reacting systems on
Sephadex. II. Molecular weights of monomers in rapid association equilibrium. J.
Phys. Chem. 68, 338-343.
(70) Glenner, G. G., Wong, C. W. (1984) Alzheimer's Disease: initial report of the
purification and characterization of a novel cerebrovascular amyloid protein.
Biochem. Biophys. Res. Commun. 120.
(71) Krueger, S., Ho, D., and Tsai, A. (2006) Small-Angle Neutron Scattering as a Probe
for Protein Aggregation at Many Length Scales, in Misbehaving Proteins. Protein
(Mis)Folding, Aggregation, and Stability (Murphy, R. M., Tsai, A. M., Ed.) pp 125-
146, Springer, New York.
37
(72) Yong, W., Lomakin, A., Kirkitadze, M. D., Teplow, D. B., Chen, S.-H., and
Benedek, G. B. (2002) Structure Determination of Micelle-like Intermediates in
Amyloid -protein Fibril Assembly by using Small Angle Neutron Scattering. P.
Nat. Acad. Sci. 99, 150-154.
(73) Gosal, W. S., Myers, S. L., Radford, S. E., Thomson, N. H. (2006) Amyloid Under
the Atomic Force Microscope. Protein Pept. Lett. 13, 261-270.
(74) Hansma, H. G., Hoh, J. H. (1994) Biomolecular imaging with the atomic force
microscope. Annu. Rev. Biophys. Biomol. Struct. 23, 115-139.
(75) Yang, J., Sho, Z. (1995) Recent advances in biological atomic force microscopy.
Micron 26.
(76) Fulop, L. Z., M., Datki, Z., Soos, K., and Penke, B. (2004) -Amyloid-derived
Pentapeptide RIIGLa Inhibits A (1-42) Aggregation and Toxicity. Biochem. and
Biophys. Res. Comm. 324, 64-69.
(77) Guinier, A. a. F., G. (1955) Small Angle Scattering of X-ray, Wiley, New York.
(78) Gallas, J. M., Littrell, K. C., Seifert, S., Zajac, G. W., and Thiyagarajan, P. (1999)
Solution structure of copper ion-induced molecular aggregates of tyrosine melanin.
Biophys. J. 77, 1135-42.
(79) Bu, Z., and Engelman, D. M. (1999) A method for determining transmembrane helix
association and orientation in detergent micelles using small angle X-ray scattering.
Biophys. J. 77, 1064-1073.
(80) Glatter, O. (1979) The interpretation of real-space information from small-angle
scattering experiments. J. App. Cryst. 12, 166-175.
(81) Glatter, O. (1977) A new method for the evaluation of small-angle scattering data. J.
Appl. Cryst. 10, 415-421.
(82) Heller, W. T., Krueger, J. K., and Trewhella, J. (2003) Further Insights into
Calmodulin-Myosin Light Chain Kinase Interaction from Solution Scattering and
Shape Restoration. Biochemistry 42, 10579-10588.
(83) Svergun, D. I., Petoukhov, M. V., and Koch, M. H. J. (2001) Determination of
domain structure of proteins from x-ray solution scattering. Biophys. J. 80, 2946-
2953.
38
CHAPTER II: Probing Lysozyme Conformation with Light Reveals
a New Folding Intermediate
Andrea C. Hamill, Shao-Chun Wang, and C. Ted Lee, Jr.
Department of Chemical Engineering, University of Southern California, Los
Angeles, California 90089-1211
Biochemistry 2005, 44, 15139-15149
Received August 18, 2005; Revised Manuscript Received September 20, 2005
2. 1 Abstract
A means to control lysozyme conformation with light illumination has been
developed using the interaction of the protein with a photoresponsive surfactant.
Upon exposure to the appropriate wavelength of light, the azobenzene surfactant
undergoes a reversible photoisomerization, with the visible-light (trans) form being
more hydrophobic than the UV-light (cis) form. As a result, surfactant binding to the
protein and, thus, protein unfolding, can be tuned with light. Small-angle neutron
scattering (SANS) measurements were used to provide detailed information of the
protein conformation in solution. Shape-reconstruction methods applied to the SANS
data indicate that under visible light the protein exhibits a native-like form at low
surfactant concentrations, a partially swollen form at intermediate concentrations,
and a swollen/unfolded form at higher surfactant concentrations. Furthermore, the
SANS data combined with FT-IR spectroscopic analysis of the protein secondary
structure reveal that unfolding occurs primarily in the -domain of lysozyme, while
39
the -domain remains relatively intact. Thus, the surfactant-unfolded intermediate of
lysozyme appears to be a separate structure than the well-known -domain
intermediate of lysozyme that contains a folded -domain and unfolded -domain.
Because the interactions between the photosurfactant and protein can be tuned with
light, illumination with UV light returns the protein to a native-like conformation.
Fluorescence emission data of the non-polar probe Nile red indicate that hydrophobic
domains become available for probe partitioning in surfactant-protein solutions
under visible light, while the availability of these hydrophobic domains to the probe
decrease under UV light. Dynamic light scattering and UV-vis spectroscopic
measurements further confirm the shape-reconstruction findings and reveal three
discrete conformations of lysozyme. The results clearly demonstrate that visible light
causes a greater degree of lysozyme swelling than UV light, thus allowing for the
protein conformation to be controlled with light.
2.2 Introduction
The biological function of a protein is largely determined by the structure of
the protein, and thus, much work has been aimed at investigating this form-function
relationship. The method of choice in this regard has been X-ray crystallography,
which has been shown to be quite successful at elucidating the structure of soluble
proteins in the native state. However, the form-function relationship is not static in
nature, and instead, it is often necessary for the protein to adopt non-native or
partially folded conformations to perform certain functions, whether beneficial (e.g.,
40
enzymatic catalysis, transport across cellular membranes, immune protection, etc.) or
harmful [e.g., ailments such as Alzheimer’s disease, cystic fibrosis, Mad Cow
disease, amyloid diseases, prion diseases, and even many cancers are believed to
result from misfolded proteins(1, 2)]. In the specific case of lysozyme, mutations of
the protein have been found to produce structural transitions that lead to
amyloidosis(3), with large quantities, sometimes kilograms, of aggregated protein
accumulating in organs such as the liver, kidney, and spleen(4, 5).
Despite the fact that intermediately folded states represent key structures
along folding pathways, high-resolution structural information of unfolded or
partially folded states is limited, largely because intermediately folded proteins
typically aggregate over the long times required for crystallization, a result of
exposure of the hydrophobic regions of the protein to solvent(6). Thus, developing
novel methods of investigating intermediate states in solution, as opposed to in the
crystalline form, is warranted. However, direct high-resolution structural studies of
these intermediates are usually precluded because of their transient nature (lifetimes
< 100 ms), thereby relegating the investigation of these kinetic intermediates
primarily to the field of simulations(7, 8), although novel experimental strategies
have been developed to study the folding process, such as time-resolved small angle
X-ray scattering, nuclear magnetic resonance (NMR), and various spectroscopic
techniques combined with stoppedflow methods(9-12). An alternate approach is to
study equilibrium intermediates, partially folded states stabilized by selected
solvents, denaturants, or surfactants, because a variety of evidence suggests that
41
equilibrium intermediates are similar in conformation to the corresponding kinetic
intermediates(13-20). In this manner, “snapshots” along the protein-folding pathway
can be stabilized as equilibrium intermediates, allowing an investigation of the
mechanisms of protein folding.
It is generally observed that c-type lysozymes do not exhibit intermediately
folded states during equilibrium folding experiments(21), except for extreme
conditions such as low pH or when mixed with a solvent. In kinetic folding
experiments, however, a populated intermediate has been observed(22-25). Upon
initiating refolding from the denatured state, lysozyme rapidly (within a few
milliseconds) forms an ensemble of different hydrophobically collapsed states(11).
After this rapid collapse, the majority of the molecules (70%) accumulate into a
stable kinetic intermediate, often termed the -intermediate (I
), with a structured
domain and disordered domain. The rate-limiting step for formation of the native
state is then refolding of the domain, allowing formation of the hydrophobic
interface between the and domains, with molecular dynamics simulations(26)
and mutation studies(27-29) suggesting that the final step in the folding process may
be insertion of the domain residues Leu-55 and Ile-56 into a hydrophobic pocket of
the domain. Of the remaining 30% of molecules that do not fold via the formation
of the I
intermediate, ~2/3 fold very rapidly as the and domains form
simultaneously (I
/ intermediate), while 1/3 fold very slowly, potentially limited by
a cis-trans isomerization of proline.
42
The manner in which a protein folds depends upon the interplay between
electrostatic, hydrogen-bonding, van der Waals, and hydrophobic interactions among
the amino acid residues making up the protein. In the native state, the charged and
polar amino acid groups typically reside on the exterior of the protein exposed to
water, while the nonpolar amino acid groups are largely found inside the folded
structure of the protein, protected from unfavorable solvent interactions. Protein
unfolding can then be induced to different extents by changing variables such as pH,
temperature, pressure, or through the addition of a chemical denaturant (i.e., urea or
GdmCl)(13, 15). Surfactants have also been shown to unfold proteins(19, 20),
including lysozyme, where the hydrophobic moieties of the surfactant cause
unfolding by interacting with nonpolar amino acids, thereby eliminating unfavorable
solvent contacts. Because surfactant hydrophobicity increases as the length of the
hydrocarbon tail increases, it has been found that a greater degree of protein
unfolding occurs with increased surfactant molecular weight for a given
headgroup(30).
In a previous study, it was shown that with the use of a photoresponsive
surfactant (“azoTAB”), simple light illumination could be used to induce reversible
changes in the conformation of proteins(19). An azobenzene group in the
hydrocarbon tail allowed the surfactant hydrophobicity to be tuned with light, with
the visible-light (trans) isomer being more hydrophobic and thus having a higher
affinity to bind to proteins than the UV-light (cis) form. Small-angle neutron
scattering (SANS) data combined with shape reconstruction techniques (to be
43
discussed below) were used to determine the structure of bovine serum albumin
(BSA) in solution, and three conformations of the protein were observed. At low
surfactant concentrations, the native, heartshaped structure (N form) of BSA was
evident. However, upon increasing the azoTAB concentration under visible light, the
C-terminal portion of the protein was seen to separate from the molecule, giving rise
to a partially unfolded intermediate structure (F form). This unfolding mechanism
had been suspected on the basis of indirect evidence (e.g., UV-vis, CD, and FT-
IR)(31), now “visualized” from the SANS data. As the surfactant concentration was
further increased under visible light, a highly unfolded/elongated conformation (E
form) was detected, although a significant R-helical content still remained, which
was observed as “kinks” in the protein chain from the shape-reconstructed SANS
fits. It was also shown that light illumination could be used to reversibly switch
between these folded forms.
In the present study, we demonstrate that intermediately folded states of
lysozyme can be stabilized in solution with the azoTAB surfactant and that light
illumination can be used to induce changes between these conformations. SANS,
dynamic light scattering (DLS), FT-IR, UV-visible, and fluorescence spectroscopic
measurements are performed on the lysozyme-azoTAB system to investigate the
folding of the protein. At low surfactant concentrations, native-like structures of
lysozyme are observed, while increasing the azoTAB concentration under visible
light causes the domain of the protein to swell, resulting in a partially unfolded
structure at intermediate surfactant concentrations, and to eventually form a
44
swollen/unfolded form of lysozyme at high concentrations. Under UV light at all
azoTAB surfactant concentrations, however, the protein appears to be in a native-like
state; thus, light can be used to manipulate the protein, allowing for reversible
control of protein folding.
2.3 Experimental Methods
2.3.1 Materials
An azobenzene-trimethylammonium bromide surfactant (azoTAB) of the
form
similar to the surfactant used in a previous study(19), was synthesized according to
published procedures(32, 33) via an azo-coupling reaction of 4-ethlyalaniline with
phenol subsequently followed by reactions with 1,2 dibromoethane and finally
trimethylamine. The surfactant undergoes a photoisomerization when illuminated
with UV light that is fully reversible with exposure to visible light(19, 34). The
visible-light form of the surfactant is primarily in the trans state (75:25 trans/cis
ratio), while the UV-light form is mostly in the cis state (>90% cis)(34).
Photoisomerization with exposure to light changes the dipole moment across the -
N=N- bond in the surfactant. The visible-light (trans, planar) form of the surfactant
has a lower dipole moment and is therefore more hydrophobic than the UV-light (cis,
45
bent) form, resulting in the UV-light form having a higher affinity for water and the
visible-light form a higher affinity for protein binding. Two illumination methods
were used for surfactant photoisomerization. For DLS, FT-IR, and UV-visible
spectroscopy, conversion to the cis form was achieved with the 365 nm line from a
200 W mercury arc lamp (Oriel, model number 6283), isolated with the combination
of a 320 nm band-pass filter (Oriel, model number 59 800) and a heat-absorbing
filter (Oriel, model number 59 060). A 400 nm long-pass filter (Oriel, model number
59 472) was used to convert back to the trans form. Before measurements, solutions
were illuminated with this lamp for at least 10 min. For FT-IR and DLS, light
exposure was continued throughout the entire collection of data (hours) by directly
illuminating the samples using a liquid light guide (Oriel, model number 77 557). In
the small-angle neutron scattering and fluorescence spectroscopy experiments, the
solutions were exposed to an 84 W long wave UV lamp, 365 nm (Spectroline, model
number XX-15A), for at least 30 min to convert to the cis form; to convert back to
the trans form, the samples were simply left in room light. For the neutron scattering
experiments, the samples were continuously exposed to the same UV light
throughout the data collection (to maintain the cis form). Note that 365 nm
corresponds to UV-A radiation, as opposed to UV-C, which is know to inactivate
lysozyme(35). Control experiments confirmed that neither UV nor visible light
affected the structure of lysozyme ( A < 1% throughout the amide I region from
FT-IR).
46
Highly purified and lyophilized hen-egg-white lysozyme (catalog number L-
7651) and a low ionic strength phosphate buffer (pH 7.2, 8.3 mM) were purchased
from Sigma; all other chemicals were obtained from Aldrich in the highest purity.
Because of the relatively small size of lysozyme, a protein concentration of 11.8
mg/mL was used in the neutron scattering measurements to achieve a reasonable
scattering intensity (e.g., each spectra took ~4 h to collect). Additionally, higher
protein concentrations were not studied to avoid the formation of lysozyme dimers,
which are well-known to begin to form beyond this concentration at neutral pH. For
DLS (and subsequent UV-vis) experiments, a protein concentration of 1.0 mg/mL
was employed to approximate the diffusion coefficient at infinite dilution.
2.3.2 Small Angle Neutron Scattering
The neutron scattering data were collected on the 30-m NG-3 SANS
instruments at NIST(36) using a neutron wavelength of = 6 Å. Two sample-
detector distances were used, 1.33 and 7.0 m, with a detector offset of 25 cm to
produce a Q range of 0.0048-0.46 Å
-1
, where Q = 4 -1
sin( /2) and is the scattering
angle. The net intensities were corrected for the background and empty cell (pure
D2O), followed by accounting for the detector efficiency using the scattering from
an isotropic scatterer (Plexiglas) and then converted to an absolute differential cross
section per unit sample volume (in units of cm
-1
) using an attenuated empty beam.
The incoherent scattering intensities from the hydrogen atoms in lysozyme (0.004
47
cm
-1
) and the surfactant (0-0.0012 cm
-1
) were subtracted from the experimental
intensities to obtain the coherent scattering from each sample.
The SANS data were analyzed using several complementary techniques to
develop consistent conclusions and to cross-check the data, including Guinier
analysis, a shape-reconstruction algorithm, and calculation of the pair distance
distribution functions (PDDFs). The shape reconstructions were performed using the
GA_STRUCT program(37). The data files are fit by treating the protein as a
collection of 1000 spherical scattering centers, the positions of which were adjusted
until the simulated data agreed with the experimental data. The Q range used for the
data fits was 0.01 < Q < 0.3 Å
-1
. This range was used to exclude potential protein
interactions that would be exhibited at low Q and to avoid length scales too small for
protein continuity at high Q.
The PDDF is a measure of the probability P(r) of finding two scattering
centers at a distance r apart(38, 39); thus, information about protein conformation
can be readily obtained from the PDDF using a model-independent, straightforward
procedure. PDDFs were calculated assuming a monodisperse system using
GNOM(40) over a Q range of 0.02-0.3 Å
-1
, again to exclude interactions at low Q.
The maximum particle diameter, D
max
, was selected to be the lowest value that gave
a smooth return of the PDDF to zero at D
max
.
48
2.3.3 Dynamic Light Scattering
DLS measurements were performed at 25°C on a Brookhaven model BI-
200SM instrument (Brookhaven Instrument Corp.) with a 35 mW (Melles Griot,
model number 05-LHP-928) helium neon ( =632.8 nm) laser. The scattered light
was collected at an angle of 30° to decrease Q and, hence, increase the decay time,
thereby allowing the full correlation function, even for a small protein such as
lysozyme, to be observed. The laser wavelength was high enough to not convert the
azoTAB cis isomer into the trans form during the course of the experiments, thus,
allowing long sample times (up to 120 min). The data were analyzed with both the
NNLS and CONTIN routines (difference < 2 Å) using a BI-9000AT digital
correlator (Brookhaven Instrument Corp.). Before each measurement, the protein-
surfactant solutions were passed through a 0.2 μm PVDF filter and then through a
0.02 μm Anatop filter. UV-vis spectra were taken after filtration to determine both
the surfactant and protein concentrations. The experimental results represent the
average obtained from two runs using independent solutions, both of which produced
analogous results ((5%). The hydrodynamic radius, R
H
, was calculated assuming a
spherical shape, according to the Stokes-Einstein equation, R
H
= kBT/6 D, where
kB is Boltzmann’s constant, T is the temperature, is the viscosity of the solvent,
and D is the experimentally determined diffusion coefficient.
49
2.3.4 FT-IR Spectroscopy
FT-IR measurements were performed on a Genesis II FT-IR system (Mattson
Instruments). For each spectrum, 250 interferograms were collected with a 4 cm
-1
resolution. The protein and surfactant were dissolved in buffered D
2
O at room
temperature for a minimum of 24 h before experimentation. Each sample solution
was placed in a demountable liquid cell equipped with CaF
2
windows and a 50 μm
Teflon spacer. The sample compartment was continuously purged with CO
2
-free dry
air for 1 h before data collection to avoid any complication to the spectra from water
vapor. IR spectra of solutions without protein were measured under identical
conditions and subtracted from the original spectra of solutions containing protein to
remove surfactant peaks at ~1600 cm
-1
. The spectra were Fourier self-deconvoluted
with a K factor = 2.0.
2.3.5 UV-visible Spectroscopy
Absorption measurements were performed on an Agilent model 8453 UV-
visible spectrophotometer. Cells with a path length of 2 mm were used. When
employed, a crystal violet concentration of 10 μM was used. Crystal violet exhibits a
maximum absorbance at 590 nm in pure water, far enough away from the azoTAB
absorbance(34) to be readily detected in azoTAB solutions. The wavelength at which
the maximum absorption of crystal violet occurred for each sample was estimated by
fitting each absorption spectrum with a fifth-order polynomial.
50
2.3.6 Fluorescence Spectroscopy
Fluorescence measurements using Nile red as the probe were performed on a
Quanta-Master spectrofluorometer model QM-4 (Photon Technology International)
at 25 °C. The results were obtained with an excitation wavelength of 560 nm and an
emission wavelength of 650 nm, with excitation and emission slit widths of 4 nm.
The spectrofluorometer was loaded with 3 mL of a surfactant-protein solution
([protein] = 10.0 mg/mL), and 30 μL of a 1.0 mM Nile red solution in ethanol was
then pipetted into each sample. After this, the samples were stirred for 30 min to
reach steady state, as assured by monitoring the change in fluorescence with time. To
avoid potential photodegredation of Nile red upon UV light illumination(41), when
necessary, the lysozyme-zoTAB solutions were pre-exposed to UV light for 2 h
before the addition of Nile red. UV-visible spectroscopic measurements were then
taken immediately following the fluorescence scan to ensure that the surfactant
remained in the UV-light form during the course of the measurement.
2.4 Results and Discussion
The ability to control the lysozyme structure with light is shown in the SANS
data in Figure 2.1 as a function of the azoTAB concentration and light illumination.
The broad Q range of neutron scattering, in this case Q = 0.0048-0.46 Å
-1
, allows a
thorough investigation of protein-folding phenomena by simultaneously
investigating a wide range of length scales, L 2 /Q (L = 13.7-1300 Å). As seen in
Figure 2.1, at a surfactant concentration of either 7.9 or 12.2 mM azoTAB under
51
visible light illumination, the data deviate from the remaining scattering curves
beginning at ~Q = 0.2 Å
-1
or L = 31 Å. When compared to the literature value
reported for the diameter of lysozyme (~36 Å)(18) and to our own value determined
by DLS (35 Å, see below), the deviation at this Q value suggests that under visible
light azoTAB causes the protein to “swell”. However, illuminating these solutions
with UV light results in spectra similar to that of pure lysozyme, indicating that the
protein refolds back to a size similar to the native state when exposed to UV light.
Figure 2.1: SANS data of lysozyme-azoTAB solutions as a function of surfactant
concentration and light illumination. [lysozyme] = 11.8 mg/mL
52
To quantify the effect of light and surfactant on lysozyme folding, Guinier
analysis of the SANS data was employed. From the Guinier approximation(42), the
scattering intensity I(Q) is given by
) 3 / exp( ) 0 ( ) (
2 2
g
R Q I Q I =
where I(0) is the extrapolated intensity at Q = 0 and R
g
is the radius of gyration(43).
Thus, the radius of gyration can be determined from the slope of a plot of ln(I(Q))
versus Q
2
in the region where QR
g
~ 1. Values of the radius of gyration calculated in
this manner are shown in Table 2.1. For the “native-like” conformations of pure
lysozyme, 5.1 mM surfactant under visible light, and all UV-light surfactant
concentrations, the R
g
values are in good agreement with the literature value for R
g
measured with SANS in D
2
O (13.3 Å)(44). For 7.9 and 12.2 mM azoTAB under
visible light, however, Rg increases to 14.3 and 14.7 Å, respectively. While this
roughly 10% increase in the radius of gyration is not as large as the increase
observed upon denaturation induced thermally (20%)(45) or upon the addition of
sodium dodecyl sulfate (30% increase for SDS/lysozyme ~ 300:1, compared to a
maximum azoTAB/lysozyme ~ 15:1 in Figure 1)(17), urea (40%)(9, 14), or methanol
(40%)(46), it does, however, support the idea that lysozyme swells under visible
light with the azoTAB surfactant and can further refold upon illumination with UV
light to a size akin to the native state. Similarly, the radius of gyration of the I
kinetic intermediate has been observed to be 9% larger than the native state with
time-resolved small-angle X-ray scattering experiments upon refolding lysozyme
from a GdmCl-unfolded state(47).
53
Table 2.1: Values of the Radius of Gyration and I(0) determined from Guinier Analysis of
the SANS data in Figure 2.1, as well as the protein maximum dimension (± 3 Å) obtained
from the Pair Distance Distribution Function.
From the Guinier fits, I(0) was also determined and then used to obtain the
molecular weight of the protein to examine the possibility of protein self-association.
The weight-average molecular weight of the protein can be determined using the
equation
A
s g w
N
c M
I
2
2
) (
) 0 (
=
where s
and p
are the scattering length densities of the solvent and protein,
respectively, c is the protein concentration, and is the protein-specific volume(48).
As shown in Table 2.1, the values of I(0) determined from the data in Figure 2.1 for
the native-like conformations (0.124 ± 0.008 cm
-1
) agrees nicely with that calculated
from the amino acid sequence using the Debye equation [I(0) = 0.128 cm
-1
](49),
indicating that self association does not occur.
54
The I(0) value is very consistent for all but 7.9 and 12.2 mM under visible
light, where an approximately 40 and 80% increase in I(0) is observed, respectively.
This increase may appear to indicate protein self-association; however, when
considered along with the increase in R
g
, it can be seen that this is not the case. For
globular proteins(50) (see PDDFs below), R
g
Mw
0.369
; thus, an 80% increase in the
protein average molecular weight (from pure monomer to a mixture of monomer and
oligomers) would be expected to result in a 24% increase in R
g
, as opposed to the
10% observed. Therefore, the increase in I(0) in Figure 2.1 does not appear to result
from protein association. Instead, this increase in I(0) is likely to be a result of an
increase in the effective molecular weight of the surfactant-protein complex [15 or
30 bound surfactant molecules would explain the 40 or 80% increase in I(0),
respectively].
While the above results demonstrate that light illumination combined with
photoresponsive surfactants can be used to control lysozyme conformation, the
values of R
g
obtained from the Guinier analysis only provide low-resolution
information on the folding process. To gain more insight into this phenomena, a
shape-reconstruction algorithm(37) was applied to the SANS data by treating the
protein as a collection of 1000 scattering centers with positions that are adjusted to
the best fit of the data. Shown in Figure 2.2 are the structures for the runs best fitting
each set of data (in blue), along with the resulting “consensus envelopes” (in red)
obtained by averaging the 10 runs at each surfactant concentration. Shape-
reconstruction techniques require the fitting of the positions of thousands of
55
scatterers with only a few hundred data points; thus, the structures obtained for
different runs are not unique and instead depend upon the path of the fits. The
consensus envelopes indicate the consistency of the multiple runs.
Figure 2.2: Conformation of lysozyme in solution determined from shape reconstruction of
the SANS data (best fit in blue, consensus envelope representing the average of 10 runs in
red), compared with the X-ray crystallographic structure(51) (PDB code 6LYZ). The -
domain is shown to consist of helices A (residues 4-15), B (residues 24-36), C (residues 89-
99), and D (residues 108-115), along with a 310 helix (residues 120-124)(26). Also shown
are the structures of the folded state and the I
intermediate determined from molecular
dynamic simulations(25). [lysozyme] = 11.8 mg/mL.
As seen in Figure 2.2, all of these shape-reconstruction fits, except 7.9 and
12.2 mM azoTAB under visible light, look very similar in both size and shape and
agree quite well with the X-ray crystallographic structure (PDB code 6LYZ)(51).
56
This is to be expected because the conformation of soluble proteins in the native
state, in solution generally agree with the respective crystal structures(37, 49). For
lysozyme, this similarity between the solution and crystal structure has been obtained
using several techniques such as X-ray and neutron scattering, 2D-NMR, molecular
dynamic simulation, differential scanning calorimetry, and laser Raman scattering.
The solution structure determined from the SANS data for pure lysozyme, while not
providing for nearly as high of a resolution as X-ray crystallography, does allow the
active-site cleft to be observed in the upper right of the molecule between the and
-domains. Furthermore, upon increasing the surfactant concentration under visible-
light illumination, lysozyme is observed to unfold, represented most notably by a
progressive swelling of the lower-left side of the protein in Figure 2.2, although
because of rotational averaging as a result of the protein being in solution, it is
impossible to assign which domain ( or ) is unfolding from the SANS data alone,
as will be discussed below.
PDDFs were calculated from the experimental SANS data to further
investigate the effects of surfactant and light on the protein structure. A PDDF is the
measure of the probability, P(r), of finding two scattering centers (i.e., atomic nuclei)
at a distance r apart(38, 39), and calculation of a PDDF from small-angle scattering
data is a common method of examining the protein structure(19, 20), providing a
simple, model-independent method to probe the myriad of complex structures that
proteins can adopt in solution. The results in Figure 2.3 show that the maximum of
the PDDF curve shifts to higher values of r for 7.9 and 12.2 mM azoTAB under
57
visible light. For globular proteins, the maximum of the PDDF (i.e., the most
probable dimension) is approximately the protein radius, once again indicating that
lysozyme is swollen under visible light for 7.9 and 12.2 mM azoTAB. Furthermore,
the PDDF curves at these two concentrations also indicate a higher probability of this
most common dimension and are more symmetric about P(r)max, implying that the
protein-surfactant complex is becoming more “globular”(19). Furthermore, the point
at which P(r) returns to 0 at large r values defines the maximum dimension within
the protein, D
max
. As shown in Table 2.1, the D
max
values for 7.9 and 12.2 mM under
visible light are very close to that of pure lysozyme, confirming that the protein does
not dramatically elongate but rather “swells” with azoTAB addition. Comparing
these model-independent PDDFs to Figure 2.2 provides a consistency check of the
shape-reconstruction analysis. The shape-reconstruction fits show a swelling of the
protein, giving rise to the observed increase in r at P(r)max, while only a slight
increase in the protein length is seen in Figure 2.2, supported by the fact that the
location of D
max
changes little with surfactant or light and generally agrees with D
max
for pure lysozyme.
58
Figure 2.3: PDDFs of lysozyme-azoTAB solutions as a function of the surfactant
concentration and light illumination. [lysozyme] = 11.8 mg/mL
The lack of shoulders or secondary peaks in the calculated pair distance
distribution functions also confirm that protein self-association (i.e., the forming of
n-mers) does not occur. For example, in the case of dimer formation, a secondary
peak would develop at an r value equal to the distance between the centers of each
monomer, clearly not the case in Figure 2.3. In terms of larger protein aggregates, a
modest increase in the scattering intensity can be seen in Figure 2.1 at very low Q
(<0.01 Å
-1
), which could possibly indicate the formation of a small amount of such
aggregates(18). An alternative explanation for this increase could be an imperfect
59
correction of the scattering from the solvent (D2O), which was also found to exhibit
a ~2-fold increase in scattering intensity below Q = 0.01 Å
-1
[data not shown, see
also Svergun and Koch(52)]. Note that a slight increase in scattering at low Q has
also been observed in similar polyelectrolyte systems(19, 49). Thus, the results
obtained from the various techniques used to analyze the SANS data support the
conclusion that any aggregation corresponds to a very small fraction of the total
lysozyme.
To determine whether Figure 2.2 represents a swelling of the or -domain
of lysozyme, changes in the lysozyme secondary structure were detected with FT-IR
spectroscopy in the amide I region (1700-1600 cm
-1
, due mainly to C=O stretching).
Figure 2.4 shows the FT-IR spectra of pure lysozyme and lysozyme with 5.0 mM
azoTAB under both visible and UV light, all in deuterated buffer with 10.0 mg/ mL
protein. Seven peaks were generally observed in the FSD spectra and assigned to
secondary structure elements according to well-established protocols(53): 1683 cm
-1
( -turn), 1674 cm
-1
( -turn), 1666 cm
-1
( -turn), 1653 cm
-1
( -helix), 1641 cm
-1
(unordered structures), 1630 cm
-1
( -sheet), and 1610 cm
-1
(side chain). Upon the
addition of the surfactant, the spectra show a clear decrease in the -helix peak at
1653 cm
-1
and a growth in the peak at 1641 cm
-1
representing unordered structures,
while the peaks representing the -structures remain relatively unchanged. This
indicates that the azoTAB surfactant swells the -domain of the protein, a result of a
helix-to-unordered transition. Furthermore, under visible light, the intensity of the
peak at 1653 cm
-1
is smaller than under UV light, indicating that the visible-light
60
form of the surfactant results in a greater degree of unfolding of the -domain.
Similar conclusions regarding the surfactant primarily leading to unfolding of the -
domain of lysozyme have been observed in the lysozyme-SDS and lysozyme-
cetyltrimethylammonium bromide (CTAB) systems, with the helix-to-unordered
transition generally occurring with the addition of 10-40 surfactant molecules/protein
at near neutral pH(54, 55), similar to the surfactant/protein ratio in Figure 2.4 (~7).
When the 5.1 mM azoTAB data in Figure 2.2 is compared with the data in Figure
2.4, it is clear that changes in the protein secondary structure are observed with FT-
IR before significant changes in the tertiary structure (overall shape) are detected
with SANS. Furthermore, at azoTAB concentrations higher than 5 mM, there was no
observed difference between FT-IR spectra collecting under visible and UV light
(data not shown), despite the fact that the SANS data demonstrate considerable
changes in the tertiary structure. Similar trends have been observed in the BSA-
azoTAB system when comparing SANS and FT-IR data(56).
61
Figure 2.4: Fourier self-deconvoluted FT-IR spectra of lysozyme in the presence of 5.0mM
azoTAB as a function of the light illumination. [lysozyme] = 10.0 mg/mL.
From Figures 2.2 and 2.4, it appears that with the addition of the surfactant
the lysozyme adopts an intermediately folded state unique from the I
intermediate
mentioned above. This “surfactant-unfolded intermediate (I
S
)” of lysozyme contains
a swollen -domain, a likely result of the surfactant tails interacting with protein -
helices, which are generally hydrophobic in nature. Furthermore, by noting the
progressive swelling of the lower-left side of the protein in Figure 2.2 with surfactant
addition under visible light, it appears that unfolding of helix A in the -domain may
be primarily responsible for formation of the I
S
intermediate (with some contribution
62
from helix C also possible). However, it should be cautioned that SANS is an
ensemble technique, meaning that if multiple states exist in solution (i.e., a
fluctuating intermediate, a mixture of unfolded states, etc.) the structures obtained at
7.9 and 12.2 mM azoTAB under visible light would be considered “average”
structures. However, as will be shown below, these structures appear to be true
intermediates; thus, when tertiary structures obtained from SANS are combined with
the information on the secondary structure from FT-IR, the general mechanism of
lysozyme unfolding with the surfactant occurring through denaturing of the -
domain can be elucidated.
As the surfactant binds to the protein, hydrophobic interactions between
surfactant tails and the non-polar amino acids occur, allowing these hydrophobic
amino acids, which are normally folded within the protein interior, to be exposed to
the solvent as the protein swells. For azoTAB, these surfactant-protein interactions
can be manipulated with light because the visible-light (trans) form of the surfactant
is more hydrophobic than the UV-light (cis) form. This allows for reversible control
of lysozyme folding/swelling. Evidence of increased surfactant hydrophobicity
influencing protein folding has also been seen in traditional (i.e., non-
photoresponsive) surfactants, where increasing the hydrocarbon tail length for a
given headgroup has been shown to result in a greater degree of protein
unfolding(30).
To study the intermediately folded states of lysozyme that occur with
azoTAB addition and light illumination, fluorescence measurements with Nile red as
63
a probe were performed. Nile red is a nonionic and hydrophobic molecule that
exhibits low fluorescence in polar media such as water, but when Nile red is
preferentially partitioned into non-polar environments, the molecule will exhibit a
large increase in fluorescence. Thus, Nile red fluorescence would be expected to be
low in a protein solution with the protein in the native state, because the tightly
packed core would prevent Nile red from partitioning into this non-polar
environment. However, if the protein were to swell and the interior of the protein
were to become more loosely packed, Nile red could partition into the core,
evidenced by an increase in fluorescence. Thus, Nile red fluorescence measurements
provide a sensitive technique for differentiating between native and non-native
states(41). This use of Nile red to probe the formation of hydrophobic domains is
conceptually similar to experiments involving ANS, a probe molecule that binds to
hydrophobic sites on the protein and has been used to test for the formation of
molten globules(57). However, the use of ANS ( excite
~ 400 nm) is precluded in the
present study because of strong absorption of azoTAB at wavelengths less than 500
nm(34); thus, Nile red ( excite
~ 560 nm) was used.
As seen in Figure 2.5, for the lysozyme-azoTAB system, the fluorescence of
Nile red begins to increase at surfactant concentrations as low as 1 mM and
eventually reaches a plateau at around 5 mM under visible light. This indicates that
hydrophobic domains become increasingly available for Nile red to partition into the
bulk water phase as the surfactant concentration is increased from 1 to 5 mM, despite
the fact that the tertiary structure of the protein exhibits only modest differences from
64
the native state over a similar region, as seen in Figure 2.2. Thus, both FT-IR and
fluorescence techniques suggest that a loosely packed protein core develops before
the dramatic changes in the protein tertiary structure, observed with SANS. Under
UV light, however, the fluorescence of Nile red remains low for all surfactant
concentrations, indicating that photoisomerization of the surfactant to the cis form
causes lysozyme to refold to a native-like conformation, leading to a decrease in
accessible hydrophobic domains.
Figure 2.5: Fluorescence of Nile red as a function of the azoTAB surfactant concentration as
measured under both visible and UV light illumination. [lysozyme] = 10.0 mg/mL.
65
DLS was also used to investigate the effect of the surfactant and light
illumination on the conformation and swelling of lysozyme. As shown in Figure
2.6a, the measured diffusion coefficients decreased with an increasing surfactant
concentration under visible or UV light illumination, indicating an increase in the
overall size (hydrodynamic radius) of the protein-surfactant complex. The value
obtained for the hydrodynamic radius of pure lysozyme with no surfactant (R
H
=
17.5 Å) agrees well with literature values (R
H
= 17.4 ± 0.1 Å)(17, 58) under similar
conditions (i.e., low ionic strength, pH ~7).
To compare the light-scattering results in Figure 2.6a to the SANS fits in
Figure 2.2, estimated hydrodynamic radii (diffusion coefficients) were calculated
directly from each shape-reconstruction fit using Kirkwood’s theory(59) for a
collection of spherical subunits. The radius of each subunit was taken to be that of a
sphere with a volume of 1/1000
th
of the molecular volume of lysozyme (1000
scattering centers where used in the shape-reconstruction fits). Water molecules that
are tightly bound to the protein-surfactant complex were also accounted for by
adding a 3 Å thick water hydration shell to the overall radii(19). As seen in Figure
2.6a, the hydrodynamic properties estimated from the SANS fits agree nicely with
the experimental values determined from DLS. Specifically, both SANS and DLS
appear to reveal three different protein forms under visible light: a folded (native-
like) form at low azoTAB concentrations, a partially folded form at intermediate
concentrations, and a swollen/unfolded form at higher azoTAB concentrations.
66
Figure 2.6: (a) Lysozyme swelling as measured by DLS and (b) max
of crystal violet
measured by absorption spectroscopy under both visible and UV light illumination.
[lysozyme] = 1.00mg/mL
67
In the case of neutron scattering, the UV-light form of the surfactant had little
to no effect on the conformation of lysozyme, in contrast to Figure 6a. However, the
surfactant/protein molar ratio was by necessity much lower for SANS (6.2-14.9) than
for DLS (10.5-126), because of the fact that a lower protein concentration was used
in the DLS experiments to approximate the diffusion coefficient at infinite dilution.
Thus, only one protein conformation is observed with UV-light illumination from
SANS data; however, there appear to be at least two forms under UV light from the
DLS data, because more surfactant is available to bind to the protein in the case of
light scattering. A direct comparison of the SANS and DLS results is therefore not
possible, but in general, it is observed that the protein swells with the addition of the
visible-light form of the surfactant, while the addition of the UV-light form of the
surfactant has a lesser effect. The more hydrophobic, visible-light form of the
surfactant has a greater affinity to bind to the protein than the relatively hydrophilic,
UV-light form(19). Because hydrophobic interactions between surfactant tails and
non-polar amino acids, which are normally folded within the protein interior, allow
these amino acids to be exposed to the solvent, protein folding and swelling can be
reversibly manipulated with light. This is evident from Figure 2.6a, because a greater
increase in the size of the protein is observed under visible light as opposed to UV
light with an increasing surfactant concentration. As previously stated, similar results
with conventional (i.e., not photoresponsive) surfactants have shown that a greater
degree of protein unfolding occurs in the presence of a more hydrophobic
surfactant(30). DLS experiments have also been used to examine mixtures of
68
lysozyme and SDS(16-18). The hydrodynamic radius of the protein-surfactant
complex was observed to increase from ~15 to 25 Å as the SDS/lysozyme molar
ratio increased from 1.6 to 16.7(16), similar to the results in Figure 2.6a.
The photoinduced unfolding and refolding of lysozyme observed in this study
are consistent with results of the effect of azoTAB and light illumination on BSA
conformation(19). In the case of BSA, however, a more dramatic unfolding was
observed with azoTAB addition, as opposed to swelling in the case of lysozyme. In
general, small proteins such as lysozyme are harder to unfold than larger proteins
such as BSA, and lysozyme is known to be a tightly folded protein especially in the
-domain(12), which consists of a hydrophobic pocket enclosed by four helices
(although surfactants swell the -domain, as discussed above). Furthermore, the
difference in isoelectric points (BSA, pI = 4.7; lysozyme, pI = 11) and net charges
[-18 for BSA(19) and +8 for lysozyme at neutral pH] may cause binding of the
cationic azoTAB to be more difficult in the case of lysozyme and, hence, result in
less unfolding. The fact that in this study swelling of lysozyme is observed primarily
in the -domain, which contains the aforementioned hydrophobic pocket,
demonstrates the importance of hydrophobic surfactant-protein interactions in the
unfolding process. Similarly, in solutions of lysozyme mixed with CTAB,
hydrophobic interactions were found to dominate over electrostatic repulsion
between lysozyme and the positively charged surfactant(60).
UV-visible spectroscopy was used to analyze the binding of crystal violet (a
nearly planar, cationic probe molecule) to lysozyme. The absorption maximum of
69
crystal violet depends upon the microenvironment in which the probe is located. In a
polar solvent such as water, max
= 590 nm, while in a non-polar solvent such as
benzene, max
= 605 nm(61). Thus, investigating the absorbance of crystal violet
upon binding to lysozyme and with the addition of azoTAB allows for the
determination of the micropolarity of the local environment surrounding the probe,
which can in turn be used to infer about the binding of azoTAB to lysozyme.
As seen in Figure 2.6b, under visible light, a slight increase in max starts at
concentrations as low as 0.75 mM azoTAB; however, on average, the max
values
indicate that crystal violet is experiencing a polar environment similar to that of
water for surfactant concentrations <2 mM. This demonstrates that there is not yet
enough surfactant to dramatically affect the conformation of lysozyme, and crystal
violet remains either unbound to lysozyme (likely because both crystal violet and the
lysozyme have a positive charge and there is not as strong of a hydrophobic binding
force as in azoTAB) or perhaps bound to the surface of the protein (and therefore
still exposed to water). Between 2 and 4.5 mM azoTAB under visible light, max
increases to a nearly constant value of ~598 nm, demonstrating a significant decrease
in the local polarity of crystal violet. This is likely a result of protein
unfolding/swelling with surfactant addition, which would make normally interior
hydrophobic domains of the protein available to crystal violet, combined with the
ability of planar crystal violet to “stack” with planar, trans azoTAB molecules bound
to the protein. Note that the increase in max
occurs over similar azoTAB
concentrations that an increase in the hydrodynamic radius was observed from DLS,
70
and, furthermore, both sets of data demonstrate a possible plateau in the range of 2-4
mM azoTAB, implying that the lysozyme has adopted a new “partially swollen”
form. Increasing the surfactant concentration beyond ~4 mM under visible light
causes max
to again increase, approaching ca. 602 nm, similar in polarity to a
benzene-like environment, and indicating that lysozyme has adopted a second
“swollen” form. Again, note the similarity in this increase in crystal violet max
to the
increase in the hydrodynamic radius observed over the same surfactant
concentrations in Figure 2.6a. Furthermore, the three protein forms observed in parts
a and b of Figure 2.6 (native-like, partially swollen, and swollen) are consistent with
the three forms determined from neutron scattering under visible light (native-like at
5.1 mM azoTAB, partially swollen at 7.9 mM, and swollen at 12.2 mM).
Admittedly, the SANS data (11.8 mg/mL) and the DLS and UV-vis data (both at 1.0
mg/ mL) are at different protein concentrations; however, the consistency of the
results does indicate that the shape reconstruction fits in Figure 2.2 demonstrate
swelling of the protein.
This decrease in micropolarity observed with crystal violet is not seen with
increasing surfactant concentrations under UV light; indeed, max
stays relatively
constant at around 590 nm for concentrations ranging from 0 to 5 mM. At a
surfactant concentration of 5 mM under UV light, max
slowly starts to increase with
an increasing surfactant concentration, and above ~6 mM, there appears to be a rapid
increase in max
until a value of ~600 nm is approached. This increase is again at
about the same point that the light scattering data demonstrate a plateau, indicating a
71
partially swollen form of lysozyme. Furthermore, beyond this plateau, once the
hydrodynamic radius begins to increase again with the surfactant concentration (~8
mM), a steep increase in crystal violet max
is seen, again similar to the behavior
under visible light. The fact that a dampened increase in max
is observed under UV
light compared to visible light is likely a result of the trans (planar) surfactant being
more effective at “stacking” with crystal violet, compared to the cis (bent) surfactant
form.
2.5 Conclusion
This study demonstrates that intermediately folded states of lysozyme can be
stabilized through the addition of photoresponsive surfactants, thereby allowing
lysozyme unfolding and refolding to be controlled with light illumination. The
visible-light (trans) form of the surfactant is more hydrophobic than the UV-light
(cis) form; hence, surfactant binding to the protein and the protein structure can be
tuned with light. Three conformations of the protein were observed under visible
light, including a native-like structure at low surfactant concentrations (R
g
= 13.5 Å),
a partially swollen form at intermediate surfactant concentrations (14.3 Å), and a
swollen/unfolded conformation at higher surfactant concentrations (14.7 Å), while
UV-light exposure caused the protein to refold. Furthermore, by applying shape-
reconstruction methods to SANS data, the precise nature of surfactant-induced
protein unfolding was visualized. Combined SANS and FT-IR measurements
revealed that lysozyme unfolding occurred primarily in the -domain, potentially
72
initiated by a surfactant-induced denaturing of helix A (residues 4-15). Thus, the
surfactant-unfolded conformation appears to represent a folding intermediate unique
from the well-known -domain intermediate of lysozyme that, in contrast, contains a
folded -domain and unfolded -domain. The ability to directly track lysozyme
unfolding in solution, particularly when combined with photoresponsive surfactants
as a means of probing protein structure, illustrates the potential impact of the SANS
technique to study protein folding.
2.6 Acknowledgements
We thank W. T. Heller for graciously supplying the GA_STRUCT program.
We acknowledge the support of the National Institute of Standards and Technology,
U.S. Department of Commerce, in providing the neutron research facilities used in
this work. This work utilized facilities supported in part by the National Science
Foundation under Agreement number DMR-9986442. We also acknowledge the
Charles Lee Powell Foundation for support of this research.
2.7 References
(1) Horwich, A. (2002) Protein aggregation in disease: A role for folding intermediates
forming specific multimeric interactions, J. Clin. Invest. 110, 1221-1232.
(2) Dobson, C. M. (1999) Protein misfolding, evolution, and disease, Trends Biochem.
Sci. 24, 329-332.
(3) Morozova-Roche, L. A., Zurdo, J., Spencer, A., Noppe, W., Receveur, V., Archer,
D. B., Joniau, M., and Dobson, C. M. (2000) Amyloid fibril formation and seeding
by wild-type human lysozyme and its disease-related mutational variants, J. Struct.
Biol. 130, 339-351.
73
(4) Pepys, M. B., Hawkins, P. N., Booth, D. R., Vigushin, D. M., Tennent, G. A.,
Soutar, A. K., Totty, N., Nguyen, O., Blake, C. C. F., et al. (1993) Human lysozyme
gene mutations cause hereditary systemic amyloidosis, Nature 362, 553-557.
(5) Valleix, S., Drunat, S., Philit, J.-B., Adoue, D., Piette, J.-C., Droz, D., MacGregor,
B., Canet, D., Delpech, M., and Grateau, G. (2002) Hereditary renal amyloidosis
caused by a new variant lysozyme W64R in a French family, Kidney Int. 61, 907-
912.
(6) Bowie, J. U. (2001) Stabilizing membrane proteins, Curr. Opin. Struct. Biol. 11,
397-402.
(7) Brooks, C. L., III (1998) Simulations of protein folding and unfolding, Curr. Opin.
Struct. Biol. 8, 222-226.
(8) Duan, Y., and Kollman, P. A. (1998) Pathways to a protein folding intermediate
observed in a 1-ís simulation in aqueous solution, Science 282, 740-744.
(9) Jones, J. A., Wilkins, D. K., Smith, L. J., and Dobson, C. M. (1997) Characterization
of protein unfolding by NMR diffusion measurements, J. Biomol. NMR 10, 199-203.
(10) Uversky, V. N., Karnoup, A. S., Segel, D. J., Seshadri, S., Doniach, S., and Fink, A.
L. (1998) Anion-induced folding of Staphylococcal nuclease: Characterization of
multiple equilibrium partially folded intermediates, J. Mol. Biol. 278, 879-894.
(11) Bachmann, A., Segel, D., and Kiefhaber, T. (2002) Test for cooperativity in the
early kinetic intermediate in lysozyme folding, Biophys. Chem. 96, 141-151.
(12) Yokota, A., Hirai, K., Miyauchi, H., Iimura, S., Noda, Y., Inoue, K., Akasaka, K.,
Tachibana, H., and Segawa, S. (2004) NMR characterization of three-disulfide
variants of lysozyme, C64A/C80A, C76A/C94A, and C30A/C115AsA marginally
stable state in folded proteins, Biochemistry 43, 6663-6669.
(13) Dobson, C. M. (1994) Solid evidence for molten globules, Curr.Biol. 4, 636-640.
(14) Chen, L., Hodgson, K. O., and Doniach, S. (1996) A lysozyme folding intermediate
revealed by solution X-ray scattering, J. Mol. Biol. 261, 658-672.
(15) Matthews, C. R. (1993) Pathways of protein folding, Ann. Rev. Biochem. 62, 653-
683.
(16) Gimel, J. C., and Brown, W. (1996) A light scattering investigation of the sodium
dodecyl sulfate-lysozyme system, J. Chem. Phys. 104, 8112-8117.
(17) Valstar, A., Brown, W., and Almgren, M. (1999) The lysozyme sodium dodecyl
sulfate system studied by dynamic and static light scattering, Langmuir 15, 2366-
2374.
74
(18) Stenstam, A., Montalvo, G., Grillo, I., and Gradzielski, M. (2003) Small angle
neutron scattering study of lysozyme-sodium dodecyl sulfate aggregates, J. Phys.
Chem. B 107, 12331-12338.
(19) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2005) Photocontrol of protein
folding: The interaction of photosensitive surfactants with bovine serum albumin,
Biochemistry 44, 524-536.
(20) Ibel, K., May, R. P., Kirschner, K., Szadkowski, H., Mascher, E., and Lundahl, P.
(1990) Protein-decorated micelle structure of sodium-dodecyl-sulfate-protein
complexes as determined by neutron scattering, Eur. J. Biochem. 190, 311-318.
(21) Griko, Y. V., Freire, E., Privalov, G., van Dael, H., and Privalov, P. L. (1995) The
unfolding thermodynamics of c-type lysozymes: A calorimetric study of the heat
denaturation of equine lysozyme, J. Mol. Biol. 252, 447-459.
(22) Kiefhaber, T. (1995) Kinetic traps in lysozyme folding, Proc. Natl. Acad. Sci. U.S.A.
92, 9029-9033.
(23) Matagne, A., Radford, S. E., and Dobson, C. M. (1997) Fast and slow tracks in
lysozyme folding: Insight into the role of domains in the folding process, J. Mol.
Biol. 267, 1068-1074.
(24) Radford, S. E., Dobson, C. M., and Evans, P. A. (1992) The folding of hen lysozyme
involves partially structured intermediates and multiple pathways, Nature 358, 302-
307.
(25) Dinner, A. R., Sali, A., Smith, L. J., Dobson, C. M., and Karplus, M. (2000)
Understanding protein folding via free-energy surfaces from theory and experiment,
Trends Biochem. Sci. 25, 331-339.
(26) Gilquin, B., Guilbert, C., and Perahia, D. (2000) Unfolding of hen egg lysozyme by
molecular dynamics simulations at 300 K: Insight into the role of the interdomain
interface, Proteins 41, 58-74.
(27) Booth, D. R., Sunde, M., Bellotti, V., Robinson, C. V., Hutchinson, W. L., Fraser, P.
E., Hawkins, P. N., Dobson, C. M., Radford, S. E., Blake, C. C. F., and Pepys, M. B.
(1997) Instability, unfolding, and aggregation of human lysozyme variants
underlying amyloid fibrillogenesis, Nature 385, 787-793.
(28) Canet, D., Sunde, M., Last, A. M., Miranker, A., Spencer, A., Robinson, C. V., and
Dobson, C. M. (1999) Mechanistic studies of the folding of human lysozyme and the
origin of amyloidogenic behavior in its disease-related variants, Biochemistry 38,
6419-6427.
(29) Shih, P., Holland, D. R., and Kirsch, J. F. (1995) Thermal stability determinants of
chicken egg-white lysozyme core mutants: Hydrophobicity, packing volume, and
conserved buried water molecules, Protein Sci. 4, 2050-2062.
75
(30) Lee, P. K. J., and Jirgensons, B. (1971) Conformational transitions of
immunoglobulin fragment Fe(t) incited by alkyl sulfates of various hydrophobic
chain lengths, Biochim. Biophys. Acta 229, 631-641.
(31) Carter, D. C., and Ho, J. X. (1994) Structure of serum albumin, Adv. Protein Chem.
45, 153-203.
(32) Hayashita, T., Kurosawas, T., Miyata, T., Tanaka, K., and Igawa, M. (1994) Effect
of structural variation within cationic azosurfactant upon photoresponsive function
in aqueous solution, Colloid Polym. Sci. 272, 1611-1619.
(33) Shang, T., Smith, K. A., and Hatton, T. A. (2003) Photoresponsive surfactants
exhibiting unusually large, reversible surface tension changes under varying
illumination conditions, Langmuir 19, 10764-10773.
(34) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2004) Photoreversible viscosity
changes and gelation in mixtures of hydrophobically modified polyelectrolytes and
photosensitive surfactants, Macromol. 37, 5397-5405.
(35) Durchschlag, H., Hefferle, T., and Zipper, P. (2003) Comparative investigations of
the effects of X- and UV-irradiation on lysozyme in the absence or presence of
additives, Radiat. Phys. Chem. 67, 479-486.
(36) Glinka, C. J., Barker, J. G., Hammouda, B., Krueger, S., Moyer, J. J., and Orts, W. J.
(1998) The 30 m small-angle neutron scattering instruments at the National Institute
of Standards and Technology, J. Appl. Crystallogr. 31, 430-445.
(37) Heller, W. T., Krueger, J. K., and Trewhella, J. (2003) Further insights into
calmodulin-myosin light chain kinase interaction from solution scattering and shape
restoration, Biochemistry 42, 10579-10588.
(38) Glatter, O. (1977) A new method for the evaluation of smallangle scattering data, J.
Appl. Crystallogr. 10, 415-421.
(39) Feigin, L. A., and Svergun, D. I. (1987) Structure Analysis by Small-Angle X-ray
and Neutron Scattering, Plenum Press, New York.
(40) Svergun, D. I. (1992) Determination of the regularization parameter in indirect-
transform methods using perceptual criteria, J. Appl. Crystallogr. 25, 495-503.
(41) Daban, J. R., Samso, M., and Bartolome, S. (1991) Use of Nile red as a fluorescent
probe for the study of the hydrophobic properties of protein-sodium dodecyl sulfate
complexes in solution, Anal. Biochem. 199, 162-168.
(42) Guinier, A., and Fournet, G. (1955) Small Angle Scattering of X-ray, Wiley, New
York.
76
(43) Gallas, J. M., Littrell, K. C., Seifert, S., Zajac, G. W., and Thiyagarajan, P. (1999)
Solution structure of copper ion-induced molecular aggregates of tyrosine melanin,
Biophys. J. 77, 1135- 1142.
(44) Stuhrmann, H. B., and Fuess, H. (1976) Neutron small-angle scattering study of hen
egg-white lysozyme, Acta Crystallogr., Sect. A: Found. Crystallogr. 32 (part 1), 67-
74.
(45) Nicoli, D. F., and Benedek, G. B. (1976) Study of thermal denaturation of lysozyme
and other globular proteins by light scattering spectroscopy, Biopolymers 15, 2421-
2437.
(46) Kamatari, Y. O., Konno, T., Kataoka, M., and Akasaka, K. (1998) The methanol-
induced transition and the expanded helical conformation in hen lysozyme, Protein
Sci. 7, 681-688.
(47) Segel, D. J., Bachmann, A., Hofrichter, J., Hodgson, K. O., Doniach, S., and
Kiefhaber, T. (1999) Characterization of transient intermediates in lysozyme folding
with time-resolved small-angle X-ray scattering, J. Mol. Biol. 288, 489-499.
(48) Bu, Z., and Engelman, D. M. (1999) A method for determining transmembrane helix
association and orientation in detergent micelles using small-angle X-ray scattering,
Biophys. J. 77, 1064-1073.
(49) Velev, O. D., Kaler, E. W., and Lenhoff, A. M. (1998) Protein interactions in
solution characterized by light and neutron scattering: Comparison of lysozyme and
chymotrypsinogen, Biophys. J. 75, 2682-2697.
(50) Uversky, V. N. (1993) Use of fast protein size-exclusion liquid chromatography to
study the unfolding of proteins which denature through the molten globule,
Biochemistry 32, 13288-13298.
(51) Diamond, R. (1974) Real-space refinement of the structure of hen-egg-white
lysozyme, J. Mol. Biol. 82, 371-391.
(52) Svergun, D. I., and Koch, M. H. J. (2003) Small-angle scattering studies of
biological macromolecules in solution, Rep. Prog. Phys. 66, 1735-1782.
(53) Fabian, H., and Mantele, W. (2002) in Handbook of Vibrational Spectroscopy
(Chalmers, J. M., and Griffiths, P. R., Eds.) pp 3399-3424, Wiley, New York.
(54) Lad, M. D., Ledger, V. M., Briggs, B., Green, R. J., and Frazier, R. A. (2003)
Analysis of the SDS-lysozyme binding isotherm, Langmuir 19, 5098-5103.
(55) Liu, H., Yang, W., and Chen, J. (1998) Effects of surfactants on emulsification and
secondary structure of lysozyme in aqueous solutions, Biochem. Eng. J. 2, 187-196.
(56) Wang, S.-C., and Lee, C. T., Jr. (2006) Protein Secondary Structure Controlled with
Light and Photoresponsive Surfactants, J. Phys. Chem. B. 110, 16117-16123.
77
(57) Semisotnov, G. V., Rodionova, N. A., Razgulyaev, O. I., Uversky, V. N., Gripas, A.
F., and Gil’manshin, R. I. (1991) Study of the “molten globule” intermediate state in
protein folding by a hydrophobic fluorescent probe, Biopolymers 31, 119-128.
(58) Colvin, J. R. (1952) Size and shape of lysozyme, Can. J. Chem. 30, 831-834.
(59) Kirkwood, J. G. (1996) The general theory of irreversible processes in solutions of
macromolecules, J. Polym. Sci. Pol. Phys. 34, 1-14.
(60) Chatterjee, A., Moulik, S. P., Majhi, P. R., and Sanyal, S. K. (2002) Studies on
surfactant-biopolymer interaction. I. Microcalorimetric investigation on the
interaction of cetyltrimethylammonium bromide (CTAB) and sodium dodecyl
sulfate (SDS) with gelatin (Gn), lysozyme (Lz), and deoxyribonucleic acid (DNA),
Biophys. Chem. 98, 313-327.
(61) Mackay, R. A., Letts, K., and Jones, C. (1977) in Micellization, Solubilization, and
Microemulsions (Mittal, K. L., Ed.) pp 801, Plenum, New York.
78
CHAPTER III: Control of Ribonuclease A Structure through the
use of Photoresponsive Surfactants
3.1 Abstract
A photosensitive surfactant is used to stabilize intermediately folded sates of
ribonuclease A (RNase A), making it possible to shape reconstruct such intermediate
states. The surfactant has two different forms depending on light illumination. A
trans planar form under visible light and a cis bent form under UV light. The trans
form of the surfactant is more hydrophobic than the cis form, thus it more readily
binds to the protein. This photosensitivity of the surfactant allows for tunable and
reversible surfactant to protein binding. As a result, photoreversible control of
RNase A from a swollen state under visible light to a compact state, consistent with
the native state, under UV light is achieved. This study proves that surfactant
concentration and not protein concentration is key to conformational changes of the
protein. Shape reconstruction of small angle neutron scattering data show the effects
of surfactant concentration on the conformation of the protein to be independent of
protein concentration. This lends to the idea of a surfactant saturation limit, where
each protein molecule only has a certain number of binding sites available to the
surfactant. Once these sites are filled binding and hence conformational changes no
longer take place. Therefore, independent of protein concentration, the fully
unfolded form of RNase A remains the same.
79
3.2 Introduction
Proteins carry out essentially all biological functions. The structure of a
protein determines the function of that protein. A change in protein environment
(e.g., temperature, pressure, hydrophobicity, or pH) can cause a change in the
protein structure, which can result in a change in protein function. Non-native
protein structures are very difficult to characterize. The two techniques used to
determine protein structure, X-ray crystallography and NMR, are typically limited to
native conformations. To investigate intermediate protein structures it is necessary
to have a means to induce changes in the protein structure and a method to determine
in vitro conformations. The present study will demonstrate both a means and
method for the structural determination of intermediately folded RNase A.
RNA depolymerases (i.e., “ribonucleases”) catalyze the degradation of RNA.
Specifically, ribonuclease A (RNase A) is a 124 residue monomeric enzyme that
catalyses the cleavage of the P-O
5
bond in single-stranded RNA. RNase A from
bovine pancreas is one of the model systems of protein science; it was the most
studied enzyme of the 20
th
century.(1) The “A” refers to the predominant form of the
enzyme in the pancreas of Bos Taurus (domestic cow).(2) The active site of RNase A
is located in its cleft. It is composed of three -helices ( 1, the N-terminal helix,
residues 3-13; 2, residues 24-34; and 3, residues 50-60) and three -hairpins
(residues 61-74, 79-104 and 105-124, the C terminal hairpin) arranged in two -
sheets. Hairpins 61-74 and 105-124 form a four-stranded antiparallel -sheet ( 1).
A three-stranded antiparallel -sheet ( 2) is formed from the major -hairpin
80
residues (79-104) and a short -strand (residues 42-45).(3) The enzyme is tightly
linked by four disulphide bonds (Cys26-Cys84, Cys58-110, Cys40-95 and Cys65-
72), which are critical to the proteins stability and significantly limit conformational
changes. The Cys26-Cys84 and Cys58-110 disulfide bonds are between an -helix
and a -sheet, 2- 2 and 3- 1, respectively.(3) These bonds contribute more to the
stability than the two other disulfide bonds which simply connect loop segments.
RNase A (pI = 9.3) is positively charged at neutral pH and contains all but
one of the 20 naturally occurring amino acids (tryptophan), the three residues
believed to be most important for catalysis are His12, His 119 and Lys 41. A
grouping of hydrophobic residues makes up the major hydrophobic core (residues
58-110) of RNase A, composed primarily of the residues form the C-terminal -
hairpin as well as the first and third -helix. A minor hydrophobic core also exists
that contains residues from the second -helix and the major -hairpin. RNase A is
shaped like a kidney bean and divided into two halves a front and a back half (as if
the kidney bean where sliced lengthwise in two); an N-terminal, predominantly -
helical half (residues 1-60) and a C-terminal, predominantly -sheet half (residues
65-124).
RNase A also contains four X-Pro peptide bonds with Tyr92-Pro93 and
Asn113-Pro114 are in the cis conformation while Lys41-Pro42 and Val116-Pro117
are in the trans conformation.(4) Five intermediate species of unfolded RNase A
have been observed experimentally, including the very-fast, fast, medium, and two
slow folding phases.(5) Before all five of these form were even discovered Brandts
81
et al. (1975) suggested that the isomerization of X-Pro peptide bonds was responsible
for the variation in the unfolded states of RNase A.(6) Different refolding rates have
been observed for the different unfolded species. Specifically the conformations
about the X-Pro peptide bonds are believed to be responsible for these differences.
Upon thermal unfolding, 2 becomes destabilized at moderate temperatures
this loosens the packing of 1 in the major hydrophobic core.(3) The thermally
denatured state has a considerably smaller radius of gyration in comparison to that of
the chemically denatured state.(7) On the other hand the radius of gyration for the
“denaturant-induced unfolded state of RNase A” is significantly lower than that
estimated for a random coil, most likely because of the disulfide bonds.(8) The
unfolding of RNase A by sodium dodecyl sulfate (SDS) is similar to the extent of
unfolding reported from thermal denaturation.(9)
Recently, azobenzene-based light-responsive surfactants (“azoTABs”) have
been used to reversibly photo-control protein structures(10-13) and even protein
activity(14). The azobenzene group in the photosurfactant undergoes a
photoisomerization from the trans form under visible exposure (436 nm) to the cis
form under UV exposure (365 nm). The trans form is more hydrophobic than the
cis form and thus more readily binds to the protein. Essentially surfactant to protein
binding can be turned “on” under visible light and “off” under UV light. It has been
shown that azoTAB prefers to bind to the helical segments of several proteins
(lysozyme, BSA, and -chymotrypsin)(10, 11, 13). In the case of lysozyme,
azoTAB was used to swell the protein under visible light and then refold the protein
82
under UV light. This control of protein structure allowed for the control of enzyme
activity. In the swollen form under visible light the enzyme exhibited an activity
eight fold higher than the native state, while in the folded/native form under UV light
the activity remained the same as the native protein.(14) Lysozyme superactivity was
attributed to the fact that the protein swelling occurred far from the active site near
the hinge region (where the domain attaches to the domain). This allows for
enhanced domain motions and thus increased enzyme flexibility.
In the present study, small-angle neutron scattering (SANS), dynamic light
scattering (DLS), fluorescence and UV-vis spectroscopy will be used to investigate
the effect of azoTAB on the structure of RNase A. A swollen/slightly-unfold form is
induced by the photoresponsive surfactant under visible light. The photoresponsive
property of this surfactant then allows for reversible control of this structure change,
from a swollen form to a compact/native form. Both 10 mg/mL and 1 mg/mL RNase
A concentrations are employed in the SANS experiments to investigate the effect of
protein concentration on surfactant induced structure changes.
3.3 Experimental Methods
3.3.1 Materials
An azobenze-trimethyammonium bromide surfactant (azoTAB) of the form
N
N
CH
3
CH
2
O(CH
2
)
4
N+(CH
3
)
3
Br-
,
83
similar to surfactants used in previous studies (10-16) was synthesized according to
published procedures(17, 18). This surfactant undergoes a photoisomerization from
a trans to a cis form when incident light is changed from visible to UV(10-16). For
SANS, DLS, fluorescence, and UV-vis spectroscopy experiments, conversion to the
cis form was achieved with an 84 W long-wave UV lamp (365 nm; Spectroline,
model XX-15A) with solutions exposed to UV light for at least 1hr prior to
measurements. In the cases where measurements took more than a couple of
minutes (SANS and DLS), continuous UV exposure was applied throughout data
collection. For FT-IR measurements, a 200 W mercury arc lamp was used for
conversion, as described previously (10, 11).
RNase A from bovine pancreas was purchased from Sigma (cat. No. R5500)
along with a low ionic strength phosphate buffer (pH 7.2, I = 8.3mM). All other
chemicals were purchased form Aldrich in their highest purity. An RNase A
concentration of ~1 mg/mL was used for all experiments except for in SANS were
experiments were performed at both 0.9 and 10.0 mg/mL.
3.3.2 Small Angle Neutron Scattering
Small Angle Neutron Scattering experiments were performed on the 30m
NG3 SANS instrument at NIST(19) as previously described(10, 11). GA_STRUCT,
a shape reconstruction algorithm was used to generate protein conformations from
the experimental scattering data using 2000 spherical scattering centers, similar to
84
previous studies(10-12). GNOM(20) was also used to calculated pair distance
distribution functions as described previously(10, 11).
3.3.3 Dynamic Light Scattering
Dynamic Light scattering measurements were performed at 25 °C on a
Brookhaven model BI-200SM instrument (Brookhaven Instrument Corp.) with a
35mW helium neon laser ( = 632.8 nm) (Melles Griot, model number 05-LHP-
928). The samples were filtered through a 0.02 μm Anatop filter to minimize the
presence of dust and its effect on scattering. Measurements were taken at three
different angles (90°, 60°, 30°) to obtain an accurate measure of the diffusion
coefficient and, hence, the hydrodynamic radius, R
H
, of the protein. Three different
angles were used to include more of the correlation function and to insure
reproducibility independent of angle. Experimental results were obtained from
averaging six different measurements on the same sample (two measurements at
each angle). The protein concentration employed was 0.93 mg/mL.
3.3.4 Fluorescence Spectroscopy
Fluorescence measurements were performed at 25° C on a Quanta Master
spectrofluorometer model QM-4 (Photon Technology International) using Nile red as
the fluorescence probe. An emission range of 600-725 nm was studied with an
excitation wavelength of 560 nm, while both the excitation and emission slit widths
were 4 nm. The surfactant was preconverted to the cis form prior to both protein and
85
Nile red injection in order to avoid any photodegradation of Nile red and any
prolonged exposure of the protein to UV light. Prolonged exposure of the protein
could produce free radicals that can detrimentally effect the proteins activity.
Concentrated solutions of protein and Nile red were injected to obtain final
concentrations of RNase A equal to that used in the DLS measurements (0.93
mg/mL) along with 10 μM Nile Red.
3.3.5 UV-visible Spectroscopy
Absorption measurements were preformed on an Agilent model 8453 UV-
visible spectrophotometer. Again to avoid photodegradation and prolonged exposure
of the protein to UV, the surfactant was preconverted and then a concentrated RNase
A-crystal violet solution was injected prior to measurements to achieve final
concentrations of 10 μM crystal violet and 0.98 mg/mL RNase A.
3.4 Results and Discussion
The effect of azoTAB on RNase A is shown by the SANS data in Figure 3.1,
at two different protein concentrations (0.90 and 10.0 mg/mL). SANS has a very
broad Q range and can, therefore, probe many length scales as Q L / 2 . In these
experiments, a Q range of 0.0044 to 0.49 Å
-1
was investigated and, therefore length
scales from 12.8 to 1428 Å were probed. The scattering curves for pure RNase A at
the two different protein concentrations differ by an order of magnitude at Q < 0.01
Å
-1
, as expected since scattering intensity is directly proportional to protein
86
concentration. At an RNase A concentration of 0.9 mg/mL (Figure 3.1a), the
presence of azoTAB under visible light causes an increase in scattering intensity at Q
< 0.16 Å
-1
(L > 39 Å, roughly the size of the native protein) this is indicative of an
increase in the size or “swelling” of RNase A. However at these same surfactant
concentrations under UV light the scattering curves resemble that of the pure protein
(with the exception of 8.33 mM), suggesting that the protein remains in the native
state at these conditions. The data presented in Figure 3.1a show that enhanced
protein occurs with increasing relatively hydrophobic trans surfactant concentration
under visible light, while converting the surfactant to the relatively hydrophilic cis
form with UV light causes the protein to refold back to a native-like form. In the
case of 8.33 mM azoTAB, a modest degree of protein unfolding is observed under
even UV light, suggesting that even the relatively hydrophilic cis form of the
surfactant begins to interact with the protein at this elevated concentration.
87
Figure 3.1: SANS data of RNase A-azoTAB solutions as a function of surfactant
concentration and light illumination. (a) [RNase A] = 0.90 mg/mL (b) [RNase A] = 10.0
mg/mL
88
Figure 3.1b shows deviations in the scattering intensity with surfactant
concentration, again at length scales larger than the size of native RNase A Q < 0.16
Å
-1
, (L > 39 Å). Although the scattering curves 10 mg/mL are initially and order of
magnitude higher and more closely grouped together than those at 0.90 mg/ mL the
results are actually quite similar, this will be discussed below. Both light
illumination and surfactant concentration again are seen to have effect on the protein
scattering. At the highest concentration under visible light (8.13 mM), a peak in the
scattering data appears at Q ~ 0.04 Å
-1
. This is an interaction peaks indicative of
enhanced electrostatic repulsion between (protein) particles. This would be expected
as increasingly more cationic surfactant binds to the already positively charged
protein. Also noteworthy is that based on contrast matching experiments in both the
RNase A system (discussed below) and previous studies,(10-12, 14) surfactant
micelle formation does not occur under these conditions.
To quantify the size changes of RNase A, Guinier analysis using the equation
) 3 / exp( ) 0 ( ) (
2 2
g
R Q I Q I =
was performed to determine the radius of gyration, R
g
, and scattering extrapolated to
zero angle, I(0), as reported previously(11). The results are shown in Table 1. The
I(0)/c, where c is the protein concentration, values calculated from the Guinier
analysis of the scattering data of the pure protein at 0.90 mg/ mL (0.0115 mg·cm
-
1
mL
-1
) and 10 mg/mL (0.0106 mg·cm
-1
mL
-1
) are in good agreement with the value
expected (0.0112 mg·cm
-1
mL
-1
) based on previous SANS measurements using a
range of native proteins, as shown in Figure 3.2, Furthermore, these values are in
89
agreement with the value calculated (0.0101 mg·cm
-1
mL
-1
) using the equation
I(0)/c = (M /N
A
)( / c)
2
(21, 22), where, c is the protein concentration, M is the
protein molecular weight, N
A
is Avogadro’s number, and / c is the scattering
length density increment of the solution (for RNase A / c = -22.2 x 10
9
(23).
As shown in Table 1, the radii of gyration and I(0) values increase with
increasing surfactant concentration from 13.3 to 20.2 Å and from 0.0115 to 0.4367
cm
-1
, respectively, at a protein concentration of 0.90 mg/mL. The 52% increase in
R
g
is similar to the 65% increase observed upon denaturing in guanidine
hydrochloride.(24) Since I(0) can be related to the molecular weight of the scatter as
seen in the equation above,, it is important to note that the measured increase in I(0)
cannot be simply the result of surfactant binding leading to an increase in the
effective molecular weight of the protein-surfactant complex. Over the range of
surfactant concentrations studied, between 33 and 126 surfactant molecules are
available for binding to each protein molecule. Using the highest concentration (126
surfactants per protein) and assuming all of the surfactant molecules are bound (a
conservative estimate), an I(0) value of at most 0.046 cm
-1
could be achieved. Thus,
when compared to the measured I(0) value (0.437 cm
-1
), it is clear that surfactant
alone does not account for the dramatic increase in I(0) and instead a protein
conformational change must be occurring.
Nevertheless, contrast matching techniques were used to rule out any
scattering from the formation of surfactant micelles, either free in solution or bound
to the protein. The protein was made “invisible” to neutrons by increasing the
90
scattering of the solvent to match that of the protein (contrast matching point for a
protein is 60:40, H
2
O:D
2
O), so that only scattering from the surfactant would be
observed. No scattering (within experimental error) was detected from the surfactant
(results not shown). Thus, the increase in I(0) in Table 1 is a direct result of protein
swelling.
The radii of gyration and I(0) values determined for all of the samples under
UV light (with the exception of 8.33 mM) are in good agreement with those obtained
for pure RNase A (13.5 ± 0.3 Å and 0.0138 ± 0.0027 cm-
1
, respectively). This R
g
value would equate to a protein diameter of D = 2*sqrt(5/3)R
g
~ 35 Å . From the
literature using a variety of techniques (NMR, DLS, and dielectric spectroscopy) the
size of RNase A is 38.8 Å(24-26), from the DLS measurements in this study (to be
discussed in detail below) it is 38.7 Å, from shape reconstruction of the SANS (also
to be discussed in detail below) is 38.9 Å, and from a calculation of the
hydrodynamic radius of PDB 1RBX(27) is 38.3 Å (using Kirkwood’s theory(28)).
This supports the idea that the swollen protein (by azoTAB under visible light) can
be refolded to its native state with UV light. At 8.33 mM the protein can be refolded
significantly (from an R
g
of 20.23 to 15.04 Å) but not completely to native-like
values due to the fact that even under UV light at high enough concentrations some
surfactant may begin to bind to the protein.
91
Figure 3.2: SANS experimental data collected for proteins of varying molecular weight (10-
12)
At a protein concentration of 10.0 mg/mL, R
g
and I(0) values also increase with
increasing surfactant concentration. As seen in Table 1, the radii of gyration increase
from 14.8 to 18.3 Å, consistent with the results at 0.90 mg/mL. In terms of I(0), as
expected the value for the pure protein at 10.0 mg /mL is approximately an order of
magnitude higher than that obtained at 0.90 mg/mL. Interestingly, however, the I(0)
value at the highest surfactant concentration is approximately the same at both
protein concentrations, a clear indication that the protein structure is primarily
determined by the surfactant concentration. This result indicates that there is likely
some sort of surfactant saturation level of surfactant binding that has been reached.
Only so many surfactants can bind to one RNase A and once all of these spots are
92
taken no more surfactant can bind preventing RNase A from continuing to swell. A
similar effect was observed with lysozyme, once a concentration of 12 mM azoTAB
was introduced no further changes in the protein structure were observed.(11, 13)
Table 3.1: Values of radius of gyration (R
g
) and I(0) as determined from Guinier analysis
and pair distance distribution function calculations (shown in Figure 3.2) of the SANS data
shown in Figure 3.1. Also shown are the hydrodynamic radii (R
H
) values calculated form
the shape reconstructions shown in Figures 3.4 and 3.5 using Kirkwood’s theory
Although the radius of gyration gives a rough measure of the size of the
protein, Guinier analysis assumes that the protein is relatively globular. While true
for native RNase A, this assumption was tested through determination of pair
distance distribution functions, P(r), calculated using GNOM(20) as shown in Figure
3.3. P(r) is related to the probability of two scattering centers (nuclei for SANS)
being a distance r + dr apart. For globular proteins, P(r) should have a symmetric,
inverse parabolic shape, with the peak height given approximately by the position of
the maximum of the peak, namely the most common intermolecular dimension. As
can be seen from Figure 3.3a, the most common dimension within the protein, r
max
,
increases slightly with increasing azoTAB concentration, supporting the idea that the
93
protein is unfolding. However, the probability of this dimension (the height of the
maximum) and the maximum dimension within the protein, D
max
(where P(r) returns
to zero), do not change substantially, implying that the protein remains globular and
is, thus, swelling during the unfolding process (as opposed to elongation, for
example).
Figure 3.3: Pair Distance Distribution Function of RNase A-azoTAB solutions as a function
of azoTAB concentration and light illumination; (a) [RNase A] = 0.90 mg/mL (b) [RNase A]
= 10.0 mg/mL
(a)
(b)
94
Although Figure 3.3b shows very similar results to Figure 3.3a, indicating
very little difference between the results at different protein concentrations, there is a
slight difference in the increase of r. In Figure 3.1a there appears to be a discrete
step change when transitioning from 2.17 mM to 4.56 mM azoTAB under visible
light, followed by gradual smooth increases at higher surfactant concentrations. In
contrast, Figure 3.2b displays a smooth gradual increases in r
max
at all
concentrations. This is likely related to the surfactant saturation effect mentioned
above. Despite this small difference, the most common dimension, r
max
, is found to
increase with increasing azoTAB concentrations, and the dimensions obtained are
similar to those obtained at 0.90 mg/mL. Also under UV light the PDDFs match that
of the pure protein and again under all conditions D
max
is not changing.
Guinier analysis along with PDDF calculations give a good approximation of
the shape and size of the protein, however, they do not give precise detail about the
tertiary structure of the protein. Therefore, shape reconstruction (29) was applied to
the SANS data, as seen in Figures 3.4 (0.90 mg/mL) and 3.5 (10.0 mg/mL). The
shape reconstruction procedure begins with an initial guess of 2000 randomly
oriented scattering centers. The positions of the scattering centers are then
rearranged through a genetic algorithm to best fit the experimental scattering data. In
blue are the best fits for each condition with consensus envelopes (the average of 10
runs) are shown in red. The fact that each consensus envelope agrees nicely with its
respective best fit lends confidence to the shape reconstruction results.
95
As can be seen from Figure 3.4, the shape reconstruction results for the pure
protein are consistent with the X-ray crystallographic structure for pure RNase A
(PDB 1RBX) (27). As surfactant is added under visible light, the protein is seen to
swell and become increasingly globular. Upon exposure to UV light the protein
refolds into a state resembling that of the native state, with the exception of 8.33 mM
that refolds to a slightly swollen form resembling that of a visible concentration
somewhere between 2.17 and 4.56 mM. This is due to the fact that under visible light
the relatively hydrophobic trans form of the surfactant has a much higher affinity to
bind to the protein than the relatively hydrophilic cis form under UV light, however,
under UV light a small percentage of the surfactant still binds to the protein,
particularly at elevated surfactant concentrations.
Figure 3.4: Conformations of RNase A determined form shape reconstruction of the SANS
data (best fit in blue, consensus envelopes representing the average of 10 runs in red),
compared with the X-ray crystallographic structure (RDB code 1RBX)(27). [RNase A] =
0.90 mg/mL
96
Figure 3.5 shows the respective shape reconstruction fits for the scattering
collected at 10 mg/mL. As expected the shapes are almost identical to those
obtained at 0.90 mg/mL. This supports the conclusion that only surfactant
concentration and not protein concentration has a significant effect on the structural
changes observed in the RNase A/azoTAB system. An indication that hydrophobic
forces are responsible. Again implying that there is a surfactant saturations level
where protein-binding sites are no longer available to azoTAB. From the
GA_STRUCTS shown in Figure 3.4 and Figure 3.5 along with what has been shown
in previous studies with lysozyme and BSA, it appears that the hydrophobic regions
of the photo surfactant interact with the -helical regions of RNase A. This can be
seen by the swelling of the bottom region of the protein. Although the exact regions
of the protein cannot be determined from shape reconstruction the location of the
active site allows us to consistently orient the shape reconstructions. These
interactions with the helical regions are what cause swelling of RNase A.
97
Figure 3.5: Conformations of RNase A determined form shape reconstruction of the SANS
data (best fit in blue, consensus envelopes representing the average of 10 runs in red),
compared with the X-ray crystallographic structure (RDB code 1RBX)(27). [RNase A] =
10.0 mg/mL
Dynamic light scattering was preformed on the RNase A-azoTAB system, as
shown in Figure 3.6a, to measure the diffusion coefficient of the protein as a function
of surfactant concentration and light illumination. This was done as an additional
measure of the corresponding sizes (hydrodynamic radii, R
H
) at each condition.
From Figure 3.6a one can see that these values are very consistent with the SANS
data. Under visible light, increasing surfactant concentration causes the protein to
swell, while under UV light the protein remains with a size similar to the native state
independent of surfactant concentration. The hydrodynamic radius of RNase A
increases from 19.3 to 24.9 Å with increasing surfactant concentration under visible
light, while it remains relatively constant at 19.8 ± 0.7 Å under UV light. As
expected from the SANS data, a slight increase in the hydrodynamic radius is
98
observed under UV light as the concentration begins to approach 8 mM. DLS was
only preformed at 0.90 mg/mL to avoid any effects of protein-protein interactions.
Figure 3.6: RNase A swelling as a function of azoTAB concentration and light
illumination; as measured by (a) DLS [RNase A] = 0.93 mg/mL (b) Nile Red
Fluorescence; [RNase A] = 0.93 mg/mL and Crystal Violet max
[RNase A] = 0.984
mg/mL.
99
In order to compare the sizes measured with DLS to those measured with
SANS, Kirkwood’s theory(28) was used to calculated R
H
for each sample from the
PDB files created using GA_STRUCT(29) (blue structures shown in Figure 3.4 and
3.5). Once these values were determined, 5 Å was added to account for water
molecules tightly bound to the protein-surfactant complex(23). These values are
listed in Table 1. When compared to the data in Figure 3.6a, they clearly agree
within experimental error. (Table 1 also list the respective R
H
values calculated using
Kirkwood’s theory for the 10 mg/mL system).
To get a better idea of what is happening to the protein as it unfolds Nile red
fluorescence was used to probe the internal structure of the protein, as seen Figure
3.6b. With increasing surfactant concentration under visible light the fluorescence of
Nile Red increases. This increase is observed at an azoTAB concentration as low as
1 mM, indicating that Nile red is able to partition into the protein core as
hydrophobic domains become increasingly available even at low surfactant
concentrations. This suggests that a loosely packed protein core is beginning to
develop and it appears that the packing becomes less dense as surfactant
concentration is increased to 5 mM azoTAB. Above this concentration the
fluorescence changes only slightly indicating that the core of the protein is no longer
significantly changing. Under UV light as expected there is little to no increase in
Nile red fluorescence, supporting the idea that under UV light the protein remains in
a relatively compact, native-like state.
100
Crystal violet, a micropolarity probe, was also used to investigate the interior
of the protein. The absorption maxima of crystal violet depends upon the polarity of
the microenvironment surrounding the probe. For example, in a polar solvent such
as water, max
= 590 nm, while in a nonpolar solvent such as benzene, max
= 605
nm.(30) From Figure 3.5b it is clear that the maximum of the crystal violet
absorption spectra, max
, is shifting to higher wavelengths as the surfactant
concentration is increased under visible light. Since crystal violet is a relatively
planar molecule as is azoTAB, a stacking interaction between the molecules may be
expected, thus, both molecules would experience the same microenvironment. The
fact that at concentrations up to 2mM azoTAB max
has not yet increased
significantly indicates that crystal violet is still exposed to water, thus, both crystal
violet and azoTAB are bound mostly to the surface of the protein. However, at
concentrations higher than 2mM azoTAB, crystal violet is clearly experiencing a
much less polar environment indicating that crystal violet along with azoTAB has
entered the protein core. When compared to the Nile red fluorescence, we can see
that only a small increase is observed at azoTAB concentrations < 2 mM, while a
significant increase does not occur until azoTAB concentrations > 2 mM.
Furthermore, at concentrations > 5 mM the curves no longer change significantly.
The Nile red and crystal violet data agree with the shape reconstructions and are
similar to the DLS in that the protein does not appear to swell significantly until
azoTAB concentrations > 2 mM, while the protein continue to swell only slightly at
concentration > 5 mM. Again under UV light, max
, remains relatively constant until
101
a concentration of 8 mM is reached. As there is no SANS data at concentrations
above ~8 mM the shape of RNase A is unknown at these conditions. However, from
the DLS, Nile red fluorescence, and crystal violet max
values, one can speculate that
the protein is maximally swollen at a concentration of 8 mM and that it remains in
this state up to concentrations as high as 10 mM under visible light, supporting the
idea of surfactant saturation. However, under UV light, although max
is still
increasing significantly indicating that the protein is more swollen than at 8 mM
azoTAB, the hydrodynamic radius measured from DLS and the NR fluorescence
would indicate that the protein is not continuing to drastically swell.
3.5 Conclusion
Photoreversible conformations of RNase A are dependent only upon
surfactant concentration; they are independent of protein concentration and thus
surfactant to protein ratio. This is surfactant saturation effect is indicate of the
structural changes being largely dependent on hydrophobic forces. From shape
reconstructions of small angle neutron scattering data it is clear that the addition of
photosurfactant results in a swollen form of RNase A. The swelling appears to be
the direct result of surfactant interactions with the -helical regions making up the
major hydrophobic core within the protein. Upon exposure to UV light shape
reconstruction indicates that the protein refolds back to a conformation similar to its
native state. These structures are consistent with the unfolding of RNase A from
literature and the native X-ray crystallographic structure from the protein data bank.
102
3.6 References
(1) Zenkova, M. A. (2004) Artificial Nucleases, Vol. 13, Springer, Berlin.
(2) Raines, R. T. (1998) Ribonuclease A. Chem. Rev. 98, 1045-1065.
(3) Scheraga, H. A., Wedemeyer, W. J., and Welker, E. (2001) Bovine pancreatic
ribonuclease A: oxidative and conformational folding studies. Method. Enzymol.
341, 189-221.
(4) Wlodawer, A., Svennson, L. A., Sjolin, L., and Gilliland, G. L. (1988) Structure of
phosphate-free ribonuclease A refined at 1.26 A. Biochemistry 27, 2705-2717.
(5) Houry, W. A., and Harold, A. S. (1996) Nature of the Unfolded State of Evidence A:
Effect of Cis-Trans X-Pro Peptide Bond Isomerization. Biochemistry 35, 11719-
11733.
(6) Brandts, J. F., Halvorson, H. R., and Brennan, M. (1975) Consideration of the
Possibility that the slow step in protein denaturation reactions is due to cis-trans
isomerism of proline residues. Biochemistry 14, 4953-4963.
(7) Harrington, W. F., and Schellman, J. A. (1956) Evidence for the instability of
hydrogen-bonded peptide structures in water, based on studies of ribonuclease and
oxidized ribonuclease. Cr. Trav. Lab. Carlsb. 30, 21-43.
(8) Salahuddin, A., and Tanford, C. (1970) Thermodynamics of the denaturation of
ribonuclease by guanidine hydrochloride. Biochemistry 9, 1342-1347.
(9) Paz Andrade, M. I., Boitard, E., Saghal, M. A., Manley, P., Jones, M. N., and
Skinner, H. A. (1981) Enthalpy of interaction of ribonuclease A and n-alkyl
sulphates in aqueous solution. J. Chem. Soc., Faraday T. 77, 2939-2948.
(10) Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2007) Solution Structure of an
Amyloid-Forming Protein During Photoinitiated Hexamer-Dodecamer Transitions
Revealed Through Small-Angle Neutron Scattering. Biochemistry 46, 7694-7705.
(11) Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2005) Probing lysozyme
conformation with light reveals a new folding intermediate. Biochemistry 44, 15139-
15149.
(12) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2005) Photocontrol of Protein
Folding: The Interaction of Photosensitive Surfactants with Bovine Serum Albumin.
Biochemistry 44, 524-536.
103
(13) Wang, S.-C., and Lee, C. T., Jr. (2006) Protein Secondary Structure Controlled with
Light and Photoresponsive Surfactants. J. Phys. Chem. B 110, 16117-16123.
(14) Wang, S. C., and Lee, C. T., Jr. (2007) Enhanced Enzymatic Activity Through
Photoreversible Conformation Changes Biochemistry 46, 14557-14566.
(15) Le Ny, A.-L. M., and Lee, C. T., Jr. (2006) Photoreversible DNA Condensation
Using Light-Responsive Surfactants. J. Am. Chem. Soc. 128, 6400-6408.
(16) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2004) Photoreversible Viscosity
Changes and Gelation in Mixtures of Hydrophobically Modified Polyelectrolytes
and Photosensitive Surfactants. Macromolecules 37, 5397-5405.
(17) Hayashita, T., Kurosawas, T., Miyata, T., Tanaka, K., and Igawa, M. (1994) Effect
of structural variation within cationic azo-surfactant upon photoresponsive function
in aqueous solution. Colloid Polym. Sci. 272, 1611-1619.
(18) Shang, T., Smith, K. A., and Hatton, T. A. (2003) Photoresponsive Surfactants
Exhibiting Unusually Large, Reversible Surface Tension Changes under Varying
Illumination Conditions. Langmuir 19, 10764-10773.
(19) Glinka, C. J., Barker, J. G., Hammouda, B., Krueger, S., Moyer, J. J., and Orts, W. J.
(1998) The 30 m small-angle neutron scattering instruments at the National Institute
of Standards and Technology. J. App. Cryst. 31, 430-445.
(20) Svergun, D. I. (1992) Determination of the Regularization Parameter in Indirect-
Transform Methods Using Perceptual Criteria. J. Appl. Cryst. 25, 495-503.
(21) Eisenberg, H. (1976) Biological Macromolecules and Polyelectrolytes in Solution,
Lodon: Oxford Univ. Press.
(22) Eisenberg, H. (1981) Forward scattering of light, X-rays and neutrons. Q. Rev.
Biophys. 14, 141-172.
(23) Lehmann, M. S., and Zaccai, G. (1984) Neutron Small-Angle Scattering Studies of
Ribonuclease in Mixed Aqueous SOlutions and Determination of the Preferentially
Bound Water. Biochemistry 23, 1939-1942.
(24) Noeppert, A., Gast, K., Mueller-Frohne, M., Zirwer, D., and Damaschun, G. (1996)
Reduced-denatured ribonuclease A is not in a compact state. FEBS Lett. 380, 179-
82.
(25) Cole, R., and Loria, J. P. (2002) Evidence for flexibility in the function of
ribonuclease A. Biochemistry 41, 6072-6081.
(26) Oleinikova, A., Sasisanker, P., and Weingaertner, H. (2004) What Can Really Be
Learned from Dielectric Spectroscopy of Protein Solutions? A Case Study of
Ribonuclease A. J. Phys. Chem. B 108, 8467-8474.
104
(27) Dunbar, J., Yennawar, H. P., Banerjee, S., Luo, J., and Farber, G. K. (1997) The
effect of denaturants on protein structure. Protein Sci. 6, 1727-1733.
(28) Kirkwood, J. G. (1996) The general theory of irreversible processes in solutions of
macromolecules. J. Polym. Sci. Pol. Phys. 34, 1-14.
(29) Heller, W. T., Krueger, J. K., and Trewhella, J. (2003) Further Insights into
Calmodulin-Myosin Light Chain Kinase Interaction from Solution Scattering and
Shape Restoration. Biochemistry 42, 10579-10588.
(30) Mackay, R. A., Letts, K., and Jones, C. (1977) in Micellization, Solubilization, and
Microemulsions (Mittal, K. L., Ed.) pp 801, Plennum, New York.
105
CHAPTER IV: Solution Structure of an Amyloid-Forming Protein
During Photoinitiated Hexamer-Dodecamer Transitions Revealed
through Small-Angle Neutron Scattering
Andrea C. Hamill, Shao-Chun Wang, and C. Ted Lee, Jr.*
Department of Chemical Engineering and Materials Science, University of Southern
California, Los Angeles, California 90089-1211
Biochemistry 2007, 46, 7694-7705
Received February 2, 2007; Revised Manuscript Received April 18, 2007
4. 1 Abstract
Shape-reconstruction analysis applied to small angle neutron scattering
(SANS) data is used to determine the in vitro conformations of R-chymotrypsin
oligomers that form as a result of partial unfolding with a photoresponsive surfactant.
In the presence of the photoactive surfactant under visible light, the native oligomers
(dimers or compact hexamers) rearrange into expanded corkscrew-like hexamers.
Converting the surfactant to the photopassive form with UV light illumination causes
the hexamers to laterally aggregate and intertwine into dodecamers with elongated,
twisted conformations containing crosssectional dimensions similar to amyloid
protofilaments. Secondary-structure measurements with FT-IR indicate that this
photoinduced hexamer-to-dodecamer association occurs through intermolecular -
sheets stabilized with hydrogen bonds, similar to amyloid formation. Traditional
structural characterization techniques such as X-ray crystallography and NMR are
106
not easily amenable to the study of these nonnative protein conformations; however,
SANS is ideally suited to the study of these associated intermediates, providing
direct observation of the mechanism of oligomeric formation in an amyloid-forming
protein. Combined with photoinitiated hexamer-to-dodecamer associations in the
presence of the photoresponsive surfactant, this study could provide unique insight
into the amyloidosis disease pathway, as well as novel disease treatment strategies.
4.2 Introduction
Proteins interact with a variety of molecules during the course of activity,
ranging from small ions and ligands to other proteins through either heterogeneous
or homogeneous association. Indeed, the dynamic and multifarious response of
proteins to these interactions is utilized to stimulate or regulate virtually every
biological process. In some cases, however, protein interactions can result in
unwanted or deleterious effects, such as protein-protein associations leading to
amyloid fibril formation. The most well-known example of this process involves the
aggregation of the amyloid- (A )1 peptide fragments A 40 and A 42 implicated in
Alzheimer’s disease, although amyloid fibrils have been observed in an array of
proteins largely independent of the native secondary structure(1), including
ribonuclease A(2), an SH3 domain, lysozyme(3), insulin(4), and -chymotrypsin(5).
This process is generally believed to result from the formation of unstable slightly
unfolded conformations, leading to a cascading aggregation process from monomers
to oligomers [unstructured aggregates of typically multiples of six molecules in the
107
case of A 42(6)] to protofibrils (structured aggregates exhibiting -sheet structure)
to protofilaments (elongated aggregates 2-5 nm in diameter) to fibrils (2-6 entwined
protofilaments)(1). Perhaps most importantly, the prefibrillar intermediates (i.e.,
oligomers and protofibrils), which can induce cognitive impairment, have become
increasingly viewed as the primary pathogenic species(1, 7).
To date, however, the solution structure of these important intermediate
species remains unknown as the two preferred methods to determine protein
structure, namely, X-ray crystallography and solution NMR, are generally limited to
the study of native proteins in the solid state or relatively small protein assemblies,
respectively, since protein crystallization is often supplanted by unwanted
aggregation and crystal-packing constraints largely dominate protein orientations in
multimolecular complexes. Thus, the development of novel structural
characterization methods capable of examining partially folded proteins in non-
native conformations and supramolecular complexes undergoing self- or
heteroassociation is highly desired. For example, through the use of small-angle
neutron scattering (SANS), the in vitro structures of A 40 protofibrils were found to
be cylindrical with 24 Å cross-sectional radii and 110 Å long(8), while through AFM
and TEM measurements a variety of protofibril arrangements have been observed,
including twisted chains 2-5 nm in diameter. However, due to the relatively low
resolution of the cylindrical models employed in the above SANS analysis,
combined with potential influences of surface interactions with AFM and TEM, the
precise conformation of protofibrils in solution remains unknown. As a result, to
108
properly investigate intermediate conformations in an amyloid protein necessitates
two complementary approaches:(1) a means to induce changes in protein folding
and, hence, association in a controlled and preferably reversible manner and(2) a
method to determine the conformation of non-native and associated proteins at
relatively high resolution.
Recently, we have shown that light illumination can be used to induce
photoreversible changes in both the secondary(9) and tertiary(10, 11) structure of
proteins. This method utilizes the interaction of proteins with photosensitive
“azoTAB” surfactants containing an azobenzene group that undergoes a trans
(relatively hydrophobic) to cis (relatively hydrophilic) photoisomerization upon
exposure to visible (434 nm) or UV (350 nm) light illumination, respectively. Hence,
light can be used to reversibly bind the surfactant to the hydrophobic domains of
proteins, leading to photocontrol of protein folding. Furthermore, we have applied
small-angle neutron scattering (SANS) to study the in vitro structure of the non-
native protein conformations that form in response to photosurfactant and light.
Small-angle neutron and X-ray scattering have been used for several decades to
investigate the structure of soluble proteins in solution(12-15) and membrane
proteins in surfactant assemblies(16, 17). The obtained structures have typically been
low resolution, however, a consequence of modeling proteins with a single
dimension (radius of gyration) or as ellipsoids (axial radii). These procedures,
although convenient, belie the wealth of structural information contained within the
measured scattering intensity. From the range of momentum vectors
109
Q = 4
1
sin( /2), where is the neutron wavelength (6 Å) and is the scattering
angle, in a typical SANS experiment (Q = 0.005-0.5 Å
-1
), it can be seen that the data
span length scales (L = 2 / Q) ranging from 12.5 to 1250 Å, ideal for protein
conformational studies. Indeed, application of shapereconstruction techniques such
as the ab initio methods of GASBOR(14) and GA_STRUCT(15) reveals a high
degree of similarity between the native structure in solution (SANS) and in the solid
state (X-ray crystallography), a seemingly general property of soluble proteins(12).
In the present study, the ability to photoinitiate changes in protein quaternary
structure through photocontrol of -chymotrypsin self-association is demonstrated.
Native -chymotrypsin is well-known to self-associate through either a monomer-
dimer (pH 3) or monomer-hexamer (pH 7) equilibrium, while the addition of
trifluoroethanol, a solvent known to induce partially folded structures(18, 19), has
been reported to result in -chymotrypsin amyloid-fibril formation(5). Mixing -
chymotrypsin with the photoresponsive azoTAB surfactant is found to result in
partial unfolding of the protein, giving rise to changes in both the degree and type of
self-association. Shape-reconstruction analysis applied to SANS data allows
determination of the in vitro conformation of -chymotrypsin oligomers. In the
presence of azoTAB under visible light, native oligomers (dimer or compact
hexamers) are converted to expanded corkscrew-like hexamers, while upon UV light
illumination the hexamers laterally aggregate, wrapping around each other to form
dodecamers with twisted conformations. FT-IR measurements of the protein
secondary structure reveal that dodecamer formation is accompanied by hydrogen
110
bond stabilized intermolecular sheets, commonly observed in amyloid fibrils. TEM
measurements following incubation further confirm the formation of fibrillar
structures, while photocontrol of the hexamer-to-dodecamer association process is
studied with small-angle X-ray scattering (SAXS) measurements. Together, these
results provide the first direct observation of the mechanism of formation of the key
intermediates in an amyloid-forming protein, which could provide unique insight
into the amyloidosis disease pathway.
4.3 Experimental Procedures
4.3.1 Materials
An azobenzenetrimethylammonium bromide surfactant (azoTAB) of the form
similar to surfactants used in previous studies(9-11, 20) was synthesized according to
published procedures(21, 22). When illuminated with 350 nm UV light, the
surfactant undergoes a photoisomerization predominantly to the cis form (90/10
cis/trans), which can be rapidly reversed upon exposure to visible light (434 nm,
75/25 trans/cis) or in the dark in about 24 h (100% trans isomer)(23). For the SAXS
and FT-IR measurements, conversion to the UV light form was achieved with the
365 nm line from a 200 W mercury arc lamp (Oriel, model 6283), isolated with the
combination of a 320 nm band-pass filter (Oriel, model 59800) and an IR filter
111
(Oriel, model 59060). A 400 nm long-pass filter (Oriel, model 59472) was used to
convert back to the visible light form. In the SANS experiments, the solutions were
exposed to an 84 W long-wave UV lamp (365 nm; Spectroline, model XX-15A) for
at least 30 min prior to sample collection to convert to the UV light form and were
continuously exposed to the same UV light throughout the data collection.
Type II, essentially salt-free -chymotrypsin from bovine pancreas (Sigma,
catalog number C-4129, lot 105K7670), five times crystallized chymotrypsinogen-A
from bovine pancreas (Worthington, catalog number LS005630), and phosphate
buffer (Sigma, catalog number P-3288, pH 7.2, 8.3 mM) were used as received. All
other chemicals were obtained from Aldrich in the highest purity. For the
experiments performed at pH 3, HCl (37%) was added to the pH 7.2 buffer as
needed.
4.3.2 Small Angle Neutron Scattering
Small-angle neutron scattering experiments were performed on the 30 m
NG3 SANS instrument at NIST(24). A neutron wavelength of = 6 Å and a detector
offset of 25 cm with two sample-detector distances of 1.33 and 7.0 m were utilized to
achieve a Q range of 0.0048-0.46 Å
-1
. The net intensities were corrected for the
background and empty cell (pure D2O), accounting for the detector efficiency using
the scattering from an isotropic scatterer (Plexiglas), and converted to an absolute
differential cross section per unit sample volume (in units of cm-1) using an
attenuated empty beam. The data were then corrected for incoherent scattering by
112
subtracting a constant background. The shapereconstruction algorithm
GA_STRUCT(15) was used to generate solution conformations, similar to previous
studies(10, 11). Beginning with an initial guess of randomly distributed scattering
centers, the program rearranges the position of the scattering centers to best fit the
experimental scattering data.
The weight-average molecular weight (Mw) of each sample was calculated
from the equation
M
W
=
1000I(0)N
A
c 2
( P
S
)
2
where s
and P
are the scattering length densities of the solvent (6.36 x 10
10
cm
-2
)
and protein (3.23 x 10
10
cm
-2
), respectively, c is the protein concentration (11.6
mg/mL at pH 3 and 11.4 mg/mL at pH 7), and is the protein specific volume
(0.734 cm
3
/g) (25). I(0) values were determined from Guinier plots(25) using I(Q) =
I(0) exp(-Q
2
R
g
2
/ 3), where R
g
is the radius of gyration. Guinier plots, generally valid
for QR
g
< 1.3, can be influenced by solution structuring due to intermolecular
interactions between charged proteins, which becomes increasingly important as Q
decreases below 1.5/R (10) (or <0.05 Å
-1
using an -Ch radius ~30 Å). Thus, pair
distance distribution functions were calculated from the SANS data using the
program GNOM(26) according to the equation
I(Q) = 4 P(r)
sin(Qr)
Qr
dr
0
D
max
where P(r) is related to the probability of two scattering centers (nuclei for SANS)
being a distance r + dr apart and D
max
is the maximum distance between scattering
113
centers within the protein or protein oligomer. I(0) values were then obtained from
the PDDFs through I(0) = 4 P(r)dr
0
D
max
, which has the advantage of utilizing the
entire Q range to determine I(0), as opposed to just the low Q values as in Guinier
analysis(27).
4.3.3 Small Angle X-ray Scattering
The small-angle X-ray scattering data were measured using the X21 beamline
at the National Synchrotron Light Source at the Brookhaven National
Laboratory(28). The X-ray wavelength was set to 1.24 Å with a pair of Si(111)
monochromator crystals. The sample-to-detector distance was calibrated to be 1.69
m using a silver behenate standard. To avoid radiation damage, solutions were
continuously passed at a flow rate of 60 μL/min through a 1 mm glass capillary
housed within an aluminum block containing Plexiglas observation windows(28).
The net intensities were corrected for the background and solvent scattering, as well
as sample transmission, and were put on an absolute scale by comparison with a
calibration standard [10 mg/mL BSA(10)].
4.3.4 Fourier Transform Infrared Spectroscopy
Infrared spectra were measured with a Genesis II FT-IR spectrometer
(Mattson Instruments). Solutions were loaded in a demountable liquid cell equipped
with a circulating water jacket (T = 20 °C) between a pair of CaF
2
windows using a
50 μm Teflon spacer. A liquid light guide (Oriel, model 77557) was used to directly
114
illuminate the sample with UV or visible light for 2 h prior to and during data
collection, as previously described(9). The sample chamber was continuously
purged with dry air to eliminate the influence of water vapor. For each spectrum, a
500-scan interferogram was collected with a 2 cm
-1
resolution. The relatively sharp
surfactant peaks at ~1600 cm
-1
were removed by subtracting the spectra measured for
a pure surfactant solution under otherwise identical conditions, resulting in corrected
spectra that were flat in the region between 2000 and 1750 cm
-1
. Fourier self-
deconvolution (FSD) was applied to spectra to resolve the overlapping bands in the
amide I region using a band-narrowing factor k = 2.0 and a full width at half-height
of 12.6 cm
-1
. Second derivative spectra were obtained with the Savitsky-Golay
function for a third order polynomial, using a 13 data point window. Difference
spectra were obtained by subtracting the spectra collecting under visible light from
the spectra collecting under UV light illumination. Difference spectra obtained for
pure -chymotrypsin solutions without surfactant show no significant absorbance
changes (<1% throughout the amide I region).
4.3.5 Microscopy
Optical Microscopy was performed on an Olympus IX71 inverted
fluorescence microscope equipped using a 50x lens (Olympus, model SLCPlanFl)
and a U-N41027 CAL CRIM C58158 filter cube (Chroma, model C58158). Images
were recorded with a Hamamatsu digital CCD camera (model C4742-95). Aliquots
(5-10 μL) of the protein-surfactant solution were deposited onto glass slides and
115
dyed with an equal volume of a 400 μM Congo red aqueous solution. Cross
polarizers were used to image birefringence from amyloid fibrils and maltese-cross
patterns form amyloid sphereulites.
Transmission electron microscopy was performed on a Philips EM420 TEM
operating at 80 kV. A drop of protein solution was placed on a carbon-coated grid
for 10 s and then blotted with filter paper, followed by repeating this procedure with
a second drop. The grid was then placed in a freshly made 1 wt % uranyl acetate
solution for 30 s.
4.3.6 UV-visible Spectroscopy
Absorption measurements were performed on an Agilent model 8453 UV-
visible spectrophotometer using a 5 mm path length quartz cell. 15 μL of a
concentrated bromophenol blue aqueous solution were added to 1.5 mL of protein
solution to obtain a final dye concentration of 10 μM. Bromophenol blue exhibits a
maximum absorbance at 592 nm in water, thus, detectible independent of the
azoTAB adsorption peaks.(20, 23)
4.4 Results and Discussion
SANS data for -Ch/azoTAB mixtures are shown in Figure 4.1 as a function
of pH, surfactant concentration, and light illumination. AzoTAB undergoes a
photoisomerization to the relatively hydrophilic cis form when illuminated with 350
nm UV light, which can be reversed back to the relatively hydrophobic trans form
116
upon exposure to 434 nm visible light(23). In inverse space (e.g., with Q in units of
Å
-1
), the transitions responsible for SANS intensity changes in Figure 4.1 can be
difficult to conceptualize; thus, the real space length scale L = 2 / Q) is plotted on
the upper x-axis. The addition of azoTAB causes an increase in scattering at low Q
(i.e., L > 100 Å), suggesting the surfactant induces monomer oligomer
associations. UV light illumination further enhances this effect, with a shift in the
scattering curves to lower Q indicating greater protein aggregation when the
surfactant is converted to the cis form. Thus, the trans isomer appears to be capable
of replacing protein-protein interactions with protein-surfactant interactions. Beyond
Q > 0.2 Å
-1
(L < 30 Å, or length scales less than the protein diameter) the SANS data
converge, suggesting that the individual protein subunits remain relatively intact.
However, at high Q the limiting sensitivity of the SANS data is approached due to
weak sample scattering relative to incoherent scattering from the protein (0.003 cm
-1
)
and solvent (0.0004 cm
-1
for 99.9% D2O).
117
Figure 4.1: SANS data of -chymotrypsin/azoTAB solutions as a function of surfactant
concentration and light illumination at (a) pH 3.0, 11.6 mg/mL protein, and (b) pH 7.2, 11.4
mg/mL protein.
For associating systems, SANS has two advantageous properties. First,
SANS is an absolute technique with the weight-average molecular weight (Mw) of
the sample given directly by I(0), the scattering at zero angle (see also Experimental
Procedures). Thus, the weight fraction of protein existing as monomer (x
1
) and n-mer
(x
n
= 1 - x
1
) can be calculated from Mw = x
1
Mw,
1
+ (1 - x
1
)Mw,
n
, where Mw,
1
and
Mw,
n
are the monomer and n-mer molecular weights, respectively. Second, SANS is
additive with the scattering for a mixture of monomer (1-mer) and n-mer species
given by the sum of the contributions from each oligomer o from 1 to n(29)
I(Q) =n
p
N
o
N
F
o
(Q)
2
o=1
n
S(Q)
118
where n
p
=1/VN
o
o=1
n
=N /V is the total number of particles per volume, N
o
/N is the
number fraction of a given type of oligomer, F
o
(Q) is the form factor for that
oligomer, and S(Q)is the averaged structure factor related to the partial structure
factors S
ij
(Q). Hence for a noninteracting mixture of monomer and a single n-mer
(30), the scattering intensity can be shown to beI =v
1
I
1
+v
n
I
n
, where v
1
and v
n
are the
fractions of protein existing as monomer and n-mer on a volume basis (not to be
confused with the volume fraction in solution, =c /1000, where c is the protein
concentration in mg/mL units and is the protein specific volume), while I
1
and I
n
are the scattering from pure monomer and n-mer, respectively(27). Since v
i
~x
i
,
assuming that is constant independent of oligomeric state, the total scattering
intensity is then also given by the linear combination I = x
1
I
1
+ x
n
I
n
(31). Thus, these
two properties of SANS can be utilized to assign the contributions of the overall
scattering to the monomer and n-mer, followed by shape reconstruction to determine
the in vitro structure of -Ch oligomers. In the sections that follow this will first be
illustrated for pure -Ch and then extended to solutions containing azoTAB to
demonstrate photocontrolled -Ch association.
4.4.1 Pure Protein Solutions
-Ch is well-known to selfassociate through either a monomer-dimer (pH 3)
or monomer-hexamer (pH 7) equilibrium at low ionic strength(32-34), with a
reduction in the overall positive charge of the protein (pI = 9.1) with increased pH
119
generally allowing for greater association. The SANS data for pure -Ch solutions
shown in Figure 4.1, measured at conditions where self-association is expected to be
prevalent (10 mg/mL protein), are replotted in Figure 4.2.
Figure 4.2: SANS data of pure -chymotrypsin solutions at (a) pH 3 and (c) pH 7: raw data
( ), scaled PDB (2CHA) (–), and raw data minus the scaled PDB ( ). Insets show Guinier
analysis of the raw data. Also shown are pair distance distribution functions at (b) pH 3.0
and (d) pH 7: raw data (–) and oligomer-only data (dotted). For comparison the raw data
minus the oligomer data (dashed) and 2CHA (–) are displayed.
The raw data are largely featureless due to the presence of both monomer and
oligomer in solution, complicating quantitative analysis of the self-association
process. To deconvolute the scattering data, the weight-average molecular weight
120
(Mw) of each sample was calculated from the scattering at zero angle, reported as the
effective oligomer size (n
eff
= M
w
/M
1
) in Table 4.1.
Table 4.1: Values of Radius of Gyration (R
g
), I(0), Resulting n-mer (n
eff
), Fraction of
Oligomer (x
n
), and Radius of Gyration of Oligomer-Only Data (R
g
n
) Determined from
Guinier or PDDF Analysis of SANS Data in Figure 4.1
I(0) values determined from both Guinier plots(25) using I(Q) = I(0) exp(-
Q
2
R
g
2
/ 3), technically valid for QR
g
< 1.3 as discussed below, where R
g
is the radius
of gyration (Figure 4.2a,c, insets), and from the entire Q range using pair distance
distribution functions (PDDFs) in Figure 4.2b,d are generally consistent. Note that
121
the steep upturn in the Guinier plots at Q < 0.01 Å-1 (L > 600 Å) could be due to the
presence of a small amount of higher order aggregates, which due to the
characteristic Q-4 decay would not be expected to influence the data analyses
employed below.
From the Mw value determined at each pH, the weight fractions of monomer
and n-mer were calculated. The portion of the scattering resulting from free
monomers, estimated from the monomer PDB file 2CHA using the program
CRYSON(35) and scaled with a monomer concentration of x
1
c as shown in Figure
4.2a,c, was then subtracted from the overall scattering to give the oligomer-only
SANS data. This procedure presumes that the structure of the monomer in vitro is
well represented by the native X-ray crystallographic structure, shown to be true for
a range of soluble proteins(12, 36, 37). Compared to the featureless raw data, the
oligomer scattering curve at pH 7 displays a prominent peak at Q = 0.14 Å
-1
,
translating in real space to L ~ 45 Å, also detected as a peak in the corresponding
PDDF curve. This dimension corresponds to a highly probable distance within the
protein oligomer, namely, the separation distance between monomers. Peaks in this
Q range signify well-ordered oligomer conformations and are often used as
qualitative tests for oligomer formation(38), lending confidence in the deconvolution
procedure. Oligomer peaks become more pronounced with increasing n-mer size;
thus, it is not surprising that the dimer data at pH 3 do not display this peak.
Following deconvolution, shape reconstructions of the oligomer-only data
were performed, conceptually similar to previous studies used to determine the in
122
vitro structures of partially folded BSA(9, 10) and lysozyme(11). The GA_STRUCT
program begins with chains of randomly oriented “scattering centers” (i.e., atomic
nuclei for SANS), with a genetic algorithm consisting of matings, mutations, and
extinctions used to update the shape(15). Despite this general procedure, the dimer
(pH 3) and hexamer (pH 7) structures indeed contain n subunits for each n-mer, as
shown in Figure 4.3. Interestingly, the SANS-based in vitro dimer is not consistent
with the “face-to-face” (active site-to-active site) crystal packing of R-chymotrypsin
and is instead better represented by the “back-to-side” packing of chymotrypsinogen
(2CGA; note that -Ch results from the removal of two dipeptides at positions 14-15
and 147-148 in chymotrypsinogen). For example, the maximum dimension of 6CHA
70 Å, while the PDDF in Figure 4.2b gives a Dmax of 90Å compared to 85 Å for
2CGA. This serves to highlight the influence that crystal-packing constraints can
have on molecular orientations, a significant advantage of SANS in the study of
protein aggregates, and could explain why the role of the active site in -Ch
association remains unsolved in the literature with different techniques yielding
conflicting results(32).
123
Figure 4.3: Shape reconstructions of the oligomer-only SANS data for pure -chymotrypsin
at (a) pH 3 and (b) pH 7 (in blue) compared to the X-ray crystallographic structures of the -
chymotrypsin dimer (PDB code 6CHA) and chymotrypsinogen-A dimer (PDB code 2CGA)
at pH 3. The insert shows a hypothetical hexamer built from three 2CGA subunits
(alternating monomers shown in blue and green) along with the consensus envelope and
worst fit (shown in red).
Shape reconstruction of the pH 7 data reveals the compact, “W-shaped”
hexamer shown in Figure 4.3b. The average distance between nearest-neighbor
subunits is 43 ± 5 Å, in agreement with the 0.14 Å
-1
peak in Figure 4.2c, while the
orientation angle between the centers of mass of three successive subunits is
estimated as 70 ± 10° from the 3D shape reconstructions. The consistency of these
values suggests that specific intermolecular interactions are responsible for hexamer
formation in solution, resulting in the twisted arrangement of the subunits.
Interestingly, the ribbon diagram of a hypothetical hexamer constructed by
continuing the relative orientation of the two macromolecules in 2CGA (with
alternate proteins color-coded blue and green) exhibits a similar twisted orientation,
124
with all but the final protein in nearly identical locations. In contrast, the face-to-face
arrangement of 6CHA would not support higher order association, as opposed to the
“heterologous association” apparently observed in Figure 4.3a(31). Also shown in
the inset of Figure 4.3 is the consensus envelope obtained by docking and averaging
ten independent fits of the GA_STRUCT program, along with the run that
statistically produced the worst fit to the data. Both of these structures agree with the
W-shaped hexamer conformation, demonstrating that the coupled deconvolution /
shape-reconstruction technique can be applied to protein oligomers in solution. Non-
native protein conformations such as partially folded or associated states challenge
existing crystallographic and NMR methods. However, as demonstrated in Figure
4.3 and in recent studies of photocontrolled protein folding(10, 11), SANS can
provide valuable information on these important yet understudied class of structures.
4.4.2 -Chymotrypsin/azoTAB Solutions
As discussed above, qualitative assessment of the SANS data in Figure 4.1
indicates enhanced -Ch association with either increased surfactant concentration
or upon converting azoTAB from the trans to the cis form with UV light
illumination. To quantitatively investigate this phenomenon, Guinier plots of the
SANS data for -Ch/azoTAB solutions were generated, as shown in Figure 4.4.
Two unique slopes can be detected at each condition, the first in the region of Q
2
<
0.002 Å
-2
corresponding to the z-average radius of gyration of the mixture and the
second at Q
2
~ 0.01-0.03 Å
-2
with R
g
1
values ranging from 17.0 to 18.2 Å, as shown
125
in Table 4.1. These latter values are consistent with the R
g
of monomeric -Ch in the
literature of 16.9 Å(39), thus indicating a monomer/n-mer equilibrium(40, 41)
similar to the monomer --> oligomer equilibrium observed during the early stages of
fibril formation of A proteins(1, 8). The 7% increase in R
g
1
with the addition of
azoTAB in Table 4.1 indicates that a slight unfolding of the protein could be the
cause of increased association, consistent with the general observation that partially
unfolded protein conformations can lead to amyloid fibril formation(1). For example,
intermediate trifluoroethanol concentrations where non-native conformations are
stabilized through still favorable hydrogen bonding between peptides(42) can induce
amyloid formation in a variety of proteins including -Ch(5). The amphiphilic nature
of azoTAB has also been shown to stabilize partially unfolded intermediates in
proteins such as BSA and lysozyme (10, 11). It should be pointed out, however, that
the Guinier region is strictly valid only for QR
g
< 1.3, while the above fits span Q
2
=
0.01-0.03 Å
-2
(QR
g
= 1.7-3). Replacing [3
j1
-(QR
g
)/QR
g
]
2
with the approximate
expression exp(-Q
2
R
g
2
/ 3), as suggested by Guinier(25), results in deviations on the
order of 10% over this Q range.
126
Figure 4.4: Guinier analysis of the raw SANS data. (a) pH 3: pure -Ch ( ), 1.59 mM
azoTAB visible ( ) and UV ( ), 4.23 mM visible ( ) and UV ( ), and 6.70 mM visible
( ) and UV ( ). (b) pH 7: pure -Ch ( ), 1.03 mM visible ( ) and UV ( ), 4.03 mM
visible ( ) and UV ( ), and 9.92 mM visible ( ) and UV ( ). Data sets are successively
offset by 0.5 logarithmic units.
From the fits in the low Q region of Figure 4.4, the R
g
values are
approximately constant at a given pH and light condition, suggesting that the
oligomer size is primarily determined by the state of the surfactant. I(0) values
determined from either the Guinier plots in Figure 4.4 for Q
2
< 0.002 Å
-2
or PDDFs
of the overall data (not shown) are displayed in Table 4.1, along with the effective
oligomer size (n
eff
). Together, these values suggest monomer-hexamer equilibrium
for the visible light data and a monomer-dodecamer equilibrium under UV light;
however, unlike pure R-chymotrypsin the oligomer size is not known a priori.
Nevertheless, additional evidence will be presented below to support this type of
protein selfassociation. In truth, these n-mer assignments are a result of a
comprehensive iterative procedure whereby the number of protein subunits observed
from shape reconstruction of the raw data (monomer plus oligomer) was used to
127
provide initial estimates of n. However, since fitting the overall data would return the
z-averaged shape of the protein(43), which is heavily weighted toward the oligomer
conformation, 6 and 12 subunits could be consistently detected even from the raw
data (see below). Furthermore, for the UV data, the choice of n = 12 was particularly
clear given that both the SANS and SAXS data (Figure 4.8 below, to be discussed in
detail later) appear to converge to an n
eff
value of 12 with increasing surfactant
concentration.
It is important to note that the measured increases in I(0) cannot simply be
the result of surfactant binding alone. First, by comparing the surfactant to protein
molar ratios, only between 2.5 and 25 surfactant molecules could possibly be bound
per protein over the range of surfactant concentrations studied assuming complete
surfactant binding. This would equate to at most an increase in the effective
molecular weight of between 4% and 40%, much smaller that the 12-fold increase
observed during dodecamer formation. Thus, it is immediately clear that the primary
role of the surfactant is to induce protein aggregation. Nevertheless, control
experiments were performed at the protein contrast-matching point (60/40
H
2
O/D
2
O), thereby rendering the protein “invisible” to the neutron beam. The SANS
spectrum obtained from this contrast-matched sample (not shown) was within the
experimental error, indicating minimal to no scattering from azoTAB. Hence, the
increases observed in this report can be conclusively attributed to protein
aggregation.
128
From the resulting n-mer assignments above, the monomer weight fraction
was calculated, indicating that the monomeroligomer equilibrium shifts toward n-
mer formation with increasing surfactant concentration, a likely result of increased
partial unfolding as mentioned above. To gain further insight in the oligomer
structures, the SANS data were deconvoluted as above to obtain the portion of the
scattering due only to the n-mers. As shown in Figure 4.5, the x
n
-scaled oligomer
scattering data are largely consistent for a given pH and light conditions, suggesting
a sound deconvolution procedure. Some subtle changes are observed with increasing
surfactant concentration within a given data set, particularly in the high Q region
representing fine structural detail. Specifically, the peak observed at Q ~ 0.2 Å
-2
,
similar to the deconvoluted hexamer of pure -Ch at pH 7, becomes “washed out”
with increasing azoTAB concentration, suggesting that the oligomers become more
disordered with increased fluctuations in the protein subunit positions. Using Guinier
plots (not shown) to calculate of the radius of gyration from each oligomer-only
scattering profile gives the values of R
g
n
reported in Table 4.1.
PDDFs calculated from the oligomer-only data display a similar degree of
homogeneity at each condition with increasing surfactant, as shown in Figure 4.5a,b.
Interestingly, independent of oligomer type (hexamer or dodecamer)
R
g
n
(M
w
)
0.42±0.03
compared to the monomer radius of gyration, where M
w
is the
molecular weight of the oligomer. A similar scaling exponent (0.45) has been
reported for self-associating insulin, with values intermediate between those
129
expected for spheres (1/3) or Gaussian coils (1/2), suggesting relatively open
oligomer structures(30).
Figure 4.5: Pair distance distribution functions of the oligomer only data scaled by the
respective oligomer weight fractions. (a) pH 3: pure -Ch ( ), 1.59 mM azoTAB visible ( )
and UV ( ), 4.23 mM visible ( ) and UV ( ), and 6.70 mM visible ( ) and UV ( ). (b)
pH 7: pure -Ch ( ), 1.03 mM visible ( ) and UV ( ), 4.03 mM visible ( ) and UV ( ),
and 9.92 mM visible ( ) and UV ( ). Also shown are the SANS scattering curves of the
oligomer-only data. (c) pH 3: pH 7: pure -Ch ( ) 1.59 mM azoTAB visible ( ) and UV
( ), 4.23 mM visible ( ) and UV ( ), and 6.70 mM visible ( ) and UV ( ). (d) pH 7: pure
-Ch ( ), 1.59 mM azoTAB visible ( ) and UV ( ), 4.23 mM visible ( ) and UV ( ), and
6.70 mM visible ( ) and UV ( ).
130
Comparing the visible light (hexamer) PDDFs to Figure 4.2d for the pure
hexamer reveals a shift in the PDDF peak to lower r values. This suggests a potential
unraveling of the tightly packed W-shaped hexamer with the most probable
dimension being reduced to distances within the protein subunits (e.g., the protein
radius) as opposed to distances between the subunits. However, note that D
max
of the
hexamer remains at ~ 120 Å as in Figure 4.2d; thus, only partial unraveling can be
occurring, largely retaining the twisted hexamer conformation. For example, a linear
n-mer formed from a protein with a radius of 20 Å would give peaks at 20, 40, ..., (n
- 1)40 Å. For the dodecamer structures the most probable dimension returns to 40-50
Å, while D
max
undergoes a modest increase to ~ 160 Å; hence, longitudinal extension
of hexamers to form dodecamers does not appear to be an appropriate mechanism.
Shoulders can also be detected in the PDDF curves at ~ 80, 100, and 120 Å
corresponding to distances between higher order neighbors, suggesting regular, as
opposed to random, oligomer conformations. Guinier analyses of the oligomer-only
data (not shown) give radii of gyration of the n-mers (R
g
n
) consistent with the PDDF
analysis as displayed in Table 4.1, again largely independent of surfactant
concentration across a given data set. Taken together, this evidence suggests that
converting azoTAB to the cis form with UV light causes hexamers to laterally (as
opposed to longitudinally) associate into dodecamers.
To obtain a better understanding of the oligomer conformations, shape
reconstruction was applied to the deconvoluted SANS data, as shown in Figure 4.6
(above). In all cases the shapereconstruction algorithm returned conformations
131
containing either 6 or 12 subdomains, despite the fact that the program begins with a
random arrangement of scattering centers. This fact further confirms the choice of
hexamers and dodecamers, as well as the overall deconvolution procedure. The
shapereconstructed hexamers indeed support the notion above of an unraveling of the
W-shaped hexamer, as the hexamers now have extended, corkscrew-like
appearances. Upon UV illumination and conversion of the surfactant to the cis form,
hexamers are converted into the rope-like dodecamers, suggesting that dodecamer
formation results from lateral association of two hexamers. This is illustrated by the
beadmodel structures accompanying each 90° rotation view of the oligomers, used to
guide the eye as to the relative positions of each protein subunit. The observed n-mer
structures are found to be reasonably consistent across the range of pH and surfactant
concentration conditions, again pointing to the global consistency of the
deconvolution procedure.
132
Figure 4.6: Shape reconstructions of the oligomer-only SANS data at pH 3 and pH 7 as a
function of azoTAB concentration and light illumination. Inserts show four views of the
shape reconstructions rotated at 90°, along with low-resolution globular models designed to
mimic the twisted conformations detected in the structures.
4.4.3 Photoinduced -Chymotrypsin Oligomers Are Amyloid Precursors
The lateral association of hexamers into dodecamers is consistent with the
eventual rope-like conformation commonly observed in many amyloid fibrils,
indicating that SANS may be reporting on the mechanism of formation of key
prefibrillar intermediates in the amyloid cascade. To investigate whether the
oligomer structures in Figure 4.6 are true prefibrillar intermediates, several classic
amyloid tests were performed on azoTAB/ -Ch mixtures. FT-IR spectra of pure -
Ch and -Ch in the presence of azoTAB under both visible and UV light are shown
133
in Figure 4.7a,b. Two aggregation processes can be triggered in the -Ch/azoTAB
system: the first upon the addition of trans azoTAB to pure -Ch (dimers
hexamers at pH 3) and the second upon exposure of the -Ch/azoTAB system to UV
light (hexamers dodecamers). As seen in the FT-IR spectra, both of these
association processes give rise to an increase in peaks at 1612 and 1685 cm
-1
,
characteristic of intermolecular -sheet formation(5, 44, 45), at the expense of the
peak at 1637 cm
-1
commonly assigned to intramolecular -sheets(46, 47). Zurdo et
al. observe bands at 1612 and 1685 cm
-1
in SH3 domain protofibril intermediates that
eventually mature into fully developed amyloid fibrils(45), suggesting that the
oligomers observed in Figure 4.6 are indeed precursors to amyloid structures.
134
Figure 4.7: (a) FT-IR absorbance spectra at pH 3 for pure -Ch (black line) and mixtures of
-Ch with 9.04 mM azoTAB under visible light (red line) and UV light (blue line). [ -Ch] =
11.6 mg/mL. (b) FT-IR difference spectra (UV - visible) demonstrating the effect of light
illumination. Also shown are Congo red fluorescence (c) and apple green birefringence (d)
obtained under cross-polarizers, as well as TEM images of a fresh solution (e, f) (pH 3,
[azoTAB] = 4.95 mM) and an original SANS solution (g) (pH 3, [azoTAB] = 4.23 mM)
after an elapsed time of approximately 1 year.
The photomicrographs shown in Figure 4.7c,d further support this
conclusion. Congo red staining of a -Ch/ azoTAB solution aged for five days
results in characteristic Congo red fluorescence as well as “apple green”
birefringence, respectively. Congo red preferentially stains amyloid structures due to
the planar structure of the dye favoring incorporation into the -sheet structure of
amyloids(48-50). These images were also accompanied by Maltese-cross patterns
under cross-polarizers (Figure 4.10) indicative of spherulites formed by the aligning
of fibrils in a radial pattern(51).
135
TEM images in Figure 4.7 further demonstrate the formation of fibrillar
structures. Panels e and f of Figure 4.7 were obtained 2 weeks after preparing a fresh
-Ch/azoTAB solution, while Figure 4.7g was obtained from an original SANS
solution (pH 3, [azoTAB] = 4.2 mM) approximately 1 year after collecting the SANS
spectra. The fibrils shown in Figure 4.7e-g possess clear amyloid characteristics:
they are long, unbranched, and appear to be twisted, with diameters of ca. 10 nm.
Combined, these tests confirm that the structures obtained from the SANS
measurements in Figure 4.6 are indeed preamyloid oligomer intermediates.
4.4.4 Photocontrol of Protein Association
To investigate the photocontrol of protein association, small-angle X-ray
scattering (SAXS) data were collected for mixtures of chymotrypsinogen-A and
azoTAB at pH 3, as shown in Figure 4.8. Chymotrypsinogen is the zymogen of R-
chymotrypsin, activated by the removal of two dipeptides at positions Ser14-Arg15
and Thr147-Asn148 leading to the formation of the active site(52). Despite this
structural similarity, however, chymotrypsinogen does not generally associate in
solution unlike in the case of -chymotrypsin(53, 54). This phenomenon is
supported by the visible light SAXS data in Figure 4.8a, where a clear intermolecular
interaction peak is observed in contrast to Figure 4.1, consistent with increasing
electrostatic repulsion between chymotrypsinogen monomers as the cationic
surfactant binds to the positively charged protein. A Guinier plot of the pure protein
SAXS data (Figure 4.8 inset) gives R
g
= 17.1 Å, similar to the SAXS derived R
g
136
value from the literature of 17.6 Å(39). With increasing surfactant concentration, R
g
increases modestly up to 10 mM azoTAB under visible light, eventually increasing
to 19.7 Å at 24 mM azoTAB (Table 4.2). The enhanced negative deviations from the
Guinier behavior at low Q with increasing surfactant concentration are a result of
increasing intermolecular interactions.
Figure 4.8: SAXS data of chymotrypsinogen-A/azoTAB solutions as a function of surfactant
concentration and light illumination at pH 2.7 and 10.0 mg/mL protein under (a) visible light
and (b) UV light. Pure -Ch ( ), 2.5 mM azoTAB (gray diamond), 5 mM (gray square), 10
mM (gray circle), 14 mM ( ), 19 mM ( ), and 24 mM ( ). Also shown are the effects of
reillumination with visible light on UV-adapted samples at various exposure times: 19 mM,
1 h (+ in square), 14 mM, 1 h (+ in diamond), 10 mM, 1 h (+ in circle), and 2 h ( in circle).
Insets show Guinier fits with successive data sets offset by 0.5 logarithmic units.
Under UV light illumination, however, the situation is markedly different,
with large increases in the SAXS data observed at low Q (note that the y-axes of
panels a and b of Figure 4.8 differ by an order of magnitude), particularly at 10 mM
azoTAB and beyond, coincidently the surfactant concentration where the onset of
chymotrypsinogen unfolding was observed under visible light. The Guinier plots
under UV light also reveal the development of an additional larger species with
137
increasing surfactant concentration, detected by the appearance of a steep slope at
low Q. I(0) values for samples under visible light [~0.2-0.25 cm
-1
; see Table 4.2,
where the SAXS data have been put on an absolute scale by comparing to a
calibration standard of 10 mg/mL BSA (10)] are consistent with the value expected
for the monomer [I(0) = 0.24 cm
-1
], again indicating that chymotrypsinogen
association does not occur under visible light. Under UV light a 12-fold increase in
I(0) is observed at 19 and 24 mM azoTAB relative to the monomer data, suggesting
that the association equilibrium is pushed entirely toward dodecamers, providing
independent confirmation of the -chymotrypsin data.
Partial reversibility of protein self-association is shown in Figure 4.8b, where
SAXS spectra were collected for UV equilibrated samples following reexposure to
434 nm visible light. The low Q scattering decreases as a function of visible light
exposure time, with apparently several hours required for complete visible light
induced dissociation (beyond the limit of the allocated SAXS beam time). However,
it should be pointed out that this dissociation process is not limited by the cis
trans isomerization kinetics, which occurs within minutes(23). Protein association
and dissociation can generally occur on time scales ranging from seconds up to hours
or even several days(55, 56). Thus, the SAXS data demonstrate the possibility of
photoreversible control of protein oligomerization.
138
4.4.5 Photoreversibility
To further examine photoreversible protein self-association, UV-vis
spectroscopic data were collected using bromophenol blue (BPB) to probe -
chymotrypsin/azoTAB mixtures, as shown in Figure 4.9a. BPB is a common pH
indicator, exhibiting an acidic (protonated) peak at 435 nm and a basic
(deprotonated) peak at 590 nm. Furthermore, the deprotonated peak of BPB has been
shown to decrease and undergo a red shift to ~605 nm upon binding to BSA,(57) a
result of a less-polar microenvironment of the probe. As expected, at pH 3 the
absorbance of BPB (~0.38) is lower than the value measured at pH 7 (~0.5,
estimated from the value of 0.58 at 600nm minus the value of 0.08 at 800 nm due to
scattering, discussed below). Upon UV-light illumination the deprotonated peak at
each pH decreases, while at pH 7 a further red shift in the absorbance maximum is
observed. This is a result of greater protein association under UV light, creating more
non-polar regions for probe binding.(58) The shift at pH 7 is due to the BPB
experiencing a non-polar environment, this could be due to the formation of a BPB-
-chymotrypsin complex. BPB has been shown to bind to the active site of -
chymotrypsin.(59) -chymotrypsin is in the inactive state at pH 3 and the active state
at pH 7,(60) so most likely the change in active site at pH 3 keeps BPB from binding
directly to the active site.
139
Figure 4.9: (a) UV-vis absorbance spectra of 10uM BPB in -chymotrypsin/azoTAB
solutions as a function of surfactant concentration and light illumination, 10.0 mg/mL
protein (b) reversibility of -chymotrypsin association
The dynamics of these absorbance changes in the BPB system upon repeated
exposure to UV and visible light are shown in Figure 4.9b, demonstrating complete
photoreversibility of the spectral properties of BPB that result from -chymotrypsin
association. When -chymotrypsin is in the presence of surfactant under visible light
the absorbance of BPB is at its maximum, as expected the BPB absorbance is higher
140
at pH 7 than at pH 3. With UV light exposure in the presence of azoTAB, BPB
experiences a less polar environment as a result of the protein association process,
while illumination with visible light results in protein dissociation and a decrease in
BPB absorbance. Control experiments (not shown) demonstrated no change in the
absorbance properties of BPB/ -Ch with light exposure without the presence of
azoTAB.
For comparison, the dynamics of absorption changes for a sample containing
2 mM azoTAB are also shown in Figure 4.9b. At elevated surfactant concentrations,
the penetration depth of the 350-nm and 434-nm light from the mercury arc lamp
used for photoisomerization decreases considerably due to the relatively high
absorption of azoTAB.(20, 23) As seen in Figure 4.8b, the change in the BPB peak is
slower at 5 mM compared to 2 mM azoTAB, and in each case the time required for
surfactant photoisomerization was measured to be equal to the dynamics of the BPB
absorption changes. However from the SAXS data it appears that protein
dissociation takes much longer than shown with BPB, this is because BPB is probing
protein association on the microscopic level. BPB-protein interactions can be formed
and disrupted rapidly with light, while the actual association/dissociation takes much
longer. Thus SAXS which probes the system on a macroscopic level is believed to be
more closely linked to actual dissociation time. From the optical density values away
from the absorbance peak at wavelengths greater than 700 nm, a measurable degree
of light scattering is observed at pH 7 that is absent at pH 3. This signifies that, in
addition to the monomer-oligomer association processes, large-scale aggregation is
141
also observed at pH 7, resulting in amorphous protein precipitates with sizes
approaching the wavelength of light (on the order of 100 nm) With UV-light
illumination, aggregation is enhanced as evidenced by an increase in the optical
density at high wavelengths, while this process can be completely reversed with re-
exposure to visible light (see Figure 4.9a insert). Thus, just as the SAXS data
demonstrated photoreversible control of protein oligomerization, these light-
scattering measurements further demonstrate photoreversible control of protein
polymerization. In many case oligomers similar in size to those observed in this
study have been shown to further assemble into larger aggregates and eventually
fibrils.(61, 62) As to date few techniques are available to breakup high molecular
weight aggregates such as A fibrils, which are generally considered to form
irreversibly.
4.4.6 Spherulite Formation
The amyloid fibrils formed by -chymortypsin eventually form amyloid
spherulites. Amyloid fibrils are often found arranged into large ordered spheroid
structures, known as spherulites. Spherulites are predominantly composed of
radially ordered amyloid fibrils and can be seen under cross polarized light,(51) each
spherulite shows a distinctive Maltese-cross extinction pattern, as seen in Figure
4.10. Images are from the original SANS samples taken one year later.
142
Figure 4.10: -chymotrypsin azoTAB solutions exhibiting maltese-crosses under cross
polarized light. [ Ch] = 10mg/mL, [azoTAB] = 1 mM
4.5 Acknowledgements
We acknowledge the support of the National Institute of Standards and
Technology, U.S. Department of Commerce, in providing the neutron research
facilities used in this work. Special thanks are owed to Paul Butler and Boualem
Hammouda for helpful discussions. Use of the National Synchrotron Light Source,
Brookhaven National Laboratory, was supported by the U.S. Department of Energy,
Office of Science, Office of Basic Energy Sciences, under Contract DE-AC02-
98CH10886. We thank W. T. Heller for graciously supplying the GA_STRUCT
program.
4.6 References
(1) Chiti, F., and Dobson, C. M. (2006) Protein misfolding, functional amyloid, and
human disease, Annu. Rev. Biochem. 75, 333-366.
(2) Sambashivan, S., Liu, Y., Sawaya, M. R., Gingery, M., and Eisenberg, D. (2005)
Amyloid-like fibrils of ribonuclease A with three-dimensional domain-swapped and
native-like structure, Nature 437, 266-269.
143
(3) Booth, D. R., Sunde, M., Bellotti, V., Robinson, C. V., Hutchinson, W. L., Fraser, P.
E., Hawkins, P. N., Dobson, C. M., Radford, S. E., Blake, C. C. F., and Pepys, M. B.
(1997) Instability, unfolding and aggregation of human lysozyme variants
underlying amyloid fibrillogenesis, Nature 385, 787-793.
(4) Brange, J., Andersen, L., Laursen, E. D., Meyn, G., and Rasmussen, E. (1997)
Toward understanding insulin fibrillation, J. Pharm. Sci. 86, 517-525.
(5) Pallares, I., Vendrell, J., Aviles, F. X., and Ventura, S. (2004) Amyloid fibril
formation by a partially structured intermediate state of -chymotrypsin, J. Mol.
Biol. 342, 321-331.
(6) Klein, W. L., Stine, W. B., and Teplow, D. B. (2004) Small assemblies of
unmodified amyloid -protein are the proximate neurotoxin in Alzheimer’s disease,
Neurobiol. Aging 25, 569-580.
(7) Kirkitadze, M. D., Bitan, G., and Teplow, D. B. (2002) Paradigm shifts in
Alzheimer’s disease and other neurodegenerative disorders: the emerging role of
oligomeric assemblies, J. Neurosci. Res. 69, 567-577.
(8) Yong, W., Lomakin, A., Kirkitadze, M. D., Teplow, D. B., Chen, S.-H., and
Benedek, G. B. (2002) Structure determination of micelle-like intermediates in
amyloid -protein fibril assembly by using small angle neutron scattering, P. Nat.
Acad. Sci. 99, 150-154.
(9) Wang, S., and Lee, C. T., Jr. (2006) Protein secondary structure controlled with light
and photoresponsive surfactants, J. Phys. Chem. B 110, 16117-16123.
(10) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2005) Photocontrol of protein
folding: the interaction of photosensitive surfactants with bovine serum albumin,
Biochemistry 44, 524-536.
(11) Hamill, A., Wang, S.-C., and Lee, C. T., Jr. (2005) Probing lysozyme conformation
with light reveals a new folding intermediate, Biochemistry 44, 15139-15149.
(12) Timchenko, A. A., Ptitsyn, O. B., Dolgikh, D. A., and Fedorov, B. A. (1978) The
structure of ribonuclease in solution does not differ from its crystalline structure,
FEBS Lett. 88, 105-108.
(13) Chen, S. H., and Teixeira, J. (1986) Structure and fractal dimension of protein-
detergent complexes, Phys. Rev. Lett. 57, 2583-2586.
(14) Svergun, D. I., Petoukhov, M. V., and Koch, M. H. J. (2001) Determination of
domain structure of proteins from X-ray solution scattering, Biophys. J. 80, 2946-
2953.
144
(15) Heller, W. T., Krueger, J. K., and Trewhella, J. (2003) Further insights into
calmodulin-myosin light chain kinase interaction from solution scattering and shape
restoration, Biochemistry 42, 10579-10588.
(16) Marone, P. A., Thiyagarajan, P., Wagner, A. M., and Tiede, D. M. (1998) The
association state of a detergent-solubilized membrane protein measured during
crystal nucleation and growth by small-angle neutron scattering, J. Cryst. Growth
191, 811-819.
(17) Hunt, J. F., McCrea, P. D., Zaccai, G., and Engelman, D. M. (1997) Assessment of
the aggregation state of integral membrane proteins in reconstituted phospholipid
vesicles using small angle neutron scattering, J. Mol. Biol. 273, 1004-1019.
(18) Chiti, F., Taddei, N., Bucciantini, M., White, P., Ramponi, G., and Dobson, C. M.
(2000) Mutational analysis of the propensity for amyloid formation by a globular
protein, EMBO J. 19, 1441-1449.
(19) Srisailam, S., Thallampuranam, K., Krishnaswamy, S., Rajalingam, D., Kathir
Karuppanan, M., Sheu, H.-S., Jan, F.-J., Chao, P.-C., and Yu, C. (2003) Amyloid-
like fibril formation in an all beta barrel protein. Partially structured intermediate
state(s) is a precursor for fibril formation, J. Biol. Chem. 278, 17701-17709.
(20) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2004) Photoreversible viscosity
changes and gelation in mixtures of hydrophobically-modified polyelectrolytes and
photosensitive surfactants, Macromol. 37, 5397-5405.
(21) Hayashita, T., Kurosawas, T., Miyata, T., Tanaka, K., and Igawa, M. (1994) Effect
of structural variation within cationic azosurfactant upon photoresponsive function
in aqueous solution, Colloid Polym. Sci. 272, 1611-1619.
(22) Shang, T., Smith, K. A., and Hatton, T. A. (2003) Photoresponsive surfactants
exhibiting unusually large, reversible surface tension changes under varying
illumination conditions, Langmuir 19,10764-10773.
(23) Le Ny, A.-L. M., and Lee, C. T., Jr. (2006) Photoreversible DNA condensation
using light-responsive surfactants, J. Am. Chem. Soc. 128, 6400-6408.
(24) Glinka, C. J., Barker, J. G., Hammouda, B., Krueger, S., Moyer, J. J., and Orts, W. J.
(1998) The 30 m small-angle neutron scattering instruments at the National Institute
of Standards and Technology, J. Appl. Crystallogr. 31, 430-445.
(25) Guinier, A., and F., G. (1955) Small Angle Scattering of X-ray,Wiley, New York.
(26) Svergun, D. I. (1992) Determination of the regularization parameter in indirect-
transform methods using perceptual criteria, J. Appl. Crystallogr. 25, 495-503.
145
(27) Svergun, D. I., and Koch, M. H. J. (2003) Small-angle scattering studies of
biological macromolecules in solution, Rep. Prog. Phys.66, 1735-1782.
(28) Yang, L. (2005) The X21 SAXS instrument at NSLS for studying macromolecular
systems, Macromol. Res. 13, 538-541.
(29) Chen, S.-H., and Lin, T.-L. (1987) in Neutron Scattering (Price, D. L., and Skold,
K., Eds.) pp 489-543, Academic Press, San Diego.
(30) Pedersen, J. S., Hansen, S., and Bauer, R. (1994) The aggregation behavior of zinc-
free insulin studied by small-angle neutron scattering, Eur. Biophys. J. 22, 379-389.
(31) Shilton, B. H., McDowell, J. H., Smith, W. C., and Hargrave, P. A. (2002) The
solution structure and activation of visual arrestin studied by small-angle X-ray
scattering, Eur. J. Biochem. 269, 3801-3809.
(32) Ramakrishna, T., and Pandit, M. W. (1992) Involvement of active site in self-
association of proteins: a case of alpha-chymotrypsin, Indian J. Biochem. Biophys.
29, 192-200.
(33) Pandit, M. W., and Rao, M. S. N. (1975) Self-association of proteins. Self-
association of R-chymotrypsin at its isoelectric point in buffer solutions of ionic
strength 0.1, Biochemistry 14, 4106-4110.
(34) Ackers, G. K., and Thompson, T. E. (1965) Determination of stoichiometry and
equilibrium constants for reversibly associating systems by molecular sieve
chromatography, P. Nat. Acad. Sci. 53, 342-349.
(35) Svergun, D. I., Richard, S., Koch, M. H. J., Sayers, Z., Kuprin, S., and Zaccai, G.
(1998) Protein hydration in solution: experimental observation by X-ray and neutron
scattering, P. Nat. Acad. Sci. 95, 2267-2272.
(36) Durchschlag, H., and Zipper, P. (1996) Comparative determination of structural
parameters and conformational changes of proteins by small-angle scattering,
crystallography and hydrodynamic analysis, J. Mol. Struct. 383, 223-229.
(37) Ninio, J., Luzzati, V., and Yaniv, M. (1972) Comparative smallangle X-ray
scattering studies on unacylated, acylated, and crosslinked Escherichia coli transfer
RNAIVa1, J. Mol. Biol. 71, 217-229.
(38) Berret, J.-F., and Oberdisse, J. (2004) Electrostatic self-assembly in polyelectrolyte-
neutral block copolymers and oppositely charged surfactant solutions, Phys. B:
Condensed Matter (Amsterdam)350, 204-206.
(39) Perkins, S. J., Smith, K. F., Kilpatrick, J. M., Volanakis, J. E., and Sim, R. B. (1993)
Modeling of the serine-proteinase fold by X-ray and neutron scattering and
sedimentation analyses: occurrence of the fold in factor D of the complement
system, Biochem. J. 295 (Part 1), 87-99.
146
(40) Fujisawa, T., Ueki, T., Inoko, Y., and Kataoka, M. (1987) X-ray scattering from a
troponin C solution and its interpretation with a dumbbell-shaped-molecule model,
J. Appl. Crystallogr. 20, 349-355.
(41) Tranier, S., Mortier-Barriere, I., Ilbert, M., Birck, C., Iobbi-Nivol, C., Mejean, V.,
and Samama, J.-P. (2002) Characterization and multiple molecular forms of TorD
from Shewanella massilia, the putative chaperone of the molybdoenzyme TorA,
Protein Sci. 11, 2148-2157.
(42) Chiti, F., Webster, P., Taddei, N., Clark, A., Stefani, M., Ramponi, G., and Dobson,
C. M. (1999) Designing conditions for in vitro formation of amyloid protofilaments
and fibrils, P. Nat. Acad. Sci. 96, 3590-3594.
(43) Vanhoudt, J., Abgar, S., Aerts, T., and Clauwaert, J. (2000) A small-angle X-ray
solution scattering study of bovine R-crystallin, Eur. J. Biochem. 267, 3848-3858.
(44) Jaikaran, E. T. A. S., Higham, C. E., Serpell, L. C., Zurdo, J., Gross, M., Clark, A.,
and Fraser, P. E. (2001) Identification of a novel human islet amyloid polypeptide -
sheet domain and factors influencing fibrillogenesis, J. Mol. Biol. 308, 515-525.
(45) Zurdo, J., Guijarro, J. I., Jimenez, J. L., Saibil, H. R., and Dobson, C. M. (2001)
Dependence on solution conditions of aggregation and amyloid formation by an SH3
domain, J. Mol. Biol. 311, 325-340.
(46) Susi, H., and Byler, D. M. (1986) Resolution-enhanced Fourier transform infrared
spectroscopy of enzymes, Methods Enzymol. 130, 290-311.
(47) Byler, D. M., and Susi, H. (1986) Examination of the secondary structure of proteins
by deconvolved FTIR spectra, Biopolymers 25, 469-487.
(48) Klunk, W. E., Jacob, R. F., and Mason, R. P. (1999) Quantifying amyloid by Congo
red spectral shift assay, Methods Enzymol. 309, 285-305.
(49) Glenner, G. G. (1980) Amyloid deposits and amyloidosis. I. The -fibrilloses, N.
Engl. J. Med. 302, 1283-1292.
(50) Puchtler, H., and Sweat, F. (1965) Congo red as a stain for fluorescence microscopy
of amyloid, J. Histochem. Cytochem. 13, 693-694.
(51) Krebs, M. R. H., Bromley, E. H. C., Rogers, S. S., and Donald, A. M. (2005) The
mechanism of amyloid spherulite formation by bovine insulin, Biophys. J. 88, 2013-
2021.
(52) Wang, D., Bode, W., and Huber, R. (1985) Bovine chymotrypsinogen A. X-ray
crystal structure analysis and refinement of a new crystal form at 1.8 Å resolution, J.
Mol. Biol. 185, 595-624.
147
(53) Rao, M. S. N., and Kegeles, G. (1958) An ultracentrifuge study of the
polymerization of R-chymotrypsin, J. Am. Chem. Soc. 80, 5724-5729.
(54) Egan, R., Michel, H. O., Schlueter, R., and Jandorf, B. J. (1957) Physicochemical
investigation of the chymotrypsins. II. On the mechanism of dimerization of
chymotrypsin, Arch. Biochem. Biophys. 66, 366-373.
(55) Ding, F., and Dokholyan, N. V. (2005) Simple but predictive protein models, Trends
Biotechnol. 23, 450-455.
(56) Dokholyan, N. V. (2006) Studies of folding and misfolding using simplified models,
Curr. Opin. Struct. Biol. 16, 79-85.
(57) Bertsch, M., Mayburd, A. L., and Kassner, R. J. (2003) The identification of
hydrophobic sites on the surface of proteins using absorption difference
spectroscopy of bromophenol blue, Anal. Biochem. 313, 187-195.
(58) Kitano, H., Maeda, Y., and Okubo, T. (1989) Kinetic study of the effects of
solvation on the dimerization process of a-chymotrypsin, Biophys. Chem. 33, 47-54.
(59) Koren, R., and Hammes, G. G. (1976) A kinetic study of protein-protein
interactions, Biochemistry 15, 1165-71.
(60) Pal, S. K., Peon, J., and Zewail, A. H. (2002) Ultrafast surface hydration dynamics
and expression of protein functionality: a-Chymotrypsin, P. Nat. Acad. Sci 99,
15297-15302.
(61) Tito, P., Nettleton, E. J., and Robinson, C. V. (2000) Dissecting the Hydrogen
Exchange Properties of Insulin under Amyloid Fibril Forming Conditions: A Site-
specific Investigation by Mass Spectrometry, J. Mol. Biol. 303, 267-278.
(62) Nettleton, E. J., Tito, P., Sunde, M., Bouchard, M., Dobson, C. M., and Robinson, C.
V. (2000) Characterization of the oligomeric states of insulin in self-assembly and
amyloid fibril formation by mass spectrometry, Biophys. J. 79, 1053-1065.
148
Chapter V: -amyloid Peptide (1-40) Fibril Growth Influenced by
Photosurfactant
5.1 Abstract
The effect of an azobenzene-based photoresponsive surfactant on the fibril
formation of -amyloid (1-40) (A 40) has been studied using small-angle neutron
scattering (SANS), atomic force microscopy (AFM), and light scattering (LS)
measurements. Fibril formation is inhibited with a lag phase persisting for
approximately five hours in the presence of the trans isomer of the photosurfactant
under visible light (i.e., the relatively hydrophobic, activated form). Conversely,
only a two-hour lag phase is observed under UV light with the cis photosurfactant
isomer (relatively hydrophilic, passive form), while large fibril networks are
immediately observed for the pure protein. Furthermore, in situ UV illumination of
a solution of trans surfactant and protein results in rapid fibril formation. Thus, the
ability to photoreversibly inhibit and trigger the fibrilization process with light
illumination is demonstrated. Shape-reconstruction analysis of the SANS data is
used to obtain novel information on the conformation of the protein during these
initial stages of protein aggregation. Small, cylindrical protein aggregates 48 Å in
diameter and 73 Å long are initially observed during the lag phase independent of the
sample conditions. AFM images confirm both the aggregate structure and duration
of the lag phase, and further suggest that these early aggregates appear to be the
seeds for longer aggregates that develop over time.
149
5.2 Introduction
When it comes to protein-protein interactions, there has been a great deal of
focus on amyloid formation owing to the relation to neurodegenerative diseases.
Each disease is linked to a specific protein; -Synuclein to Parkinson’s Disease
(PD), Huntingtin with poylQ expansion to Huntington’s Disease (HD), and Amyloid
peptide (A ) to Alzheimer’s Disease (AD).(1) There are no clear similarities, in
structure or sequence, between any of the proteins that form amyloids, despite this
fact all amyloid fibrils display similar features. The sharing of this phenomenon has
engaged scientist in one of the most perplexing and challenging problems in recent
medical research, namely solving the mechanisms behind such fibril formation.
It is generally believed that proteins need to unfold, at least partially, to
aggregate into amyloid fibrils(2-4) via a nucleated growth mechanism.(1) Typically
there is a lag phase where these partially folded states associate into nuclei (small
unstructured oligomers of 2-4 molecules in the case of A 40, and 5-6 for A 42)(5).
The small oligomers associate further into protofibrils, aggregates that are not yet
fibrillar in their morphologies but have properties characteristic of amyloids (notably
sheet structure).(1) Protofibrils then associate into protofilaments, elongated
aggregates ~2-5 nm in diameter. Eventually the species aggregate into long
insoluble amyloid fibrils that consist of 2-6 intertwined protofibrils.(6) These fibrils
then accumulate forming microscopic deposits (plaques) in the brain.
Specifically in the case of the -amyloid peptides (A ), implicated in AD,
interest has spiked in the small oligomers as they have been detected in the brains of
150
AD patients.(7) Recent focus on what is believed to be the primary pathogenic
species has shifted to these early oligomers as the severity of cognitive impairment in
AD has been correlated with the level of prefibrillar species of A (8-10). A , a
proteolytic fragment of the amyloid- precursor protein (A PP), is a small 39-43
amino acid peptide(11) with most abundant forms of A 40 and A 42. The
premature development of Alzheimer’s disease has been linked to enhanced
protofibril formation.(12) Thus the development of strategies for treating AD could
be greatly improved if the early stages of the A 40 assembly process were better
understood.
A technique such as SANS that is capable of studying such a mechanism
requires the ability to investigate the process from nanometer sized early oligomers
through protofibrils and protofilaments to micron sized fibrils. SANS has
contributed to a better understanding of the mechanism of different types of protein
aggregation in solution.(13) Specifically in the case of A 40, SANS was used in the
structural determination of small oligomers found to be cylindrical with radii and
length of 24 and 110 Å, respectively.(14)
In a recent study, it was shown that light illumination could be used to induce
photoreversible changes in the degree of association in the initial stages of an
amyloid forming protein.(15) azoTAB is a photoresponsive surfactant having an
active (visible light) and inactive (UV light) form, therefore, light can be used to
reversibly control protein association by replacing protein-protein interactions with
protein-surfactant interactions. A shape reconstruction technique, GA_STRUCT(16)
151
was applied to SANS data to determine the structure of the pre-amyloid oligomers of
-chymotrypsin.(17) Under visible light the activated trans form of the
photosurfactant caused partial unfolding of the protein, which led to the formation of
small six unit oligomers (hexamers). Shape-reconstruction showed these hexamers
to have a corkscrew like structure. Under UV light the photoisomerization of the
surfactant to the passivated cis form caused the hexamers to laterally associate in a
slightly offset manner, intertwining two at a time to form dodecamers. The
dodecamers have an elongated twisted conformation with cross-sectional radii
similar to amyloid protofilaments. FT-IR was also used to show that this hexamer-
to-dodecamer transition occurs through intermolecular sheets stabilized with
hydrogen bonds.
The present study sets out to investigate the effect of such a surfactant on the
A 40 peptide. SANS, AFM and light scattering are used to fully investigate the
A 40 fibril growth process over time as a function of surfactant and light conditions.
The ability to inhibit/delay, and then trigger, the growth process of amyloid fibrils
with light illumination is demonstrated. The presence of azoTAB inhibits initial
fibril formation. Visible light causes a more severe delay in the fibril growth process
than UV light, thus, UV light can be used to accelerate fibrilization. Shape
reconstruction applied to the SANS data from samples containing azoTAB allowed
for the structural determination of small A 40 oligomers, which are consistent with
the literature.(14, 15, 18) Under visible light these oligomers are the only significant
scatterers at an age of 0, 2, and 3 hrs, while at an age of 5 hrs the scattering is
152
consistent with fibrils. Under UV light the 0 hr sample still shows only scattering
from the small oligomers while the 2 and 3 hr samples have scattering consistent
with a mixture of both small oligomers and fibrils. The 5 hr sample under UV light
has scattering consistent with that of a 3D network of fibrils. This illustrates that UV
light illumination directly results in a high degree of association. Light scattering
was used to demonstrate the ability to accelerate fibril growth by simple illumination
with UV light. Light illumination, UV or visible, induces a change in surfactant
form, trans or cis, respectively. The trans form under visible light in turn stimulates
protein-surfactant interactions inhibiting protein-protein interactions leading to
delayed fibril formation. While the cis form under UV light inhibits protein-
surfactant interactions allowing for protein-protein interactions to take over and thus
trigger the association process.
5.3 Experimental Methods
5.3.1 Materials
An azobenzene-trimethlyammonium bromide (azoTAB) surfactant of the
form
N
N
CH
3
CH
2
O(CH
2
)
4
N+(CH
3
)
3
Br-
was synthesized and used as in previous studies.(15, 19-22) The surfactant
undergoes a photoisomerization from a trans form to a predominately cis form
153
(90%) when exposed to 365 nm UV light. This isomerization can then be reversed
(75% trans) upon exposure to 436 nm visible light or by sitting in the dark for
~24hrs. Photoconversion was achieved with a 365 nm line from a 200 W mercury
arc lamp (Oriel, model no. 6283), isolated with the combination of a 320 nm band-
pass filter (Oriel model no. 59800), for UV conversion, or a 400 nm long pass filter
(Oriel model no. 59472), for visible conversion, and an IR filter (Oriel model no.
59060). Exposure of the peptide to UV light was avoided by exposure of surfactant
solutions to UV light prior to the addition of solid peptide.
-amyloid peptide (1-40), Ultra Pure, HCl,(23) expressed in E. coli (rPeptide
cat. no. A-1156) was used as received. All other chemicals were obtained from
Aldrich in the highest purity. The A 40 was resuspended in a 1% NH4OH solution.
A 40 was selected because it is less fibrillogenic than A 42 allowing focus on the
early oligomer species as opposed to the full fibrils.
5.3.2 Small Angle Neutron Scattering
SANS experiments were performed on the 30m NG3 and NG7 SANS
instruments at NIST.(24) Measurements were performed at 4°C in order to avoid
any association/growth during data collection.(14, 25, 26) A neutron wavelength of
= 6 Å and a detector offset of 25 cm with two sample to detector distances of 1.33
and 7 meters were utilized to obtain a Q range of 0.0048 - 0.47 Å
-1
. The net
intensities were corrected for the background and empty cell (pure D
2
O), accounting
for the detector efficiency using the scattering from an isotropic scatterer (Plexiglas),
154
and converted to an absolute differential cross section per unit sample volume (in
units of cm
-1
) using an attenuated empty beam. A constant background was
subtracted (an average of the scattering form the 15 highest Q values) in order to
correct for incoherent scattering. Shape reconstruction was done similar to previous
studies(15, 19, 21) using the GA_STRUCT(16) algorithm. The program fits the
experimental data starting with an initial guess of randomly oriented scattering
centers and rearranges them until the best fit is achieved. A peptide concentration of
0.75 mg/mL was used to obtain significant scattering, when employed a surfactant
concentration of 8 mM was used.
Reduction and analysis of the SANS data were done using Igor Pro.(27) A
Lorentzian Model, BGD QL I Q I + + = ] ) ( 1 /[ ) (
2
0
, where L is the screening length, I
0
is the scattering intensity extrapolated to zero angle and BGD is the background
scattering intensity, was used to fit the scattering curves from the pure protein
solutions. Each model was smeared to account for instrumental resolution. A
cylinder model with circular cross section and uniform scattering length density was
used to fit the data from the 8mM azoTAB solutions at time t = 0, 2, and 3 hr under
visible and the t = 0 under UV light.
The Guinier approximation(28) was used to calculate the radii of gyration
() ()() 3 / exp 0
2 2
g
R Q I Q I = ,
where I(Q) is the scattering intensity, I(0) is the extrapolated intensity at Q = 0 and
R
g
is radius of gyration, valid for globular species. When the protein was in a rod-
like formation modified Guinier analysis for rod-like forms was used(29)
155
Q I(Q) =I
C
(0)exp( Q
2
R
C
2
/2)
where R
C
is the cross-sectional radius of gyration, the cross-sectional radius is related
to the geometric radius by R
C
2
= R
2
/ 2.
A model independent approach was taken to obtain information about the
protein conformation through the calculation of a PDDF, the measure of the
probability P(r) of finding two scattering centers at a distance r apart.(30, 31)
=
0
) sin(
) ( 4 ) ( dr
Qr
Qr
r p Q I
GNOM(32) was used to calculated the PDDFs, a Q range of 0.013-0.29 Å
-1
was
used. The maximum particle diameter, D
max
, was selected to be the lowest value that
gave a smooth return to zero.
5.3.3 Atomic Force Microscopy
Tapping mode AFM in air at 25°C was performed using an Autoprobe CP
atomic force microscope (Park Scientific Instruments). The microscope was
equipped with a 100 μM scanner with a maximum xy scan range of 100 μM and a
maximum z range of 6 μM. The images were acquired by using rectangular silicon
cantilevers 135 μM long with resonant frequencies in the 200-400 kHz range and
nominal spring constants in the 20-80 N/m range. The cantilevers have integrated
tips with a radius of curvature below 10nm (model no. RTESPA,Veeco Probes).
Samples were imaged at scan sizes between and 0.2 and 6 μM using line scan rates
below 2Hz. All of the images were processed by plane fitting and then three-
156
dimensional shade rendering. All peptide solutions were diluted to 0.01 mg/mL using
a 1% NH
4
OH solution. A10 μL aliquot of solution was then deposited onto freshly
cleaved mica. The samples were then wicked dry using filter paper and imaging
occurred within 2 hrs of sample preparation.
5.3.4 Light Scattering
Light scattering measurements were performed at 25°C on a Brookhaven
model BI-200SM instrument (Brookhaven Instrument Corp.) with a 35mW (Melles
Griot, model number 05-LHP-928) helium neon ( = 632.8 nm) laser at a scattering
angle of 90°. The wavelength of the laser is far enough away from the 436 nm
absorbance peak of azoTAB to avoid any conversion back to trans (visible light)
state while in the cis (UV-light) state. A BI-9000AT digital correlator (Brookhaven
Instrument Corp.) was used. Prior to measurements solutions were filter through a
200 nm Anatop filter. After the initial filtration, if possible, the solution was filter
further through a 20 nm Anatop filter (only done for the measurements under visible
light). A peptide concentration of 0.25 mg/mL was used along with a surfactant
concentration of 8 mM.
5.4 Results and Discussion
A 40 fibril growth was studied in the presence of 8 mM azoTAB under both
visible and UV light using SANS. Figure 5.1 shows the SANS data of fibril growth
over time. Both time and azoTAB significantly effect scattering. Scattering at
157
different Q values indicates structures of different length scales (Q = 2 /L), where L
is length. Pure peptide solutions are seen to form large structures (scattering shifted
to lower Q values), while the presence of trans azoTAB under visible light results in
the formation of small structures. The cis isomer under UV light appears to trigger
the growth of these small structures.
Figure 5.1: SANS data of A in the presence and absence of 8 mM azoTAB as a function of
time and light illumination. [A ] = 0.75 mg/mL.
The scattering from the pure A 40 samples exhibits a Q
-2
slope at low Q (see
Figure 5.1), indicating fibril-fibril interactions that lead to the formation of a loose 3-
D network, akin to the semi-dilute phases of flexible polymers solutions.(33, 34)
SANS has been used previously to study the association of lysozyme into such 3D
networks.(35) The study suggests that lysozyme denatures in the presence of
dithiothreitol (DTT), which leads to the formation of -sheet-rich fibrils. These
158
fibrils aggregated into even larger fibers that physically cross-link forming 3D
networks. Thus pure A 40 appears to form these fibril networks immediately since
all of the spectra for the pure peptide, aged for various amount of time prior to
scattering measurements (t = 0, t = 5, t = 10, and t = 15), lie directly on top of one
another. The Q
-2
behavior is due to the Lorentzian structure factor at low Q; hence a
Lorentzian Model was used to fit the pure A scattering curves. The resulting
screening length at each condition, which is on the order of the mesh size of the fibril
network(36), is shown in Table 5.1. The screening length is seen to decrease over
time, implying that the fibril network is becoming more densely packed.
Table 5.1: Parameters from the model dependent fitting of A SANS data.
In contrast, the spectra from the early-time samples containing azoTAB (0 - 3
hrs under visible light and 0 hrs under UV light) exhibit a dramatic shift of the
scattering to higher Q values relative to the pure A 40 samples, indicating a much
smaller structure on the order of small protein oligomers. Thus, it appears that fibril
formation is inhibited by azoTAB. Similar results were shown with A 40 in the
159
presence of the surfactant SDS, which was shown to suppress fibril formation
through peptide/SDS interactions in aqueous solution.(37) Amyloid fibrillation,
specifically that of A , is believed to be a nucleation dependent association process.
First, monomeric A forms small oligomers that then associate into protofilaments,
which then grow into full fibrils.(38) The shape of the scattering curves from these
early-time samples containing azoTAB are characteristic of small globular
structures. These structures are believed to be the early oligomers, which begin the
fibril growth process. Scattering data from these samples could be fit using a
cylinder model with an average radius and length of 24.0 Å and 72.5 Å, respectively,
as seen in Table 5.1. SANS has been used previously to propose a schematic model
of an early A oligomer (pure A , no surfactant or other additive) suggesting a
cylinder with a radius of 24 Å and a length of 110 Å.(14)
After 5 hours the scattering from the 8 mM trans azoTAB solution under
visible light has shifted towards lower Q and exhibits a Q
-4
slope at low Q, indicating
a boundary scattering from large fibrils(37) (confirmed with AFM, see below). The
5 hr-old sample with 8 mM of azoTAB in the trans state was then exposed to UV
light for 1 hr to convert azoTAB form the trans to the cis form. This
photoisomerization was done with the sample in an ice bath to suspend the normal
time course of fibril growth (14, 25, 26) and isolate the effect of the surfactant
isomeric form of fibril formation The scattering obtain from this sample exhibits
both Q
-2
and Q
-4
behavior, suggesting that an intermediate state between individual
large fibrils and a 3D fibril network exists. This indicates that conversion of azoTAB
160
to the cis form with exposure to UV light triggers fibril growth with the resulting
fibrils likely in the early stages of 3D network formation.
Samples containing cis azoTAB incubated for 2 or 3 hrs at 25 °C under UV
illumination give spectra that exhibit a slope of Q
-4
at low Q (similar to that expected
from individual large fibrils), while a shoulder is observed at higher Q values
indicating a significant amount of small, oligomeric, species remaining at these
conditions. Thus, these samples have a combination of both small oligomers and
fibrils, this will be discussed in detail below.
Guinier analysis was used to obtain the radii of gyration of the small,
oligomeric species (Figure 5.2a), while modified Guinier analysis was used to
estimate the cross-sectional radii of gyration of the isolated or networked fibrillar
samples (Figure 5.2b). As seen in Table 5.1, the values estimated from the Guinier
analysis give an average R
g
of 25.3 ± 0.2 Å for the small oligomers independent of
time over the 3 hour time observed. These R
g
values are in generally good
agreement with values that can be estimated from the cylindrical fits through the
equation 8 12
2 2 2
d L R
g
+ = , where L and d are the length and diameter of the
cylinder, respectively. Similar sized and shaped oligomers have been observed for
pure -amyloid (A 40) with SANS(14) (r = 24 Å, L = 110 Å as well as with
TEM(18) (r = 30 Å, L = 200 Å). SANS was also used to obtained the detailed
structure of the early oligomers of -chymotrypsin amyloids in the presence of
azoTAB (r = 30 Å, L = 120 Å).(15) The amyloid state is believed to be a very stable
conformation that any protein can form under the appropriate conditions.(39)
161
Appropriate conditions typically being those that cause the protein to partially
unfold, but only slightly.
Figure 5.2: Analysis of the SANS data. (a) Guinier analysis, (b) modified Guinier analysis,
and (c) pair distance distribution functions. Pure A : t = 0 hr (o), t = 5 hr (o), t = 10 hr (o), t
= 15 hr (o); A with trans azoTAB under visible light: t = 0 hr ( ), t = 2 hr ( ), t = 3 hr
( ), t = 5 hr ( ); A with cis azoTAB under UV light: t = 0 hr ( ); A with azoTAB
initially in the trans form following conversion to the cis form with UV exposure ( ).
Modified Guinier analysis, shown in Figure 5.2b, of the sample incubated
with trans azoTAB under visible light for 5 hrs gave an estimated cross-sectional
radius, R
c
, of the individual fibrils of 39.1 Å. This value can be related to the actual
fibril radius, R, through the equation R
2
= 2 R
c
2
, giving R = 55.2 Å. Similar fibril
radii of R = 56.0 ± 0.6 Å can be obtained from the cross sectional radii of the fibers
in the fibril networks (pure -amyloid peptide solutions) (t = 0, t =5, t = 10 and t = 15
hr) as well as the sample incubated with cis azoTAB under UV light, as shown in
162
Table 5.1. The thinnest A 40 and A 42 fibrils that have been observed in electron
microscope images have diameters of 6 ± 1 nm.(40-43) The length of a -strand
formed by n residues is approximately n x 0.34 nm. Given a 10-residue disordered
N-terminal segment, a minimum diameter of 10 nm would be expected for A 40 and
A 42 fibrils if the remainder of the peptide formed a single, continuous -strand.(40)
Although fibrils as small as 6 nm in diameter have been observed it is typically
believed that amyloid fibrils have a diameter greater than or equal to 10 nm.
The dimensions obtained from the analysis of the SANS data give insight
into the actual mechanism of fibrillization. Based on the size of the cylindrical
oligomers (~4.5 nm in diameter and ~7.0 nm in length) growth into fibrils must
occur both radially and longitudinally in order to obtain fibrils of such thickness (~10
nm diameter) and length. Thus, small oligomers cannot simply aggregate end-to-
end; instead, at some point lateral aggregation must occur. The AFM images (to be
discussed below) give further insight into the mechanism of the aggregation process.
PDDFs (Figure 5.2c) were calculated from the scattering of the oligomeric
species (8 mM trans azoTAB at t = 0, 2, and 3 hr, 8 mM cis azoTAB at t = 0 hr).
The most common dimension (e.g., in a sphere the radius), r = 29.6 ± 1.5 Å and the
maximum dimension, D
max
= 70.4 ± 3.4 Å agree very well with the dimensions from
both the cylinder fits (radius ~ 23 nm and length ~70 nm) and the Guinier fits in
Table 5.1 (radius ~ 25 nm). The dimensions of the seeds change very little with
condition, indicating that the oligomeric species are very consistent and stable. This
supports the idea of a nucleation and growth process with a lag phase.
163
The results above clearly demonstrate the ability of azoTAB to inhibit the
fibrillization process of A 40 and the ability of UV light to trigger this process
through conversion of azoTAB to the relatively hydrophilic and inactivated cis form.
Although Guinier analysis and the calculation of pair distance distribution functions
provide a model of the early A oligomer (believed to be the key intermediate in the
nucleation and growth process), only low-resolution information is achieved from
these two techniques. The primary goal in the present study is to obtain high-
resolution structural information on these A oligomers, this can be done through the
shape reconstruction as done previously.(44-47) The shape reconstructions shown in
Figure 5.3, show the oligomers to be cylindrical and relatively globular, this agrees
with the parameters determined from Guinier analysis, PDDF calculations, and
cylinder fits of the SANS data. Fitting of the SANS data was done using two
different shapes: one disclike (not shown) and the other cylindrical, each resulted in
equally “good” fits. It is a common problem with model-dependent fitting techniques
to be unable to distinguish between two such globular states. However the models
calculated form shape reconstruction using GA_STRUCT(16) (see Figure 5.3) lean
almost exclusively toward a cylinder and not a disc (this can be seen from the worst
fit of the t = 0 hr sample under visible light shown in the first column). All of the
shape reconstructions shown in Figure 5.3 are similar, independent of condition (i.e.
time and/or light condition). Since each structure is the consensus envelope of a
family of 10 fits, it is conclusive that the oligomers are indeed cylindrical and not
discs. After the GA_STRUCT program has performed its genetic algorithm ten
164
times (producing 10 individual structures), it then characterizes the similarity of
these generated structures through the creation of a consensus envelope. The
program uses the individual scattering centers in each fit to determine the probability
that there is a scattering center at a certain location, it then creates a consensus
envelope by putting a scattering center at only the most probable locations,
essentially giving an average of all fits.
Figure 5.3: Shape reconstruction of the A oligomers with 8 mM trans azoTAB (t = 0, 2,
and 3 hr under visible light) and 8 mM cis azoTAB (t = 0 hr under UV light). AFM images
shown across the bottom (condition corresponds to column titles). All blue structures are
consensus envelopes, an example of a worst fit is shown in red at scale. The green
encircled regions represent structures demonstrating early stages of aggregation.
165
As discussed above the samples containing cis azoTAB incubated for 2 and 3
hrs under UV illumination give spectra indicative of samples containing a
combination of both small oligomers and fibrils. This is illustrated in Figure 5.4,
where a fraction of the full fibril scattering is subtracted from the overall scattering.
This is done by subtracting a scaled scattering curve form the fibril state (e.g., the t =
5 hr visible sample in Figure 5.1) from the original data such that the scattering
intensity becomes independent of Q in the range of Q = 0.01 – 0.05 Å, a feature of
small, globular structures. The difference results in the scattering form the
remaining species, which was found to be nearly identical to the scattering from the
small oligomers.
Figure 5.4: Scattering from small oligomers and fibrils combined to model scattering from
the (a) 2 hr and (b) 3 hr samples with azoTAB under UV light. Insets show shape
reconstruction of the small oligomer fractions.
Atomic force microscopy was performed on the SANS samples as well as on
fresh A 40 samples in order to further investigate these globular structures. The
globular species are shown in the bottom row of Figure 5.3, while Figure 5.5
illustrates the growth of the peptide from intermediates into full-grown amyloid
fibrils. Similar to the shape reconstruction models, the AFM images of the oligomers
166
(Figure 5.3a-d) are cylindrical. The oligomers appear to be similar in both shape and
size, independent of time and/or light condition. From the R
g
values calculated from
both the cylinder fits (R
g
= 25.9 ± 0.2 Å) as well as from Guinier analysis (R
g
= 26.1
± 0.2 Å) a maximum height (in the z-direction) equal to that of the diameter (~ 2R
g
) d
~ 2R = 5.2 nm would be expected, which is what is observed in these images. These
dimensions also agree well with the models from shape reconstruction. The size of
the scanning tip used in AFM imaging limits the spatial resolution that can be
achieved in the x and y directions, thus, fibril widths in the images shown appear
larger than their actual size.(48) Consequently, fibril heights (z-direction) in AFM
are a much more accurate structural feature than widths.
Both shape reconstruction and AFM (from the z-direction height) show the
oligomer to be small cylindrical species with a cross sectional diameter of ~5 nm.
These early oligomers have typically been reported as spherical particles 2.7-4.2 nm
in diameter.(49) Thus, it is possible that the species in Figure 5.3 are dimers of these
spherical “beads”. Furthermore when examining the larger aggregates such as in
Figure 5.5f its appears that many of these beads have been strung together to form
longer fibrils. The beads reported in literature have been described as protein
micelles, because A is an amphipathic surface-active peptide. Thus, oligomer
formation displays a critical concentration dependence, and their formation is
correlated with the appearance of a hydrophobic environment.(50-52) Thus, it is
possible that the surfactant is changing the hydrophobic environment experienced by
A causing a dimple and elongation in the bead (similar to the spherical to
167
cylindrical transition of surfactant micelles upon increasing the surfactant
concentration)(53). There is not enough evidence to conclusively state either case.
However, an important point to mention is the structural similarity azoTAB and
Congo red, an azobenzene-based dye know to bind to amyloid sheet structure.
Both molecules contain two benzene rings with an azo linkage. Congo red is known
to intercolate between the sheets of amyloids therefore it is likely that azoTAB is
interacting with such structures and it is even possibly influencing the structure.(54-
58)
Figure 5.5: AFM images of intermediates and full fibrils. A with 8 mM azoTAB (a) t = 5
hr with trans azoTAB, freshly made (b) t = 15 hrs with trans azoTAB, frozen SANS
sample(c) t = 5 with cis azoTAB, freshly made (d) t = 15 hrs with cis azoTAB, frozen from
SANS. (e) and (f) Pure A t = 15hrs, frozen from SANS.
In order to be confident that these small oligomers are indeed associated A
monomers and not just the effect of azoTAB aggregates on the protein in the form of
surfactant micelles, experiments were performed at the protein contrast-matching
168
point (60/40, H
2
O/ D
2
O), thus, rendering the protein “invisible” to neutrons. The
SANS spectrum obtained from this experiment is essentially zero within the
experimental error (not shown), indicating minimal to no scattering from the
azoTAB itself. This conclusively shows that these oligomers are a result of protein
aggregation and are not simply surfactant micelles or a combination of surfactant and
protein. It is important to note that these experiments were performed above the
critical micelle concentration of trans azoTAB (~ 5 mM)(59), however, it has been
shown that azoTAB micellization is significantly delayed in the presence of proteins
(>20 mM)(46). Furthermore, the critical micelle concentration of cis azoTAB is ~11
mM(59).
The images in the left column of Figure 5.5 (a and d) were taken on the
freshly made samples aged for 5 hrs. The species in the images are intermediates
anywhere from tens to hundreds of nanometers long and typically around 11 nm in
diameter (as measured from the height in the z-direction). The images in the center
column of Figure 5.5 (b and e) were taken on the 8mM azoTAB SANS samples (~
15hrs old) stored on dry ice for transport form NIST. These fibrils have a z-height of
~15 nm, and they range in length from several hundred nanometers to several
microns. The z-height agrees with the diameter of the cylinder determined from the
radii calculated from modified Guinier analysis, R = 57.1 ± 2.4 Å. The fact that the
AFM heights are slightly higher is most likely due to the fact that although amyloid
fibrils share the same characteristics no two fibrils are exactly identical (fibril
thickness typically varies). Figure 5.5f illustrates the axial periodicity (very clearly
169
seen in the 3D rendering) observed in some samples. Periodicity of this type has
been observed in the -amyloid peptide (Alzheimer’s disease),(60-62) human amylin
(type II diabetes),(63) the peptide hormone calcitonin (medullary carcinoma of the
thyroid),(64) and Sup35 (prion disease).(65) The image in Figure 5.5c illustrates the
twisting of fibrils as observed in some samples, a well known characteristic of
amyloid fibrils that has been reported previously with AFM(61, 63, 66, 67) and
TEM(68-72).
Based on the SANS data along with the AFM images, the following
aggregation mechanism involving two different pathways is proposed. Both
pathways start with the same initial step, the formation of very short cylindrical
oligomers (Figure 5.3a and b). In pathway one after the seeds have formed they
initially associate end-to-end (Figure 5.3c, green highlights), once they have
established some length they twist around one another to form thicker and longer
cylinders (Figure 5.5a), then these long cylinders continue to simultaneously grow in
length and twist with other cylinders (Figure 5.5b) to form full fibrils (Figure 5.5c).
In the second pathway after the seeds have formed they initially associate side-to-
side (Figure 5.3d) the fatter bead-like cylinders then begin to “string” together
(Figure 5.5d), the beads then grow in length (Figure 5.5e) and tightly pack to
forming long fibrils (Figure 5.5f). These are the two proposed mechanisms,
however, it is more than likely that no one path is followed completely but that either
path can be followed at anytime, as no two full fibrils are identical (Figure 5.5b, c, e
and f). A similar mechanism has been proposed for the formation of insulin fibrils as
170
“multipathway” fibrillization where both intertwining of protofilaments and the
lateral association of early prefibrillar forms followed by the lateral association of
protofilaments into fibrils was proposed.(73)
Light scattering intensity was used to investigate the fibril growth of A 40.
In Figure 5.6a the scattering intensity versus time for A 40 in the absence and
presence of 8 mM azoTAB is shown. Both the pure A 40 and the 8 mM cis azoTAB
solution under UV light initially have high scattering rates. From the SANS data we
would expect to see the small oligomers in the 8 mM cis azoTAB sample at time t =
0. However, due to the extremely high sensitivity of light scattering to any large
structures, even a very small amount of fibrils or other large aggregates would mask
the scattering form these small particles. Thus, although these small oligomers are
likely present they may not be detectable. Also note the decrease in scattering over
time, which also supports the idea of very large aggregates as they appear to be
settling out over time. Under visible light it was possible to observe the growth
process all the way from the early oligomers to full fibrils. The lag phase observed
with light scattering is ~2.5 hrs, based on the time elapsed before a steep upturn in
scattering intensity is observed, this agrees well with the lag phase seen with both
SANS and AFM.
171
Figure 5.6: (a) Light scattering intensity of pure A ( ) and A in the presence of 8 mM
azoTAB under both visible ( ) and UV light ( ) as a function of time. (b) Light scattering
intensity as a function of surfactant light illumination; visible after exposure to 1 hr UV
light( ) and UV after exposure to 1 hr visible light ( ). [A ] = 0.25 mg/mL
Figure 5.6b, illustrating the ability to trigger the growth process with UV
light, is a plot of intensity versus time from 8 mM azoTAB under both visible and
UV light. After 5 hrs both samples are exposed to the opposite light (visible to UV
and UV to visible) the sample converted from visible to UV light (trans to cis
azoTAB) has a jump in its scattering intensity, which corresponds to the scattering
intensity of the UV sample, while the sample converted from UV to visible light (cis
to trans azoTAB) exhibits no change in scattering intensity. This implies that the
aggregates, once formed, are not reversible, which agrees with the literature which
reports that even powerful denaturants can not dissociate amyloid aggregates.(39)
Comparing this to the 8 mM visible t = 5 hr and 8 mM visible-to-UV t = 5 hr from
SANS, it can be conclusively stated that that the fibril growth process can be
triggered with UV light. This agrees with literature which has shown that certain
variables (both temperature(14, 25, 26) and pH(67, 74)) can have an effect on the
speed of the fibrilization process of -amyloid peptide (1-40). The time scale at
172
which these parameters affect fibril growth ranges anywhere from a few hours to a
few days. Low pH (~5) results in an accelerated fibril growth process, however, it
was shown that these fibrils are actually less neurotoxic than those formed at neutral
pH(14, 67, 74). An increase in temperature has been shown to accelerate fibril
growth, while a decrease slows growth.(25, 26) It is then hypothesize that in the
presence of the visible light trans form of the surfactant, surfactant-protein
interactions are inhibiting protein-protein interactions. When the system is then
illuminated with UV light, the cis for of the surfactant no longer strongly interacts
with the protein allowing for protein-protein interactions to once again occur
“triggering” the growth process.
Under current experimental conditions the time frames in which reversibility
is possible could not be investigated (this is mostly due to the extremely high
absorbance of azoTAB). For example, immediately after the UV light has triggered
growth (surfactant has switched form the trans to the cis state) it is likely that if the
surfactant is switched back to its visible (trans) form fast enough (absorbance
limited) the triggered growth could be stopped or even reversed. This is likely
because it at some point during nucleation a critical size is reached where monomer
addition becomes favorable. This critical size defines the nucleus and eventually the
nucleus will be more stable than the monomer.(75, 76) The point at which this
happens is called the supercritical concentration, after this point nucleation becomes
an irreversible polymerization.(77) Thus, it is very possible that there is a point early
173
on during UV exposure at which the growth process is reversible or even that the
accelerated fibril growth actually results in less neurotoxic species.
5.5 Conclusion
In the presence of azoTAB at time t = 0 small globular oligomers are formed,
while in the absence of azoTAB fibrilization has already occurred. After 5 hrs under
visible light fibrils have begun to form, when illuminated with UV light the fibrils
further associated into 3-D networks similar to those observed for the pure peptide.
Thus azoTAB initially inhibits fibrilization in the -amyloid peptide and this delay
can be accelerated by illumination with UV light. The mechanism of fibrillization is
believed to follow two different pathways. After the initial oligomer species have
formed association proceeds in one of two ways, laterally or longitudinally. In both
cases the other type of association must eventually occur in order to develop full
amyloid fibrils. Laterally associated segments will begin to associate end to end to
form fibrils demonstrating the characteristic periodicity of amyloid fibrils. While the
longitudinally associated segments will wrap around one another to form fibrils
demonstrating the characteristic twisting of amyloids.
5.6 References
(1) Chiti, F., and Dobson, C. M. (2006) Protein Misfolding, Functional Amyloid, and
Human Disease. Annu. Rev. Biochem. 75, 333-366.
(2) Dobson, C. M. (1999) Protein misfolding, evolution and disease. Trends Biochem.
Sci. 24, 329-332.
174
(3) Kelly, J. W. (1998) The alternative conformations of amyloidogenic proteins and
their multi-step assembly pathways. Curr. Opin. Struc. Biol. 8, 101-106.
(4) Uversky, V. N., and Fink, A. L. (2004) Conformational constraints for amyloid
fibrillation: the importance of being unfolded. Biochim. Biophys. Acta 1698, 131-
153.
(5) Bitan, G., Kirkitadze, M. D., Lomakin, A., Vollers, S. S., Benedek, G. B., and
Teplow, A. B. (2003) Amyloid beta-protein (A ) assembly: A 40 and A 42
oligomerize through distinct pathways. P. Nat. Acad. Sci. 100, 330-335.
(6) Serpell, L. C., Sunde, M., Benson, M. D., Tennent, G. A., Pepys, M. B., and Fraser,
P. E. (2000) The protofilament substructure of amyloid fibrils. J. Mol. Biol. 300,
1033-1039.
(7) Roger, A. E., Chaney, M. O., Kuo, Y. M., Webster, S. D., Stine, W. B., . (1996)
Morphology and Toxicity of A -(1-42) Dimer Derived from neuritic and Vascular
Amyloid deposits of Alzheimer's Disease. J. Biol. Chem 271, 20631-20635.
(8) Leu, L. F., Kuo, Y. M., Roher, A. E., Brachova, L., Shen, Y., Sue, L., Beach, T.,
Kurth, J. H., Rydek, R. E., and Rogers, J. (1999) Soluble Amyloid beta Peptide
Concentration as a Predictor of Synaptic Change in Alzheimer's Disease. Am. J.
Pathol. 155, 853-962.
(9) McClean, C. A., Cherny, R. A., Fraser, F. W., Fuller, S. J., Smith, M. J.,
Vbeyreutheer, K., Bush, A. I., and Masters, C. L. (1999) Soluble pool of A amyloid
as a determinant of severity of neurodegeneration in Alzheimer's disease. Ann.
Neurol. 46, 860-866.
(10) Wang, D., Dickson, D. W., Trojanowski, J. Q., and Lee, V. M.-Y. (1999) The Levels
of Soluble versus Insoluble Brain A Distinguish Alzheimer's Disease from Normal
and Pathological Aging Exp. Neurol. 158, 328-337.
(11) Glenner, G. G., Wong, C. W. (1984) Alzheimer's Disease: initial report of the
purification and characterization of a novel cerebrovascular amyloid protein.
Biochem. Biophys. Res. Commun. 120.
(12) Nilsberth, C., Westllind-Danielsson, A., Eckman, C. B., Condron, M. M., Axelman,
K., Forseel, C., Stenh, C., Luthman, J., Teplow, D. B., Younklin, S. G., Naslund, J.,
and Lannfelt, L. (2001) The 'Arctic' APP mutation (E693G) causes Alzheimer's
disease by enhanced A protofibril formation. Nat. Neurosci. 4, 887-893.
(13) Krueger, S., Ho, D., and Tsai, A. (2006) Small-Angle Neutron Scattering as a Probe
for Protein Aggregation at Many Length Scales, in Misbehaving Proteins. Protein
(Mis)Folding, Aggregation, and Stability (Murphy, R. M., Tsai, A. M., Ed.) pp 125-
146, Springer, New York.
175
(14) Yong, W., Lomakin, A., Kirkitadze, M. D., Teplow, D. B., Chen, S.-H., and
Benedek, G. B. (2002) Structure Determination of Micelle-like Intermediates in
Amyloid -protein Fibril Assembly by using Small Angle Neutron Scattering. P.
Natl. Acad. Sci. USA 99, 150-154.
(15) Hamill, A. C., Wang, S-C, and Lee, C. T., Jr. (2007) Solution Structure of an
Amyloid-Forming Protein During Photoinitiated Hexamer-Dodecamer Transitions
Revealed Through Small-Angle Neutron Scattering. Biochemistry 46, 7694-7705.
(16) Heller, W. T., Krueger, J. K., and Trewhella, J. (2003) Further Insights into
Calmodulin-Myosin Light Chain Kinase Interaction from Solution Scattering and
Shape Restoration. Biochemistry 42, 10579-10588.
(17) Pallares, I., Vendrell, J., Aviles, F. X., and Ventura, S. (2004) Amyloid Fibril
Formation by a Partially Structured Intermediate State of a-Chymotrypsin. Journal
of Molecular Biology 342, 321-331.
(18) Fulop, L. Z., M., Datki, Z., Soos, K., and Penke, B. (2004) -Amyloid-derived
Pentapeptide RIIGLa Inhibits A (1-42) Aggregation and Toxicity. Biochem. and
Biophys. Res. Comm. 324, 64-69.
(19) Hamill, A. C., Wang, S-C., and Lee, C. T., Jr. (2005) Probing Lysozyme
Conformation with Light Reveals a New Folding intermediate. Biochemistry 44,
15139-15149.
(20) Le Ny, A.-L. M. a. L., C. T., Jr. (2006) Photoreversible DNA Condensation Using
Light-Responsive Surfactants. J. Am. Chem. Soc. 128, 6400-6408.
(21) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2005) Photocontrol of Protein
Folding: The Interaction of Photosensitive Surfactants with Bovine Serum Albumin.
Biochemistry 44, 524-536.
(22) Wang, S.-C., and Lee, C. T., Jr. (2006) Protein Secondary Structure Controlled with
Light and Photoresponsive Surfactants. Journal of Physical Chemistry B 110,
16117-16123.
(23) Kanek, I., Tutumi, S. (1997) Conformations of -Amyloid in Solution. J.
Neurochem. 68, 438.
(24) Glinka, C. J., Barker, J. G., Hammouda, B., Krueger, S., Moyer, J. J., and Orts, W. J.
(1998) The 30 m small-angle neutron scattering instruments at the National Institute
of Standards and Technology. Journal of Applied Crystallography 31, 430-445.
(25) Harper, J. D., Wong, S. S., Lieber, C. M., and Lansbury, P. T., Jr. . (1999) Assembly
of A Amyloid Protofibrils: An in Vitro Model for a Possible Early Event in
Alzheimer's Disease. Biochemistry 38, 8972-8980.
176
(26) Kusumoto, Y., Lomakin, A., Teplow, D. B., and Benedek, G. B. (1998) Temperature
dependence of amyloid -protein fibrillizaiton. P. Natl. Acad. Sci. USA 95, 12277-
12282.
(27) Kline, S. R. (2006) Reduction and analysis of SANS and USANS data using IGOR
Pro. J. Appl. Cryst. 39, 895.
(28) Guinier, A. a. F., G. (1955) Small Angle Scattering of X-ray, Wiley, New York.
(29) Porod, G. (1982) in Small Angle X-ray Scattering (Glatter, O., Ed.) pp 17-51,
Academic Press, London.
(30) Feigin, L. A. a. S., D. I. (1987) Structure Analysis by Small-Angle X-Ray and
Neutron Scattering, Plenum Press, New York.
(31) Glatter, O. (1977) A new method for the evaluation of small-angle scattering data. J.
Appl. Cryst. 10, 415-421.
(32) Svergun, D. I. (1992) Determination of the Regularization Parameter in Indirect-
Transform Methods Using Perceptual Criteria. J. Appl. Cryst. 25, 495-503.
(33) Yan, H., Saiani, A., Miller, A. (2007) in APS March Meeting, Denver, CO.
(34) Zhou, W., Islam, M. F., Wang, H., Ho, D. L., Yodh, A. G., Winey, K. I., and
Fischer, J. E. (2004) Small angle neutron scattering from single-wall carbon
nanotube suspensions. Evidence for isolated rigid rods and rod networks. Chem.
Phys. Lett. 384, 185-189.
(35) Yan, H., Saini, A., Gough, J. E., and Miller, A. F. (2006) Thermoreversible Protein
Hydrogel as Cell Scaffold. Biomacromolecules 7, 2776-2782.
(36) Daoud, M., Cotton, J. P., Farnoux, B., Jannink, G., Sarma, G., Benoit, H., Duplessix,
C., Picot, C., and De Gennes, P. G. (1975) Solutions of flexible polymers. Neutron
experiments and interpretation. Macromolecules 8, 804-18.
(37) Jeng, U. S., Lin, T.-L., Lin, J. M., and Ho, D. L. . (2006) Contrast Variation SANS
for the Solution Structure of the -amyloid Peptide 1-40 Influenced by SDS
Surfactants. Physica B 385-386, 865-867.
(38) Walsh, D. M., Lomakin, A., Benedek, G. B., Condron, M. M., and Teplow, D. B.
(1997) Amyloid -Protein Fibrillogenesis. J. Biol. Chem. 272, 22364-22372.
(39) Horwich, A. (2002) Protein aggregation in disease: a role for folding intermediates
forming specific multimeric interactions. J. Clin. Invest. 110, 1221-1232.
(40) Antzutkin, O. N., Balbach, J. J., and Tycko, R. (2003) Site-specific identification of
non- -strand conformations in Alzheimer's -amyloid fibrils by solid-state NMR.
Biophys. J. 84, 3326-3335.
177
(41) Antzutkin, O. N., Leapman, R. D., Balbach, J. J., and Tycko, R. (2002)
Supramolecular structural constraints on Alzheimer's -amyloid fibrils from electron
microscopy and solid state NMR. Biochemistry 41, 15436-15450.
(42) Goldsbury, C. S., Wirtz, S., Muller, S. A., Sunderji, S., Wicki, P., Aebi, U., and
Frey, P. (2000) Studies on the in vitro assembly of A 40: Implications for the search
for A fibril formation inhibitors. J. Struct. Biol. 130, 217-231.
(43) Petkova, A. T., Ishii, Y., Balbach, J. J., Antzutkin, O. N., Leapman, R. D., Delaglio,
F., and Tycko, R. (2002) A structural model for Alzheimer's -amyloid fibrils based
on experimental constraints from solid state NMR. P. Nat. Acad. Sci. 99, 16742-
16747.
(44) Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2007) Solution Structure of an
Amyloid-Forming Protein During Photoinitiated Hexamer-Dodecamer Transitions
Revealed Through Small-Angle Neutron Scattering. Biochemistry 46, 7694-7705.
(45) Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2005) Probing lysozyme
conformation with light reveals a new folding intermediate. Biochemistry 44, 15139-
15149.
(46) Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2005) Photocontrol of Protein
Folding: The Interaction of Photosensitive Surfactants with Bovine Serum Albumin.
Biochemistry 44, 524-536.
(47) Wang, S. C., and Lee, C. T., Jr. (2007) Enhanced Enzymatic Activity Through
Photoreversible Conformation Changes Biochemistry 46, 14557-14566.
(48) Wong, S. S., Harper, J. D., Lansbury, P.T., and Lieber, C. M. (1998) Carbon
nanotube tips: high resolution probed for imaging biological systems. J. Am. Chem.
Soc 120, 603-604.
(49) Kayed, R., Head, E., Thompson, J. L., Mcintire, T. M., Milton, S. C., Cotman, C.
W., and Glabe, C. G. (2003) Common Structure of Soluble Amyloid Oligomers
Implies Common Mechanism of Pathogenesis. Science 300, 486-489.
(50) LeVine, H. (2002) 4,4'-Dianilino-1,1'-binaphthyl-5,5'-disulfonate: report on no-beta-
sheet conformers of Alzheimer's peptide A (1-40). Arch. Biochem. Biophys. 404,
106-115.
(51) Soreghan, B., Kosmoski, J., and Glabe, C. (1994) Surfactant properties of
Alzheimer’s A peptides and the mechanism of amyloid aggregation. J. Biol. Chem.
269, 28551-28554.
(52) Tjernberg, L. O., Pramanik, A., Jorling, S., Thyberg, P., Thyberg, J., Nordstedt, C.,
Berndt, K. D., Terenius, L., and Rigler, R. (1999) Amyloid beta-peptide
polymerization studied using fluorescence correlation spectroscopy. Chem. Biol. 6,
53-62.
178
(53) Holmberg, K., Jonsson, B., Kronberg, B., and Lindman, B. (2003) Surfactant and
Polymers in Aqueous Solution, 2nd Edition, John Wiley and Sons, Chichester, UK.
(54) Benditt, E. P., Eriksen, N., and Berglund, C. (1970) Congo Red Dichroism with
Dispersed Amyloid Fibrils, an Extrinsic Cotton Effect P. Nat. Acad. Sci 66, 1044-
1051.
(55) Glenner, G. G. (1980) Amyloid deposits and amyloidosis. I. The b-fibrilloses. N.
Eng. J. Med. 302, 1283-1292.
(56) Klunk, W. E., Debanth, M. L., and Pettegrew, J. W. (1994) Congo Red Populates
Partially Unfolded States of an Amyloidogenic Protein to Enhance Aggregation and
Amyloid Fibril Formation. Neurobiol. Aging 15, 691-698.
(57) Klunk, W. E., Jacob, R. F., and Mason, R. P. (1999) Quantifying amyloid by Congo
red spectral shift assay. Methods Enzymol. 309, 285-305.
(58) Turnell, W. G., and Finch, J. T. (1992) Binding of the dye congo red to the amyloid
protein pig insulin reveals a novel homology amongst amyloid-forming peptide
sequences. J. Mol. Biol 227, 1205-1223.
(59) Hayashita, T., Kurosawas, T., Miyata, T., Tanaka, K., and Igawa, M. (1994) Effect
of structural variation within cationic azo-surfactant upon photoresponsive function
in aqueous solution. Colloid Polym. Sci. 272, 1611-1619.
(60) Stine, W. B., Jr., Snyder, S. W., Ladror, U. S., Wade, W. S., Miller, M. F., Perun, T.
J., Holzman, T. F., and Krafft, G. A. (1996) The nanometer-scale structure of
amyloid-b visualized by atomic force microscopy. J. Protein Chem. 15, 193-203.
(61) Harper, J. D., Lieber, C. M., and Lansbury, P. T. (1997) Atomic force microscopic
imaging of seeded fibril formation and fibril branching by the Alzheimer's amyloid-
beta protein. Chem. Biol. 4, 9851-959.
(62) Seilheimer, B., Bohrmann, B., Bondolfi, L., Mueller, F., Stueber, D., and Doebeli,
H. (1997) Alzheimer's beta amyloid peptide correlates with a distinct fiber
morphology. J. Struct. Biol. 119, 59-71.
(63) Goldsbury, C., Kistler, J., Aebi, U., Arvinte, T., and Cooper, G. J. S. (1999)
Watching Amyloid Fibrils Grow by Time-lapse Atomic Force Microscopy. J. Mol.
Biol. 285, 33-39.
(64) Bauer, H. H., Aebi, U., Haener, M., Hermann, R., Mueller, M., Arvinte, T., and
Merkle, H. P. (1995) Architecture and polymorphism of fibrillar supramolecular
assemblies produced by in vitro aggregation of human calcitonin. J. Struct. Biol.
115, 1-15.
179
(65) Serio, T. R., Cashikar, A. G., Kowal, A. S., Sawicki, G. J., Moelehi, J. J., Serpell, L.,
Arnsdorf, M. F., and Lindquist, S. L. (2000) Nucleated Conformational Conversion
and the Replication of Conformational Information by a Prion Determinant. Science
289, 1317-1321.
(66) Kad, N. M., Myers, S. L., Smith, D. P., Smith, D. A., Radford, S. E., and Thomson,
N. H. (2003) Hierarchical Assembly of 2-Microglobulin Amyloid In Vitro
Revealed by Atomic Force Microscopy. J. Mol. Biol. 330.
(67) Klug, G. M. J. A., Losic, D., Subasinghe, S. S., Aguilar, M-I., Martin, L. L., and
Small, D. H. . (2003) -Amyloid Protein Oligomers Induced by Metal Ions and Acid
pH are Distinct from those Generated by Slow Spontaneous Ageing at Neutral pH
Eur. J. Biochem. 270, 4282-4293.
(68) Guijarro, J. I., Sunde, M., Jones, J. A., Campbell, I. D., and Dobson, C. M. (1998)
Amyloid Fibril Formation by an SH3 Domain. Biochemistry 95, 4224-4228.
(69) Burkoth, T. S., Benzinger, T. L. S., Urban, V., Morgan, D. M., Gregory, D. M.,
Thiyagarajan, P., Botto, R. E., Meredith, S. C., and Lynn, D. G. (2000) Structure of
the -Amyloid(10-35) Fibril. J. Am. Chem. Soc. 122, 7883-7889.
(70) Harris, J. R. (2002) In vitro Fibrillogenesis of the Amyloid (1-42) Peptide;
Cholesterol Potentiation and Aspirin Inhibition. Micron 33, 609-626.
(71) Quijano, F. A. R., Morrow, D., Wise, B. M., Brancia, F. L., and Goux, W. J. (2006)
Prediction of Nucleation Sequences form Amyloidogenic Propensities of Tau-
Related Peptides. Biochemistry 45, 4638-4652.
(72) Krebs, M. R., MacPhee, C. E., Miller, A. F., Dunlop, I. E., Dobson, C. M., and
Donald, A. M. (2004) The Formation of Spherulites by Amyloid Fibrils of Bovine
Insulin. P. Nat. Acad. Sci. USA 101, 14420-14424.
(73) Jansen, R., Dzwolak, W., and Winter, R. (2005) Amyloidogenic Self-Assembly of
Insulin Aggregates Probed by High Resolution Atomic Force Microscopy. Biophys.
J. 88, 1344-1353.
(74) Wood, S. J., Maleef, B., Hart, T. & Wetzel, R. (1996) Physical, Morphological, and
Functional Differences between pH 5.8 and 7.4 aggregates of the Alzheimer's
amyloid peptide A . J . Mol. Biol. 256, 870-877.
(75) Everett, D. H. (1986) Some thermodynamic aspects of aggregation phenomena with
special reference to micellization. Colloid Surface 21, 41-53.
(76) Firestone, M. P., Delevie, R., and Rangarajan, S. K. (1983) On one dimensional
nucleation and growth of living polymers. 1. Homogenous nucleation. J. Theor.
Biol. 104, 535-552.
180
(77) Powers, E. T., Powers, D. L. (2006) The kinetics of Nucleated Polymerization at
High Concentration: Amyloid Fibril Formation Near and Above the "Supercritical
Concetration". Biophys. J. 91, 122-132.
181
Chapter VI: Future Studies
6.1 Introduction
In this chapter the preliminary results of several experiments will be
discussed along with suggestion for their completion. The fist section investigates
the effect of self-association on -chymotrypsin activity. This leads to the second
section which proposes a study on the effect of azoTAB on another protein,
proMMP-9, that like -chymotrypsin is inhibited by -1-antichymotrypsin. The
third is a group of experiments investigating the effect of azoTAB on a variety of
different proteins including: -lactalbumin, maltose binding protein, guanylate
kinase, fibrinogen, and two SH3 domains.
6.2 Controlling -chymotrypsin Activity with Light
6.2.1 Preliminary Data
In Chapter 4 it was demonstrated that light-responsive surfactants can be used
to photo-reversibly control -chymotrypsin self-association.(1) Although it was
concluded that the active site is not directly involved in the association, (since the
monomers association in a “back-to-side” orientation), preliminary experiments have
shown promising results in the control of -chymotrypsin activity with light, as
shown in Figure 6.1.
182
Figure 6.1: Fluorescence of -chymotrypsin solutions in the presence of various azoTAB
derivatives, see Figure 6.3. [ -chymotrypsin] =10mg /mL.
Fluorescence measurements were used to probe -chymotrypsin activity
through the use of a fluorescent dye-labeled substrate BODIPY® TR-X casein
(Molecular Probes, EnzChek® Protease Assay Kit *red fluorescence*, cat no.
E6639). BODIPY ® TR-X casein, is a casein derivative that is heavily labeled with
the red-fluorescent BODIPY ® TR-X dye, which results in an almost total quenching
of the conjugate’s fluorescence. As a results, protease-catalyzed hydrolysis releases
highly fluorescent BODIPY® TR-X dye-labeled peptides (Figure 6.2). The resulting
fluorescence is proportional to the protease activity. 0.2 mL of a 0.1 M sodium
bicarbonate solution was added to the lyophilized substrate. 20 μL of the
183
substrate/dye solution was then added to 2mL of a10mg/mL -chymotrypsin solution
with various photoresponsive surfactants.
Figure 6.2: Principle of protease detection used in Molecular Probes’ EnzChek Protease
Assay Kit.
Once visible light is turned on the surfactant starts to convert form the cis to the
trans isomer, from Figure 6.1 it can be seen that this causes an immediate increase in
activity. The slope is a measure of activity, with a steeper slope indicating higher
activity. In all cases activity increase with visible light, with the magnitude of
reactivation varying significantly with surfactant structure (see Figure 6.3).
Furthermore, increasing the concentration of a given surfactant also appears to
enhance reactivation. The longer the aliphatic tail length, the larger the increase in
activity with visible light, where the surfactant is in the trans form. The activity is
the same for all surfactants under UV light, where the surfactant is in the cis form,
and it is the same as that of the pure protein. Therefore it is very possible that with
surfactant in the visible-light form superactivity is achieved. Superactivity has been
shown previously with -chymotrypsin and non-photoresponsive surfactants(2, 3).
In this case (very similar to the lysozyme system)(4) a simple “flip of the switch”
(from UV to visible light) changes the enzyme from native activity to superactivty.
184
Figure 6.3: Chemical structures of azoTAB derivatives; S1, S4 and S10
6.2.2 Future Experiments
Although the results are promising, only a few assays have been collected. A
more in-depth study involving the above azoTAB surfactants (varying each
surfactant concentration from 0-10mM) is warranted. Also, commonly used
surfactants such as CTAB and CTBABr should be studied to reproduce activity
results form the literature(2, 3) and to then compare these to the activity achieved
with the different photoresponsive (azoTAB) surfactants. The superactivity is likely
a direct result of less association under visible light, allowing for greater access to
the active sites.
185
6.3 Using Light Sensitive Surfactants to Treat Burn Victims
6.3.1 Background
It was recently discovered that -1-antichymotrypsin is a physiological
inhibitor in controlling the proteolytic activation of proMMP-9 in tissue injury.(5)
proMMP-9 is transiently expressed after acute injury and then declines as wound
healing proceeds.(6-9) In chronic wounds and many other degenerative conditions,
proMMP-9 continues to be expressed and is primarily converted into the active
form.(10-13) Excessive and prolonged expression and activation of MMP-9 is
believed to be a cause of chronic/non-healing wounds.
6.3.2 Future Experiments
It has been shown that photoresponsive surfactants can be used to control the
association of -chymotrypsin,(1) and preliminary results also show that the activity
of -chymotrypsin can be controlled in the same manner. -1-antichymotrypsin is
known to inhibit both -chymotrypsin and proMMP-9. Therefore, it is likely that if
-chymotrypsin activity can be controlled, proMMP-9 activity can be controlled as
well (the only way to truly determine if this is the case is to study effects of azoTAB
on both the structure and activity of proMMP-9). If this is the case photoresponsive
surfactants can be used to treat burn victims by preventing excessive and prolonged
expression and activation of proMMP-9, thus, avoiding chronic/non-healing wounds.
The first experiment to perform is an enzymatic assay (such as the
fluorescence experiment discussed in section 5.1 above) to determine if
186
photoresponsive surfactants have any effect on proMMP-9 activity. If the activity of
proMMP-9 can be controlled, small angle neutron scattering should be preformed to
determine if this is an effect of association of the protein or simply a denaturing of
the active site. The result obtained from small angle neutron scattering will aid in
determining the remainder of experiments necessary to fully characterize the system.
6.4 Unpublished Small Angle Neutron Scattering Data
6.4.1 Background
Preliminary SANS data has been collected on a variety of protein/photo-
surfactant systems. These systems include: -lactalbumin/S4, maltose binding
protein/S1, guanylate kinase/S1, fibrinogen/S1, and SH3 domains/S1.
-lactalbumin is a whey protein found in the milk of many mammalian
species. It has a molecular weight (14 kDa) and an amino acid sequence similar to
lysozyme. The same surfactant (S4, structure shown in Figure 6.3) used in the
photocontrol of lysozyme in Chapter 2(14) is used again here. The goal is to use S4
to induce and reversibly control the molten globule state (a protein state exhibiting a
tertiary structure similar to it’s native state but a “molten” secondary structure) of -
Lactalbumin. The molten globule form of -Lactalbumin has been shown to cause
apoptosis in tumor cells.(15) Thus photo-control could be used to target only tumor
cells, through the use of a fiber optic, leaving healthy cells unharmed. The
preliminary data will be discussed below.
187
Maltose binding protein (MBP) is part of the maltose/maltodexrin system of
E. Coli bacteria. It has a molecular weight of ~40 kDa.
Gaunylate kinase (GK) catalyses the phosphorylation of guanosine
monophosphate (GMP) to guanosine diphosphate (GDP). The enzyme does this by
transferring a phosphate from adenosine triphospahate (ATP). It has a molecular
weight of ~ 20kDa. This protein was chosen because of its enzymatic properties, the
goal is to show that photosurfactants can be used to induce and control structural
changes which in turn will allow for control of GK enzymatic activity.
Fibrinogen is the principal protein of vertebrate blood clotting. It is a fibrillar
protein that converts to fibrin upon injury forming a mesh, together with platelets a
blood clot is formed. It has a molecular weight of ~340kDa. It was chosen after the
A study to further investigate fibrillation into meshes.
In collaboration with Salvador Ventura’s lab(16) the amyloid lag phase of
two different SH3 domains and the effect of S1 on these structures has been
explored. The SH3 domain of the p85 subunit of phosphatidylinositol 3′-kinase
(PI3-SH3) has been found to form amyloid fibrils in vitro under acidic conditions.
PI3-SH3 is atypical due to a large insertion of 15 amino acid residues in the n-Src
loop when compared with more canonical members of the family. Spectrin-SH3
(SPC-SH3) with a shorter loop does not form fibrils under similar conditions. SPC-
SH3 MW is 7.2 kDa whereas PI3-SH3 is 9.6 kDa due to a helix insertion in one
loop. The Src homology 3 domain (SH3 domain) is a small protein domain of about
60 amino acid residues. It has a characteristic fold which consists of five or six β-
188
strands arranged as two tightly packed anti-parallel β sheets. The SH3 domain is
found in proteins that interact with other proteins. They arbitrate the assembly of
specific protein complexes by binding to proline-rich peptides in their respective
binding partner.
6.4.2 Preliminary Data
SANS data collected on the -lactalbumin/S4 system is shown in Figure 6.4.
The steep slope of the scattering intensity at low Q is indicative of the formation of
large protein aggregates, as was seen with A in Chapter 5. Both the effects of light
as well as the effects of surfactant concentration on the scattering intensity are only
seen when Q < 0.08 Å
-1
. At Q > 0.08 Å
-1
the size of the protein monomer is being
approached, indicating that the monomer remains relatively intact (it is not
dramatically unfolding). From the SANS data it can be concluded that -
lactalbumin is associating similar to the -chymotrypsin(1) and A (17) systems and
not swelling as was the case with lysozyme(14) and RNase A(18). Surfactant
concentration and light illumination both appear to have an effect on this association.
As surfactant concentration increases scattering intensity increase, with the exception
of the highest concentration. This indicates that surfactant causes a slight unfolding
to a certain extent once this limit is reached increasing surfactant concentration
further no longer has an effect. Also in all cases the scattering is higher under
visible light than under UV light. Indicating that when surfactant binding is turned
on (visible light) unfolding occurs which leads to association but when surfactant
189
binding is turned off (UV light) this slight unfolding no longer occurs resulting in
less association. In the case of 0.125 mM azoTAB the scattering under UV light is
about the same as that of the pure protein.
Figure 6.4: SANS data of -Lactalbumin/azoTAB solutions as a function of surfactant
concentration and light illumination. [ -Lactalbumin] = 10mg/mL.
SANS data collected on the MBP/S1 system is shown in Figure 6.5. The
slope of the scattering intensity at low Q may be indicative of a very small
population of large aggregates but is not significant enough to indicate association.
Both light illumination and surfactant concentration have an effect on the scattering
intensity across almost the entire Q range. What this means is that the monomer
190
structure of MBP is changing (possibly swelling or unfolding) with both surfactant
concentration and light illumination. Under visible light the shape of the curves are
very similar to that of the pure protein and it appears as if the protein is swelling as
surfactant concentration is increased. The scattering curves at 22 and 30 mM are
almost identical; this indicates that the protein has reached some maximum
“swollen” state possibly the surfactant saturation limit discussed in previously(18).
Under UV light the curves have changed shape and there now appears to be a
difference between the 22 and 30mM. The change in shape suggest a transition from
a small globular structure, perhaps small oligomers similar to the -chymotrpsin and
A oligomers, to higher order association, like the seed to fibril transition in the A
study. These results suggest the ability to use azoTAB to induce and control changes
in the structure of MBP.
191
Figure 6.5: SANS data of MBP/azoTAB solutions as a function of surfactant concentration
and light illumination. [MBP] = 5 mg/mL.
SANS data collected on the GK/S1 system is shown in Figure 6.6. The lack
of data for the pure protein makes it difficult to fully analyze the data shown, as it
can not be directly related to the scattering from the pure protein. However it is quite
clear that both surfactant concentration and light illumination have an effect on the
structure of GK. The SANS data suggest a swelling or unfolding system with a
slight population of larger aggregates as opposed to an associated system, that can be
induced to a different degree by varying surfactant concentration and light
illumination.
192
Figure 6.6: SANS data of guanylate kinase in the presence of azoTAB (S1) under both
visible and UV light. [Guanylate kinase] = 10mg/mL.
SANS data collected on the fibrinogen/S1 system is shown in Figure 6.7. As
was the case with GK, the lack of data for the pure protein makes it difficult to fully
analyze the data shown. However from what is known about the protein, the data
presented here and what has been learned from the A system, it is clear that the
protein is associating into long fibrils. Under visible light the surfactant inhibits this
fibrillation while under UV light fibrillation is allowed to occur resulting in a slope
of -4 at low Q, this slope is indicative of long fibril structures. There is clearly a
significant change upon light illumination.
193
Figure 6.7: SANS data of fibrinogen in the presence of azoTAB (S1) under both visible and
UV light [fibrinogen] = 5 mg/mL
SANS data was collected on both SH3 domains as a function of time, see
Figure 6.8. For the SPC-SH3 domain spectra were collected on the pure protein at t
= 4,48, and 72 hrs, and with 8 mM S1 at t = 4hrs. Data was collected on the PI3-
SH3 domain at the same times for the pure protein (t = 4, 48, and 72 hrs) and at a
slightly later time with 8 mM azoTAB, t = 12 hrs. The slope of -4 at low Q suggests
the formation of long fibrils, which, from previous experiments, we expect to be the
beginning of fibril formation. The shoulders at Q ~ 0.1 Å
-1
are indicative of
small oligomers (seeds), thus we can observe the transition from the small particles
to larger fibers.
194
Figure 6.8: SANS data of the SPC and PI3 SH3 domains as a function of time and in the
presence of azoTAB (S1). SPC-SH3 t = 4 hrs ( ), t = 48 hrs ( ), t = 72 hrs ( ), w/ 8 mM
azoTAB, t = 4 hrs ( ). PI3-SH3 t = 4 hrs ( ), t = 48 hrs ( ), t = 72 hrs (-), w/ 8 mM
azoTAB, t = 12 hrs ( ). [SPC-SH3] = [PI3-SH3] = 8.5 mg/mL.
The PI3-SH3 system exhibits a mixture of seed particles and fibers with an
aggregated state largely independent of time (4, 48, 72 hr samples give similar
scattering). The SPC-SH3 system also exhibits a mixture of small oligomers (seeds)
and fibers with an aggregated state largely independent of time (4, 48, 72 hr samples
give similar scattering). When the azoTAB photosurfactant (“S1”) is introduced into
either system it appears to prevent fibril formation, over this time domain, since the
Q
-4
region (fibers) is no longer evident. Thus, only the scattering from the small
195
oligomers is observed. Interestingly, preliminary Guinier fits shown in Figure 6.9
suggest that the small oligomers in the SPC-SH3, Figure 6.9a, R
g
~ 13.9 Å, are
smaller than those in the PI3-SH3, Figure 6.9b, R
g
~ 19.2 Å.
Figure 6.9: Guinier Analysis of the SANS data from (a) the SPC-SH3 domain with 8 mM
azoTAB and (b) the PI3-SH3 domain with 8mM azoTAB.
(a)
(b)
196
6.4.3 Future Analysis and Experiments
All of the data above should be fully analyzed to include Guinier or Modified
Guinier Analysis, PDDF calculations, and shape reconstruction where appropriate.
In the case of the associating proteins ( -lact and fibrinogen) AFM imaging is
recommended. One might also benefit from simple optical microscope tests (i.e.
Congo red fluorescence and/or imaging under cross polarized light). In the case of
the swelling/unfolding systems (MBP and GK) enzymatic assays along with
fluorescence spectroscopy, dynamic light scattering, UV-vis spectroscopy and FT-IR
are recommended. It is highly recommend that SANS be preformed on both pure
Fibrinogen and pure GK.
In the case of the SH3 domains it should be possible to deconvolute and
separate the scattering from the small oligomers and the fibers, thus, allowing for
shape reconstruction” to determine the conformation of the small oligomers. In
principle, SPC-SH3 was not expected to form amyloid fibrils under these conditions
and some evolution of the aggregated forms in PI3-SH3 along time was expected.
Therefore it is necessary to revisit these experimental conditions to be sure that what
was observed is indeed real.
6.5 References
(1) Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2007) Solution Structure of an
Amyloid-Forming Protein During Photoinitiated Hexamer-Dodecamer Transitions
Revealed Through Small-Angle Neutron Scattering, Biochemistry 46, 7694-7705.
197
(2) Spreti, N., Alfani, F., Cantarella, M., D'Amico, F., Germani, R., and Savelli, G.
(1999) a-Chymotrypsin superactivity in aqueous solutions of cationic surfactants, J.
Mol. Catal. B-Enzym. 6, 99-110.
(3) Spreti, N., Di Profio, P., Marte, L., Bufali, S., Brinchi, L., and Savelli, G. (2001)
Activation and stabilization of a-chymotrypsin by cationic additives, Eur. J.
Biochem. 268, 6491-6497.
(4) Wang, S. C., and Lee, C. T., Jr. (2007) Enhanced Enzymatic Activity Through
Photoreversible Conformation Changes, Biochemistry 46, 14557-14566.
(5) Han, Y.-P., and Garner, W. L. (2006) Alpha 1 antichymotrypsin is a physiological
inhibitor of proMMP-9 activation by human skin.
(6) Moses, M. A., Marikovsky, M., Harper, J. W., Vogt, P., Eriksson, E., Klagsbrun, M.,
and Langer, R. (1996) Temporal study of the activity of matrix metalloproteinases
and their endogenous inhibitors during wound healing, J. Cell. Biochem. 60, 379-
386.
(7) Salo, T., Makela, M., Kylmaniemi, M., Autio-Harmainen, H., and Larjava, H.
(1994) Expression of matrix metalloproteinase-2 and -9 during early human wound
healing, Lab. Invest. 70, 176-182.
(8) Tarlton, J. F., Vickery, C. J., Leaper, D. J., and Bailey, A. J. (1997) Postsurgical
wound progression monitored by temporal changes in the expression of matrix
metalloproteinase-9, Brit. J. Dermatol. 137, 506-516.
(9) Young, P. K., and Grinnell, F. (1994) Metalloproteinase activation cascade after
burn injury: a longitudinal analysis of the human wound environment, J. Invest.
Dermatol. 103, 660-664.
(10) Ladwig Glenn, P., Robson Martin, C., Liu, R., Kuhn, M. A., Muir David, F., and
Schultz Gregory, S. (2002) Ratios of activated matrix metalloproteinase-9 to tissue
inhibitor of matrix metalloproteinase-1 in wound fluids are inversely correlated with
healing of pressure ulcers, Wound Repair Regen. 10, 26-37.
(11) Weckroth, M., Vaheri, A., Lauharanta, J., Sorsa, T., and Konttinen, Y. T. (1996)
Matrix metalloproteinases, gelatinase and collagenase, in chronic leg ulcers, J.
Invest. Dermatol. 106, 1119-1124.
(12) Wysocki, A. B., Staiano-Coico, L., and Grinnell, F. (1993) Wound fluid from
chronic leg ulcers contains elevated levels of metalloproteinases MMP-2 and MMP-
9, J. Invest. Dermatol. 101, 64-68.
(13) Yager, D. R., Zhang, L.-Y., Liang, H.-X., Diegelmann, R. F., and Cohen, I. K.
(1996) Wound fluids from human pressure ulcers contain elevated matrix
metalloproteinase levels and activity compared to surgical wound fluids, J. Invest.
Dermatol. 107, 743-748.
198
(14) Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2005) Probing lysozyme
conformation with light reveals a new folding intermediate, Biochemistry 44, 15139-
15149.
(15) Svensson, M., Hakansson, A., Mossberg, A.-K., Linse, S., and Svanborg, C. (2000)
Conversion of alpha-Lactalbumin to a protein inducing apoptosis, P. Nat. Acad. Sci.
97, 4221-4226.
(16) Ventura, S., Lacroix, E., and Serran, L. (2002) Insights into the Origin of the
Tendency of the PI3-SH3 Domain to form Amyloid Fibrils, J. Mol. Biol. 322, 1147-
1158.
(17) Hamill, A. C., and Lee, T. C., Jr. (in prep) -amyloid peptide (1-40) fibril growth
influenced by photosurfactant.
(18) Hamill, A. C., and Lee, C. T., Jr. (in prep) Control of Ribonuclease A structure
through the use of photoresponsive surfactants.
199
BIBLIOGRAPHY
Ackers, G. K., and Thompson, T. E. (1965) Determination of stoichiometry and
equilibrium constants for reversibly associating systems by molecular sieve
chromatography, P. Nat. Acad. Sci. 53, 342-349.
Antzutkin, O. N., Balbach, J. J., and Tycko, R. (2003) Site-specific identification of
non- -strand conformations in Alzheimer's -amyloid fibrils by solid-state NMR.
Biophys. J. 84,
3326-3335.
Antzutkin, O. N., Leapman, R. D., Balbach, J. J., and Tycko, R. (2002)
Supramolecular structural constraints on Alzheimer's -amyloid fibrils from electron
microscopy and solid state NMR. Biochemistry 41, 15436-15450.
Aune, K. C., and Timasheff, S. N. (1971) Dimerization of alpha-chymotrypsin. I. pH
dependence in the acid region. Biochemistry 10, 1609-1617.
Bacaloglu, R., Bunton, C. A., and Ortega, F. (1989) Micellar enhancements of rate
of SN2 reactions of halide ions: the effect of head group size J. Phys. Chem. 93,
1497-1502.
Bacaloglu, R., Bunton, C. A., Cerichelli, G., and Ortega, F. (1989) NMR study of
the location of bromide ion and methyl naphtalene-2-sulfonate in cationic micelles:
reaction to reactivity. J. Phys. Chem. 93, 1490-1497.
Bachmann, A., Segel, D., and Kiefhaber, T. (2002) Test for cooperativity in the
early kinetic intermediate in lysozyme folding, Biophys. Chem. 96, 141-151.
Bauer, H. H., Aebi, U., Haener, M., Hermann, R., Mueller, M., Arvinte, T., and
Merkle, H. P. (1995) Architecture and polymorphism of fibrillar supramolecular
assemblies produced by in vitro aggregation of human calcitonin. J. Struct. Biol.
115, 1-15.
Benditt, E. P., Eriksen, N., and Berglund, C. (1970) Congo Red Dichroism with
Dispersed Amyloid Fibrils, an Extrinsic Cotton Effect P. Nat. Acad. Sci 66, 1044-
1051.
Berret, J.-F., and Oberdisse, J. (2004) Electrostatic self-assembly in polyelectrolyte-
neutral block copolymers and oppositely charged surfactant solutions, Phys. B:
Condensed Matter (Amsterdam)350, 204-206.
200
Bertsch, M., Mayburd, A. L., and Kassner, R. J. (2003) The identification of
hydrophobic sites on the surface of proteins using absorption difference
spectroscopy of bromophenol blue, Anal. Biochem. 313, 187-195.
Bitan, G., Kirkitadze, M. D., Lomakin, A., Vollers, S. S., Benedek, G. B., and
Teplow, A. B. (2003) Amyloid beta-protein (A ) assembly: A 40 and A 42
oligomerize through distinct pathways. P. Nat. Acad. Sci. 100, 330-335.
Blow, D. M. (1971) The structure of chymotrypsin, in The Enzymes (Boyer, P. D.,
Ed.) pp 185-212, Academic Press, New York.
Bonan, C., Germani, R., Ponti, P. P., Savelli, G., Cerichelli, G., and Bunton, C. A.
(1990) Micellar head-group size and anion nucleophilicity in SN2 reactions. J. Phys.
Chem. 94, 5331-5336.
Booth, D. R., Sunde, M., Bellotti, V., Robinson, C. V., Hutchinson, W. L., Fraser, P.
E., Hawkins, P. N., Dobson, C. M., Radford, S. E., Blake, C. C. F., and Pepys, M. B.
(1997) Instability, unfolding, and aggregation of human lysozyme variants
underlying amyloid fibrillogenesis, Nature 385, 787-793.
Bowie, J. U. (2001) Stabilizing membrane proteins. Curr. Opin. Struc. Biol 11, 397-
402.
Brandts, J. F., Halvorson, H. R., and Brennan, M. (1975) Consideration of the
Possibility that the slow step in protein denaturation reactions is due to cis-trans
isomerism of proline residues. Biochemistry 14, 4953-4963.
Brange, J., Andersen, L., Laursen, E. D., Meyn, G., and Rasmussen, E. (1997)
Toward understanding insulin fibrillation, J. Pharm. Sci. 86, 517-525.
Brooks, C. L., III. (1998) Simulations of protein folding and unfolding. Curr. Opin.
Struc. Biol. 8, 222-226.
Bu, Z., and Engelman, D. M. (1999) A method for determining transmembrane helix
association and orientation in detergent micelles using small angle X-ray scattering.
Biophys. J. 77, 1064-1073.
Burkoth, T. S., Benzinger, T. L. S., Urban, V., Morgan, D. M., Gregory, D. M.,
Thiyagarajan, P., Botto, R. E., Meredith, S. C., and Lynn, D. G. (2000) Structure of
the -Amyloid(10-35) Fibril. J. Am. Chem. Soc. 122, 7883-7889.
Byler, D. M., and Susi, H. (1986) Examination of the secondary structure of proteins
by deconvolved FTIR spectra, Biopolymers 25, 469-487.
201
Canet, D., Sunde, M., Last, A. M., Miranker, A., Spencer, A., Robinson, C. V., and
Dobson, C. M. (1999) Mechanistic studies of the folding of human lysozyme and the
origin of amyloidogenic behavior in its disease-related variants, Biochemistry 38,
6419-6427.
Carter, D. C., and Ho, J. X. (1994) Structure of serum albumin, Adv. Protein Chem.
45, 153-203.
Chatterjee, A., Moulik, S. P., Majhi, P. R., and Sanyal, S. K. (2002) Studies on
surfactant-biopolymer interaction. I. Microcalorimetric investigation on the
interaction of cetyltrimethylammonium bromide (CTAB) and sodium dodecyl
sulfate (SDS) with gelatin (Gn), lysozyme (Lz), and deoxyribonucleic acid (DNA),
Biophys. Chem. 98, 313-327.
Chen, L., Hodgson, K. O., and Doniach, S. (1996) A lysozyme folding intermediate
revealed by solution X-ray scattering, J. Mol. Biol. 261, 658-672.
Chen, S. H., and Teixeira, J. (1986) Structure and fractal dimension of protein-
detergent complexes, Phys. Rev. Lett. 57, 2583-2586.
Chen, S.-H., and Lin, T.-L. (1987) in Neutron Scattering (Price, D. L., and Skold,
K., Eds.) pp 489-543, Academic Press, San Diego.
Chiti, F., and Dobson, C. M. (2006) Protein misfolding, functional amyloid, and
human disease, Annu. Rev. Biochem. 75, 333-366.
Chiti, F., Taddei, N., Bucciantini, M., White, P., Ramponi, G., and Dobson, C. M.
(2000) Mutational analysis of the propensity for amyloid formation by a globular
protein, EMBO J. 19, 1441-1449.
Chiti, F., Webster, P., Taddei, N., Clark, A., Stefani, M., Ramponi, G., and Dobson,
C. M. (1999) Designing conditions for in vitro formation of amyloid protofilaments
and fibrils, P. Nat. Acad. Sci. 96, 3590-3594.
Cole, R., and Loria, J. P. (2002) Evidence for flexibility in the function of
ribonuclease A. Biochemistry 41, 6072-6081.
Colvin, J. R. (1952) Size and shape of lysozyme, Can. J. Chem. 30, 831-834.
Czapinska, H., and Otlewski, J. (1999) Structural and energetic determinants of the
S1-site specificity in serine proteases. Eur. J. Biochem. 260, 571-595.
202
Daban, J. R., Samso, M., and Bartolome, S. (1991) Use of Nile red as a fluorescent
probe for the study of the hydrophobic properties of protein-sodium dodecyl sulfate
complexes in solution, Anal. Biochem. 199, 162-168.
Daoud, M., Cotton, J. P., Farnoux, B., Jannink, G., Sarma, G., Benoit, H., Duplessix,
C., Picot, C., and De Gennes, P. G. (1975) Solutions of flexible polymers. Neutron
experiments and interpretation. Macromolecules 8, 804-18.
Diamond, R. (1974) Real-space refinement of the structure of hen-egg-white
lysozyme, J. Mol. Biol. 82, 371-391.
Ding, F., and Dokholyan, N. V. (2005) Simple but predictive protein models, Trends
Biotechnol. 23, 450-455.
Dinner, A. R., Sali, A., Smith, L. J., Dobson, C. M., and Karplus, M. (2000)
Understanding protein folding via free-energy surfaces from theory and experiment,
Trends Biochem. Sci. 25, 331-339.
Dobson, C. M. (1994) Solid evidence for molten globules, Curr. Biol. 4, 636-640.
Dobson, C. M. (1999) Protein misfolding, evolution, and disease, Trends Biochem.
Sci. 24, 329-332.
Dokholyan, N. V. (2006) Studies of folding and misfolding using simplified models,
Curr. Opin. Struct. Biol. 16, 79-85.
Duan, Y., and Kollman, P. A. (1998) Pathways to a protein folding intermediate
observed in a 1-ís simulation in aqueous solution, Science 282, 740-744.
Dugas, H. (1996) Bioorganic Chemistry, 3rd. ed., Springer Verlag, New York.
Dunbar, J., Yennawar, H. P., Banerjee, S., Luo, J., and Farber, G. K. (1997) The
effect of denaturants on protein structure. Protein Sci. 6, 1727-1733.
Durchschlag, H., and Zipper, P. (1996) Comparative determination of structural
parameters and conformational changes of proteins by small-angle scattering,
crystallography and hydrodynamic analysis, J. Mol. Struct. 383, 223-229.
Durchschlag, H., Hefferle, T., and Zipper, P. (2003) Comparative investigations of
the effects of X- and UV-irradiation on lysozyme in the absence or presence of
additives, Radiat. Phys. Chem. 67, 479-486.
Eastoe, J., and Vesperinas, A. (2005) Self-assembly of light-sensitive surfactants.
Soft Matter 1, 338-347.
203
Egan, R., Michel, H. O., Schlueter, R., and Jandorf, B. J. (1957) Physicochemical
investigation of the chymotrypsins. II. On the mechanism of dimerization of
chymotrypsin, Arch. Biochem. Biophys. 66, 366-373.
Eisenberg, H. (1976) Biological Macromolecules and Polyelectrolytes in Solution,
London: Oxford Univ. Press.
Eisenberg, H. (1981) Forward scattering of light, X-rays and neutrons. Q. Rev.
Biophys. 14, 141-172.
Everett, D. H. (1986) Some thermodynamic aspects of aggregation phenomena with
special reference to micellization. Colloid Surface 21, 41-53.
Fabian, H., and Mantele, W. (2002) in Handbook of Vibrational Spectroscopy
(Chalmers, J. M., and Griffiths, P. R., Eds.) pp 3399-3424, Wiley, New York.
Feigin, L. A., and Svergun, D. I. (1987) Structure Analysis by Small-Angle X-ray
and Neutron Scattering, Plenum Press, New York.
Firestone, M. P., Delevie, R., and Rangarajan, S. K. (1983) On one dimensional
nucleation and growth of living polymers. 1. Homogenous nucleation. J. Theor.
Biol. 104, 535-552.
Fujisawa, T., Ueki, T., Inoko, Y., and Kataoka, M. (1987) X-ray scattering from a
troponin C solution and its interpretation with a dumbbell-shaped-molecule model,
J. Appl. Crystallogr. 20, 349-355.
Fulop, L. Z., M., Datki, Z., Soos, K., and Penke, B. (2004) -Amyloid-derived
Pentapeptide RIIGLa Inhibits A (1-42) Aggregation and Toxicity. Biochem. and
Biophys. Res. Comm. 324, 64-69.
Gallas, J. M., Littrell, K. C., Seifert, S., Zajac, G. W., and Thiyagarajan, P. (1999)
Solution structure of copper ion-induced molecular aggregates of tyrosine melanin.
Biophys. J. 77, 1135-42.
Gilquin, B., Guilbert, C., and Perahia, D. (2000) Unfolding of hen egg lysozyme by
molecular dynamics simulations at 300 K: Insight into the role of the interdomain
interface, Proteins 41, 58-74.
Gimel, J. C., and Brown, W. (1996) A light scattering investigation of the sodium
dodecyl sulfate-lysozyme system, J. Chem. Phys. 104, 8112-8117.
204
Glatter, O. (1977) A new method for the evaluation of small angle scattering data, J.
Appl. Crystallogr. 10, 415-421.
Glatter, O. (1979) The interpretation of real-space information from small-angle
scattering experiments. J. App. Cryst. 12, 166-175.
Glenner, G. G. (1980) Amyloid deposits and amyloidosis. I. The b-fibrilloses, N.
Engl. J. Med. 302, 1283-1292.
Glenner, G. G., Wong, C. W. (1984) Alzheimer's Disease: initial report of the
purification and characterization of a novel cerebrovascular amyloid protein.
Biochem. Biophys. Res. Commun. 120.
Glinka, C. J., Barker, J. G., Hammouda, B., Krueger, S., Moyer, J. J., and Orts, W. J.
(1998) The 30 m small-angle neutron scattering instruments at the National Institute
of Standards and Technology, J. Appl. Crystallogr. 31, 430-445.
Goldsbury, C. S., Wirtz, S., Muller, S. A., Sunderji, S., Wicki, P., Aebi, U., and
Frey, P. (2000) Studies on the in vitro assembly of A 40: Implications for the search
for A fibril formation inhibitors. J. Struct. Biol. 130, 217-231.
Goldsbury, C., Kistler, J., Aebi, U., Arvinte, T., and Cooper, G. J. S. (1999)
Watching Amyloid Fibrils Grow by Time-lapse Atomic Force Microscopy. J. Mol.
Biol. 285, 33-39.
Gosal, W. S., Myers, S. L., Radford, S. E., Thomson, N. H. (2006) Amyloid Under
the Atomic Force Microscope. Protein Pept. Lett. 13, 261-270.
Griko, Y. V., Freire, E., Privalov, G., van Dael, H., and Privalov, P. L. (1995) The
unfolding thermodynamics of c-type lysozymes: A calorimetric study of the heat
denaturation of equine lysozyme, J. Mol. Biol. 252, 447-459.
Guijarro, J. I., Sunde, M., Jones, J. A., Campbell, I. D., and Dobson, C. M. (1998)
Amyloid Fibril Formation by an SH3 Domain. Biochemistry 95, 4224-4228.
Guinier, A., and Fournet, G. (1955) Small Angle Scattering of X-ray, Wiley, New
York.
Hamill, A. C., and Lee, C. T., Jr. (in prep) Control of Ribonuclease A structure
through the use of photoresponsive surfactants.
Hamill, A. C., and Lee, T. C., Jr. (in prep) -amyloid peptide (1-40) fibril growth
influenced by photosurfactant.
205
Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2005) Probing lysozyme
conformation with light reveals a new folding intermediate, Biochemistry 44, 15139-
15149.
Hamill, A. C., Wang, S.-C., and Lee, C. T., Jr. (2007) Solution Structure of an
Amyloid-Forming Protein During Photoinitiated Hexamer-Dodecamer Transitions
Revealed Through Small-Angle Neutron Scattering, Biochemistry 46, 7694-7705.
Han, Y.-P., and Garner, W. L. (2006) Alpha 1 antichymotrypsin is a physiological
inhibitor of proMMP-9 activation by human skin.
Hansma, H. G., Hoh, J. H. (1994) Biomolecular imaging with the atomic force
microscope. Annu. Rev. Biophys. Biomol. Struct. 23, 115-139.
Harper, J. D., Lieber, C. M., and Lansbury, P. T. (1997) Atomic force microscopic
imaging of seeded fibril formation and fibril branching by the Alzheimer's amyloid-
beta protein. Chem. Biol. 4, 9851-959.
Harper, J. D., Wong, S. S., Lieber, C. M., and Lansbury, P. T., Jr. . (1999) Assembly
of A Amyloid Protofibrils: An in Vitro Model for a Possible Early Event in
Alzheimer's Disease. Biochemistry 38, 8972-8980.
Harrington, W. F., and Schellman, J. A. (1956) Evidence for the instability of
hydrogen-bonded peptide structures in water, based on studies of ribonuclease and
oxidized ribonuclease. Compt. rend. trav. lab. Carlsberg, Ser. chim. 30, 21-43.
Harris, J. R. (2002) In vitro Fibrillogenesis of the Amyloid (1-42) Peptide;
Cholesterol Potentiation and Aspirin Inhibition. Micron 33, 609-626.
Hayashita, T., Kurosawas, T., Miyata, T., Tanaka, K., and Igawa, M. (1994) Effect
of structural variation within cationic azosurfactant upon photoresponsive function
in aqueous solution, Colloid Polym. Sci. 272, 1611-1619.
Heller, W. T., Krueger, J. K., and Trewhella, J. (2003) Further insights into
calmodulin-myosin light chain kinase interaction from solution scattering and shape
restoration, Biochemistry 42, 10579-10588.
Hess, G. P. (1971) Chymotrypsin-chemical properties and catalysis, in The Enzymes
(Boyer, P. D., Ed.) pp 213-248, Academic Press, New York.
Holmberg, K., Jonsson, B., Kronberg, B., and Lindman, B. (2003) Surfactant and
Polymers in Aqueous Solution, 2nd Edition, John Wiley and Sons, Chichester, UK.
206
Horwich, A. (2002) Protein aggregation in disease: A role for folding intermediates
forming specific multimeric interactions, J. Clin. Invest. 110, 1221-1232.
Houry, W. A., and Harold, A. S. (1996) Nature of the Unfolded State of Evidence A:
Effect of Cis-Trans X-Pro Peptide Bond Isomerization. Biochemistry 35, 11719-
11733.
Hunt, J. F., McCrea, P. D., Zaccai, G., and Engelman, D. M. (1997) Assessment of
the aggregation state of integral membrane proteins in reconstituted phospholipid
vesicles using small angle neutron scattering, J. Mol. Biol. 273, 1004-1019.
Ibel, K., May, R. P., Kirschner, K., Szadkowski, H., Mascher, E., and Lundahl, P.
(1990) Protein-decorated micelle structure of sodium-dodecyl-sulfate - protein
complexes as determined by neutron scattering. Eur. J. Biochem. 190, 311-318.
Ismail, A. A., Mantsch, H. H., and Wong, P. T. (1992) Aggregation of
chymotrypsinogen: portrait by infrared spectroscopy. Biochim. Biophys. Acta 1121,
183-188.
Jaikaran, E. T. A. S., Higham, C. E., Serpell, L. C., Zurdo, J., Gross, M., Clark, A.,
and Fraser, P. E. (2001) Identification of a novel human islet amyloid polypeptide b-
sheet domain and factors influencing fibrillogenesis, J. Mol. Biol. 308, 515-525.
Jansen, R., Dzwolak, W., and Winter, R. (2005) Amyloidogenic Self-Assembly of
Insulin Aggregates Probed by High Resolution Atomic Force Microscopy. Biophys.
J. 88, 1344-1353.
Jeng, U. S., Lin, T.-L., Lin, J. M., and Ho, D. L. . (2006) Contrast Variation SANS
for the Solution Structure of the -amyloid Peptide 1-40 Influenced by SDS
Surfactants. Physica B 385-386, 865-867.
Jones, J. A., Wilkins, D. K., Smith, L. J., and Dobson, C. M. (1997) Characterization
of protein unfolding by NMR diffusion measurements, J. Biomol. NMR 10, 199-203.
Kad, N. M., Myers, S. L., Smith, D. P., Smith, D. A., Radford, S. E., and Thomson,
N. H. (2003) Hierarchical Assembly of 2-Microglobulin Amyloid In Vitro
Revealed by Atomic Force Microscopy. J. Mol. Biol. 330.
Kamatari, Y. O., Konno, T., Kataoka, M., and Akasaka, K. (1998) The methanol-
induced transition and the expanded helical conformation in hen lysozyme, Protein
Sci. 7, 681-688.
Kanek, I., Tutumi, S. (1997) Conformations of -Amyloid in Solution. J.
Neurochem. 68, 438.
207
Kayed, R., Head, E., Thompson, J. L., Mcintire, T. M., Milton, S. C., Cotman, C.
W., and
Glabe, C. G. (2003) Common Structure of Soluble Amyloid Oligomers Implies
Common Mechanism of Pathogenesis. Science 300, 486-489.
Kelly, J. W. (1998) The alternative conformations of amyloidogenic proteins and
their multi-step assembly pathways. Curr. Opin. Struc. Biol. 8, 101-106.
Kiefhaber, T. (1995) Kinetic traps in lysozyme folding. P. Nat. Acad. Sci. 92, 9029-
9033.
Kirkitadze, M. D., Bitan, G., and Teplow, D. B. (2002) Paradigm shifts in
Alzheimer’s disease and other neurodegenerative disorders: the emerging role of
oligomeric assemblies, J. Neurosci. Res. 69, 567-577.
Kirkwood, J. G. (1996) The general theory of irreversible processes in solutions of
macromolecules, J. Polym. Sci. Pol. Phys. 34, 1-14.
Kitano, H., Maeda, Y., and Okubo, T. (1989) Kinetic study of the effects of
solvation on the dimerization process of a-chymotrypsin. Biophys. Chem. 33, 47-54.
Klein, W. L., Stine, W. B., and Teplow, D. B. (2004) Small assemblies of
unmodified amyloid b-protein are the proximate neurotoxin in Alzheimer’s disease,
Neurobiol. Aging 25, 569-580.
Kline, S. R. (2006) Reduction and analysis of SANS and USANS data using IGOR
Pro. J. Appl. Cryst. 39, 895.
Klug, G. M. J. A., Losic, D., Subasinghe, S. S., Aguilar, M-I., Martin, L. L., and
Small, D. H. . (2003) -Amyloid Protein Oligomers Induced by Metal Ions and Acid
pH are Distinct from those Generated by Slow Spontaneous Ageing at Neutral pH
Eur. J. Biochem. 270, 4282-4293.
Klunk, W. E., Debanth, M. L., and Pettegrew, J. W. (1994) Congo Red Populates
Partially Unfolded States of an Amyloidogenic Protein to Enhance Aggregation and
Amyloid Fibril Formation. Neurobiol. Aging 15, 691-698.
Klunk, W. E., Jacob, R. F., and Mason, R. P. (1999) Quantifying amyloid by Congo
red spectral shift assay, Methods Enzymol. 309, 285-305.
Koren, R., and Hammes, G. G. (1976) A kinetic study of protein-protein
interactions, Biochemistry 15, 1165-71.
208
Krebs, M. R. H., Bromley, E. H. C., Rogers, S. S., and Donald, A. M. (2005) The
mechanism of amyloid spherulite formation by bovine insulin, Biophys. J. 88, 2013-
2021.
Krebs, M. R., MacPhee, C. E., Miller, A. F., Dunlop, I. E., Dobson, C. M., and
Donald, A. M. (2004) The Formation of Spherulites by Amyloid Fibrils of Bovine
Insulin. P. Nat. Acad. Sci. USA 101, 14420-14424.
Krueger, S., Ho, D., and Tsai, A. (2006) Small-Angle Neutron Scattering as a Probe
for Protein Aggregation at Many Length Scales, in Misbehaving Proteins. Protein
(Mis)Folding, Aggregation, and Stability (Murphy, R. M., Tsai, A. M., Ed.) pp 125-
146, Springer, New York.
Kusumoto, Y., Lomakin, A., Teplow, D. B., and Benedek, G. B. (1998) Temperature
dependence of amyloid -protein fibrillizaiton. P. Natl. Acad. Sci. USA 95, 12277-
12282.
Lad, M. D., Ledger, V. M., Briggs, B., Green, R. J., and Frazier, R. A. (2003)
Analysis of the SDS-lysozyme binding isotherm, Langmuir 19, 5098-5103.
Ladwig Glenn, P., Robson Martin, C., Liu, R., Kuhn, M. A., Muir David, F., and
Schultz Gregory, S. (2002) Ratios of activated matrix metalloproteinase-9 to tissue
inhibitor of matrix metalloproteinase-1 in wound fluids are inversely correlated with
healing of pressure ulcers, Wound Repair Regen. 10, 26-37.
Le Ny, A.-L. M. a. L., C. T., Jr. (2006) Photoreversible DNA Condensation Using
Light-Responsive Surfactants. J. Am. Chem. Soc. 128, 6400-6408.
Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2004) Photoreversible viscosity
changes and gelation in mixtures of hydrophobically modified polyelectrolytes and
photosensitive surfactants, Macromol. 37, 5397-5405.
Lee, C. T., Jr., Smith, K. A., and Hatton, T. A. (2005) Photocontrol of protein
folding: The interaction of photosensitive surfactants with bovine serum albumin,
Biochemistry 44, 524-536.
Lee, P. K. J., and Jirgensons, B. (1971) Conformational transitions of
immunoglobulin fragment Fe(t) incited by alkyl sulfates of various hydrophobic
chain lengths, Biochim. Biophys. Acta 229, 631-641.
Lehmann, M. S., and Zaccai, G. (1984) Neutron Small-Angle Scattering Studies of
Ribonuclease in Mixed Aqueous Solutions and Determination of the Preferentially
Bound Water. Biochemistry 23, 1939-1942.
209
Leu, L. F., Kuo, Y. M., Roher, A. E., Brachova, L., Shen, Y., Sue, L., Beach, T.,
Kurth, J. H., Rydek, R. E., and Rogers, J. (1999) Soluble Amyloid beta Peptide
Concentration as a Predictor of Synaptic Change in Alzheimer's Disease. Am. J.
Pathol. 155, 853-962.
LeVine, H. (2002) 4,4'-Dianilino-1,1'-binaphthyl-5,5'-disulfonate: report on no-beta-
sheet conformers of Alzheimer's peptide A (1-40). Arch. Biochem. Biophys. 404,
106-115.
Liu, H., Yang, W., and Chen, J. (1998) Effects of surfactants on emulsification and
secondary structure of lysozyme in aqueous solutions, Biochem. Eng. J. 2, 187-196.
Mackay, R. A., Letts, K., and Jones, C. (1977) in Micellization, Solubilization, and
Microemulsions
Marone, P. A., Thiyagarajan, P., Wagner, A. M., and Tiede, D. M. (1998) The
association state of a detergent-solubilized membrane protein measured during
crystal nucleation and growth by small-angle neutron scattering, J. Cryst. Growth
191, 811-819.
Matagne, A., Radford, S. E., and Dobson, C. M. (1997) Fast and slow tracks in
lysozyme folding: Insight into the role of domains in the folding process, J. Mol.
Biol. 267, 1068-1074.
Matthews, C. R. (1993) Pathways of protein folding, Ann. Rev. Biochem. 62, 653-
683.
McClean, C. A., Cherny, R. A., Fraser, F. W., Fuller, S. J., Smith, M. J.,
Vbeyreutheer, K., Bush, A. I., and Masters, C. L. (1999) Soluble pool of A amyloid
as a determinant of severity of neurodegeneration in Alzheimer's disease. Ann.
Neurol. 46, 860-866.
Merlino, A., Vitagliano, L., Ceruso, M. A., Di Nola, A., and Mazzarella, L. (2002)
Global and local motions in ribonuclease A: a molecular dynamics study.
Biopolymers 65, 274-283.
Morozova-Roche, L. A., Zurdo, J., Spencer, A., Noppe, W., Receveur, V., Archer,
D. B., Joniau, M., and Dobson, C. M. (2000) Amyloid fibril formation and seeding
by wild-type human lysozyme and its disease-related mutational variants, J. Struct.
Biol. 130, 339-351.
210
Moses, M. A., Marikovsky, M., Harper, J. W., Vogt, P., Eriksson, E., Klagsbrun, M.,
and Langer, R. (1996) Temporal study of the activity of matrix metalloproteinases
and their endogenous inhibitors during wound healing, J. Cell. Biochem. 60, 379-
386.
Nettleton, E. J., Tito, P., Sunde, M., Bouchard, M., Dobson, C. M., and Robinson, C.
V. (2000) Characterization of the oligomeric states of insulin in self-assembly and
amyloid fibril formation by mass spectrometry, Biophys. J. 79, 1053-1065.
Nicoli, D. F., and Benedek, G. B. (1976) Study of thermal denaturation of lysozyme
and other globular proteins by light scattering spectroscopy, Biopolymers 15, 2421-
2437.
Nilsberth, C., Westllind-Danielsson, A., Eckman, C. B., Condron, M. M., Axelman,
K., Forseel, C., Stenh, C., Luthman, J., Teplow, D. B., Younklin, S. G., Naslund, J.,
and Lannfelt, L. (2001) The 'Arctic' APP mutation (E693G) causes Alzheimer's
disease by enhanced A protofibril formation. Nat. Neurosci. 4, 887-893.
Ninio, J., Luzzati, V., and Yaniv, M. (1972) Comparative small angle X-ray
scattering
studies on unacylated, acylated, and crosslinked Escherichia coli transfer RNAIVa1,
J. Mol. Biol. 71, 217-229.
Noeppert, A., Gast, K., Mueller-Frohne, M., Zirwer, D., and Damaschun, G. (1996)
Reduced-denatured ribonuclease A is not in a compact state. FEBS Lett. 380, 179-
82.
Oleinikova, A., Sasisanker, P., and Weingaertner, H. (2004) What Can Really Be
Learned from Dielectric Spectroscopy of Protein Solutions? A Case Study of
Ribonuclease A. J. Phys. Chem. B 108, 8467-8474.
Pal, S. K., Peon, J., and Zewail, A. H. (2002) Ultrafast surface hydration dynamics
and expression of protein functionality: a-Chymotrypsin, P. Nat. Acad. Sci 99,
15297-15302.
Pallares, I., Vendrell, J., Aviles, F. X., and Ventura, S. (2004) Amyloid fibril
formation by a partially structured intermediate state of a-chymotrypsin, J. Mol.
Biol. 342, 321-331.
Pandit, M. W., and Rao, M. S. N. (1975) Self-association of proteins. Self-
association of R-chymotrypsin at its isoelectric point in buffer solutions of ionic
strength 0.1, Biochemistry 14, 4106-4110.
211
Paz Andrade, M. I., Boitard, E., Saghal, M. A., Manley, P., Jones, M. N., and
Skinner, H. A. (1981) Enthalpy of interaction of ribonuclease A and n-alkyl
sulphates in aqueous solution. J. Chem. Soc., Faraday T. 77, 2939-2948.
Pedersen, J. S., Hansen, S., and Bauer, R. (1994) The aggregation behavior of zinc-
free insulin studied by small-angle neutron scattering, Eur. Biophys. J. 22, 379-389.
Pepys, M. B., Hawkins, P. N., Booth, D. R., Vigushin, D. M., Tennent, G. A.,
Soutar, A. K., Totty, N., Nguyen, O., Blake, C. C. F., et al. (1993) Human lysozyme
gene mutations cause hereditary systemic amyloidosis, Nature 362, 553-557.
Perkins, S. J., Smith, K. F., Kilpatrick, J. M., Volanakis, J. E., and Sim, R. B. (1993)
Modeling of the serine-proteinase fold by X-ray and neutron scattering and
sedimentation analyses: occurrence of the fold in factor D of the complement
system, Biochem. J. 295 (Part 1), 87-99.
Petkova, A. T., Ishii, Y., Balbach, J. J., Antzutkin, O. N., Leapman, R. D., Delaglio,
F., and Tycko, R. (2002) A structural model for Alzheimer's -amyloid fibrils based
on experimental constraints from solid state NMR. P. Nat. Acad. Sci. 99, 16742-
16747.
Porod, G. (1982) in Small Angle X-ray Scattering (Glatter, O., Ed.) pp 17-51,
Academic Press, London.
Powers, E. T., Powers, D. L. (2006) The kinetics of Nucleated Polymerization at
High Concentration: Amyloid Fibril Formation Near and Above the "Supercritical
Concetration". Biophys. J. 91, 122-132.
Przybycien, T. M., and Bailey, J. E. (1989) Aggregation kinetics in salt-induced
protein precipitation. AIChE J. 35, 1779-1790.
Puchtler, H., and Sweat, F. (1965) Congo red as a stain for fluorescence microscopy
of amyloid, J. Histochem. Cytochem. 13, 693-694.
Quijano, F. A. R., Morrow, D., Wise, B. M., Brancia, F. L., and Goux, W. J. (2006)
Prediction of Nucleation Sequences form Amyloidogenic Propensities of Tau-
Related Peptides. Biochemistry 45, 4638-4652.
Radford, S. E., Dobson, C. M., and Evans, P. A. (1992) The folding of hen lysozyme
involves partially structured intermediates and multiple pathways, Nature 358, 302-
307.
Raines, R. T. (1998) Ribonuclease A. Chem. Rev. 98, 1045-1065.
212
Ramakrishna, T., and Pandit, M. W. (1992) Involvement of active site in self-
association of proteins: a case of alpha-chymotrypsin, Indian J. Biochem. Biophys.
29, 192-200.
Rao, M. S. N., and Kegeles, G. (1958) An ultracentrifuge study of the
polymerization of R-chymotrypsin, J. Am. Chem. Soc. 80, 5724-5729.
Roger, A. E., Chaney, M. O., Kuo, Y. M., Webster, S. D., Stine, W. B., . (1996)
Morphology and Toxicity of A -(1-42) Dimer Derived from neuritic and Vascular
Amyloid deposits of Alzheimer's Disease. J. Biol. Chem 271, 20631-20635.
Salahuddin, A., and Tanford, C. (1970) Thermodynamics of the denaturation of
ribonuclease by guanidine hydrochloride. Biochemistry 9, 1342-1347.
Salo, T., Makela, M., Kylmaniemi, M., Autio-Harmainen, H., and Larjava, H.
(1994) Expression of matrix metalloproteinase-2 and -9 during early human wound
healing, Lab. Invest. 70, 176-182.
Sambashivan, S., Liu, Y., Sawaya, M. R., Gingery, M., and Eisenberg, D. (2005)
Amyloid-like fibrils of ribonuclease A with three-dimensional domain-swapped and
native-like structure, Nature 437, 266-269.
Scheraga, H. A., Wedemeyer, W. J., and Welker, E. (2001) Bovine pancreatic
ribonuclease A: oxidative and conformational folding studies. Methods Enzymol.
341, 189-221.
Segel, D. J., Bachmann, A., Hofrichter, J., Hodgson, K. O., Doniach, S., and
Kiefhaber, T. (1999) Characterization of transient intermediates in lysozyme folding
with time-resolved small-angle X-ray scattering, J. Mol. Biol. 288, 489-499.
Seilheimer, B., Bohrmann, B., Bondolfi, L., Mueller, F., Stueber, D., and Doebeli,
H. (1997) Alzheimer's beta amyloid peptide correlates with a distinct fiber
morphology. J. Struct. Biol. 119, 59-71.
Semisotnov, G. V., Rodionova, N. A., Razgulyaev, O. I., Uversky, V. N., Gripas, A.
F., and Gil’manshin, R. I. (1991) Study of the “molten globule” intermediate state in
protein folding by a hydrophobic fluorescent probe, Biopolymers 31, 119-128.
Serio, T. R., Cashikar, A. G., Kowal, A. S., Sawicki, G. J., Moelehi, J. J., Serpell, L.,
Arnsdorf, M. F., and Lindquist, S. L. (2000) Nucleated Conformational Conversion
and the Replication of Conformational Information by a Prion Determinant. Science
289, 1317-1321.
213
Serpell, L. C., Sunde, M., Benson, M. D., Tennent, G. A., Pepys, M. B., and Fraser,
P. E. (2000) The protofilament substructure of amyloid fibrils. J. Mol. Biol. 300,
1033-1039.
Shang, T., Smith, K. A., and Hatton, T. A. (2003) Photoresponsive surfactants
exhibiting unusually large, reversible surface tension changes under varying
illumination conditions, Langmuir 19, 10764-10773.
Shang, T., Smith, K. A., and Hatton, T. A. (2003) Photoresponsive Surfactants
Exhibiting Unusually Large, Reversible Surface Tension Changes under Varying
Illumination Conditions. Langmuir 19, 10764-10773.
Shih, P., Holland, D. R., and Kirsch, J. F. (1995) Thermal stability determinants of
chicken egg-white lysozyme core mutants: hydrophobicity, packing volume, and
conserved buried water molecules. Protein Sci. 4, 2050-2062.
Shilton, B. H., McDowell, J. H., Smith, W. C., and Hargrave, P. A. (2002) The
solution structure and activation of visual arrestin studied by small-angle X-ray
scattering, Eur. J. Biochem. 269, 3801-3809.
Soreghan, B., Kosmoski, J., and Glabe, C. (1994) Surfactant properties of
Alzheimer’s A peptides and the mechanism of amyloid aggregation. J. Biol. Chem.
269, 28551-28554.
Spreti, N., Alfani, F., Cantarella, M., D'Amico, F., Germani, R., and Savelli, G.
(1999) a-Chymotrypsin superactivity in aqueous solutions of cationic surfactants. J.
Mol. Catal. B-Enzym. 6, 99-110.
Spreti, N., Di Profio, P., Marte, L., Bufali, S., Brinchi, L., and Savelli, G. (2001)
Activation and stabilization of a-chymotrypsin by cationic additives. Eur. J.
Biochem. 268, 6491-6497.
Srisailam, S., Thallampuranam, K., Krishnaswamy, S., Rajalingam, D., Kathir
Karuppanan, M., Sheu, H.-S., Jan, F.-J., Chao, P.-C., and Yu, C. (2003) Amyloid-
like fibril formation in an all beta barrel protein. Partially structured intermediate
state(s) is a precursor for fibril formation, J. Biol. Chem. 278, 17701-17709.
Stenstam, A., Montalvo, G., Grillo, I., and Gradzielski, M. (2003) Small angle
neutron scattering study of lysozyme-sodium dodecyl sulfate aggregates, J. Phys.
Chem. B 107, 12331-12338.
Stine, W. B., Jr., Snyder, S. W., Ladror, U. S., Wade, W. S., Miller, M. F., Perun, T.
J., Holzman, T. F., and Krafft, G. A. (1996) The nanometer-scale structure of
amyloid-b visualized by atomic force microscopy. J. Protein Chem. 15, 193-203.
214
Stuhrmann, H. B., and Fuess, H. (1976) Neutron small-angle scattering study of hen
egg-white lysozyme, Acta Crystallogr., Sect. A: Found. Crystallogr. 32 (part 1), 67-
74.
Sunde, M., and Blake, C. (1997) The structure of amyloid fibrils by electron
microscopy and X-ray diffraction. Adv. Protein Chem. 50, 123-159.
Susi, H., and Byler, D. M. (1986) Resolution-enhanced Fourier transform infrared
spectroscopy of enzymes, Methods Enzymol. 130, 290-311.
Svensson, M., Hakansson, A., Mossberg, A.-K., Linse, S., and Svanborg, C. (2000)
Conversion of alpha-Lactalbumin to a protein inducing apoptosis, P. Nat. Acad. Sci.
97, 4221-4226.
Svergun, D. I. (1992) Determination of the regularization parameter in indirect-
transform methods using perceptual criteria, J. Appl. Crystallogr. 25, 495-503.
Svergun, D. I., and Koch, M. H. J. (2003) Small-angle scattering studies of
biological macromolecules in solution, Rep. Prog. Phys. 66, 1735-1782.
Svergun, D. I., Petoukhov, M. V., and Koch, M. H. J. (2001) Determination of
domain structure of proteins from x-ray solution scattering. Biophys. J. 80, 2946-
2953.
Svergun, D. I., Richard, S., Koch, M. H. J., Sayers, Z., Kuprin, S., and Zaccai, G.
(1998) Protein hydration in solution: experimental observation by X-ray and neutron
scattering, P. Nat. Acad. Sci. 95, 2267-2272.
Tarlton, J. F., Vickery, C. J., Leaper, D. J., and Bailey, A. J. (1997) Postsurgical
wound progression monitored by temporal changes in the expression of matrix
metalloproteinase-9, Brit. J. Dermatol. 137, 506-516.
Timchenko, A. A., Ptitsyn, O. B., Dolgikh, D. A., and Fedorov, B. A. (1978) The
structure of ribonuclease in solution does not differ from its crystalline structure,
FEBS Lett. 88, 105-108.
Tito, P., Nettleton, E. J., and Robinson, C. V. (2000) Dissecting the Hydrogen
Exchange Properties of Insulin under Amyloid Fibril Forming Conditions: A Site-
specific Investigation by Mass Spectrometry, J. Mol. Biol. 303, 267-278.
Tjernberg, L. O., Pramanik, A., Jorling, S., Thyberg, P., Thyberg, J., Nordstedt, C.,
Berndt, K. D., Terenius, L., and Rigler, R. (1999) Amyloid beta-peptide
polymerization studied using fluorescence correlation spectroscopy. Chem. Biol. 6,
53-62.
215
Tranier, S., Mortier-Barriere, I., Ilbert, M., Birck, C., Iobbi-Nivol, C., Mejean, V.,
and Samama, J.-P. (2002) Characterization and multiple molecular forms of TorD
from Shewanella massilia, the putative chaperone of the molybdoenzyme TorA,
Protein Sci. 11, 2148-2157.
Turnell, W. G., and Finch, J. T. (1992) Binding of the dye congo red to the amyloid
protein pig insulin reveals a novel homology amongst amyloid-forming peptide
sequences. J. Mol. Biol 227, 1205-1223.
Uversky, V. N. (1993) Use of fast protein size-exclusion liquid chromatography to
study the unfolding of proteins which denature through the molten globule,
Biochemistry 32, 13288-13298.
Uversky, V. N., and Fink, A. L. (2004) Conformational constraints for amyloid
fibrillation: the importance of being unfolded. Biochim. Biophys. Acta 1698, 131-
153.
Uversky, V. N., Karnoup, A. S., Segel, D. J., Seshadri, S., Doniach, S., and Fink, A.
L. (1998) Anion-induced folding of Staphylococcal nuclease: Characterization of
multiple equilibrium partially folded intermediates, J. Mol. Biol. 278, 879-894.
Valleix, S., Drunat, S., Philit, J.-B., Adoue, D., Piette, J.-C., Droz, D., MacGregor,
B., Canet, D., Delpech, M., and Grateau, G. (2002) Hereditary renal amyloidosis
caused by a new variant lysozyme W64R in a French family, Kidney Int. 61, 907-
912.
Valstar, A., Brown, W., and Almgren, M. (1999) The lysozyme sodium dodecyl
sulfate system studied by dynamic and static light scattering, Langmuir 15, 2366-
2374.
Vanhoudt, J., Abgar, S., Aerts, T., and Clauwaert, J. (2000) A small-angle X-ray
solution scattering study of bovine R-crystallin, Eur. J. Biochem. 267, 3848-3858.
Velev, O. D., Kaler, E. W., and Lenhoff, A. M. (1998) Protein interactions in
solution characterized by light and neutron scattering: Comparison of lysozyme and
chymotrypsinogen, Biophys. J. 75, 2682-2697.
Ventura, S., Lacroix, E., and Serran, L. (2002) Insights into the Origin of the
Tendency of the PI3-SH3 Domain to form Amyloid Fibrils, J. Mol. Biol. 322, 1147-
1158.
Walsh, D. M., Lomakin, A., Benedek, G. B., Condron, M. M., and Teplow, D. B.
(1997) Amyloid -Protein Fibrillogenesis. J. Biol. Chem. 272, 22364-22372.
216
Wang, D., Bode, W., and Huber, R. (1985) Bovine chymotrypsinogen A. X-ray
crystal structure analysis and refinement of a new crystal form at 1.8 Å resolution. J.
Mol. Biol. 185, 595-624.
Wang, D., Dickson, D. W., Trojanowski, J. Q., and Lee, V. M.-Y. (1999) The Levels
of Soluble versus Insoluble Brain A Distinguish Alzheimer's Disease from Normal
and Pathological Aging Exp. Neurol. 158, 328-337.
Wang, S.-C., and Lee, C. T., Jr. (2007) Enhanced Enzymatic Activity Through
Photoreversible Conformation Changes. Biochemistry 46, 14557-14566.
Wang, S.-C., and Lee, C. T., Jr. (2006) Protein Secondary Structure Controlled with
Light and Photoresponsive Surfactants. J. Phys. Chem. B 110, 16117-16123.
Weckroth, M., Vaheri, A., Lauharanta, J., Sorsa, T., and Konttinen, Y. T. (1996)
Matrix metalloproteinases, gelatinase and collagenase, in chronic leg ulcers, J.
Invest. Dermatol. 106, 1119-1124.
Winzor, D. J., and Scheraga, H. A. (1964) Chemically reacting systems on
Sephadex. II. Molecular weights of monomers in rapid association equilibrium. J.
Phys. Chem. 68, 338-343.
Wlodawer, A., Svennson, L. A., Sjolin, L., and Gilliland, G. L. (1988) Structure of
phosphate-free ribonuclease A refined at 1.26 A. Biochemistry 27, 2705-2717.
Wong, S. S., Harper, J. D., Lansbury, P.T., and Lieber, C. M. (1998) Carbon
nanotube tips: high resolution probed for imaging biological systems. J. Am. Chem.
Soc 120, 603-604.
Wood, S. J., Maleef, B., Hart, T. & Wetzel, R. (1996) Physical, Morphological, and
Functional Differences between pH 5.8 and 7.4 aggregates of the Alzheimer's
amyloid peptide A . J . Mol. Biol. 256, 870-877.
Wysocki, A. B., Staiano-Coico, L., and Grinnell, F. (1993) Wound fluid from
chronic leg ulcers contains elevated levels of metalloproteinases MMP-2 and MMP-
9, J. Invest. Dermatol. 101, 64-68.
Yager, D. R., Zhang, L.-Y., Liang, H.-X., Diegelmann, R. F., and Cohen, I. K.
(1996) Wound fluids from human pressure ulcers contain elevated matrix
metalloproteinase levels and activity compared to surgical wound fluids, J. Invest.
Dermatol. 107, 743-748.
Yan, H., Saiani, A., Miller, A. (2007) in APS March Meeting, Denver, CO.
217
Yan, H., Saini, A., Gough, J. E., and Miller, A. F. (2006) Thermoreversible Protein
Hydrogel as Cell Scaffold. Biomacromolecules 7, 2776-2782.
Yang, J., Sho, Z. (1995) Recent advances in biological atomic force microscopy.
Micron 26.
Yang, L. (2005) The X21 SAXS instrument at NSLS for studying macromolecular
systems, Macromol. Res. 13, 538-541.
Yokota, A., Hirai, K., Miyauchi, H., Iimura, S., Noda, Y., Inoue, K., Akasaka, K.,
Tachibana, H., and Segawa, S. (2004) NMR characterization of three-disulfide
variants of lysozyme, C64A/C80A, C76A/C94A, and C30A/C115AsA marginally
stable state in folded proteins, Biochemistry 43, 6663-6669.
Yong, W., Lomakin, A., Kirkitadze, M. D., Teplow, D. B., Chen, S.-H., and
Benedek, G. B. (2002) Structure Determination of Micelle-like Intermediates in
Amyloid -protein Fibril Assembly by using Small Angle Neutron Scattering. P.
Nat. Acad. Sci. 99, 150-154.
Young, P. K., and Grinnell, F. (1994) Metalloproteinase activation cascade after
burn injury: a longitudinal analysis of the human wound environment, J. Invest.
Dermatol. 103, 660-664.
Zenkova, M. A. (2004) Artificial Nucleases, Vol. 13, Springer, Berlin.
Zhou, W., Islam, M. F., Wang, H., Ho, D. L., Yodh, A. G., Winey, K. I., and
Fischer, J. E. (2004) Small angle neutron scattering from single-wall carbon
nanotube suspensions. Evidence for isolated rigid rods and rod networks. Chem.
Phys. Lett. 384, 185-189.
Zurdo, J., Guijarro, J. I., Jimenez, J. L., Saibil, H. R., and Dobson, C. M. (2001)
Dependence on solution conditions of aggregation and amyloid formation by an SH3
domain, J. Mol. Biol. 311, 325-340.
Abstract (if available)
Abstract
A photoresponsive surfactant, azoTAB, is used to control protein structure. When azoTAB is combined with a protein, illumination with visible and UV light can be used to induce different protein conformational changes. A means to reversibly control the secondary and tertiary structure of two proteins, lysozyme and RNase A, has been developed. In the presence of azoTAB and under visible light illumination, the alpha-domain of lysozyme unfolds, forming a new folding intermediate. Lysozyme transitions from this intermediate back to its native state under UV light illumination. Similarly, in the presence of azoTAB and under visible light illumination, a swollen form of RNase A is observed, which also transitions back to its native state under UV light illumination. Additionally, a means to control the quaternary structure of alpha-chymotrypsin has been developed. The degree of self-association of lapha-chymotrypsin, which readily associates in aqueous solution, is controlled when combined with azoTAB and UV or visible light illumination. Under visible light, the associated form is a corkscrew hexamer. Under UV light, these corkscrews self-associate in a slightly offset manner, forming ropelike dodecamers. The dodecamers were found to be preamyloidal. A means to control the association of a well-known amyloid protein linked to Alzheimer s disease, amyloid-beta peptide, has been developed. When combined with azoTAB and illuminated with visible light, the association of amyloid-beta into fibrils is significantly delayed relative to the association of pure amyloid-beta. UV light can be used to trigger the fibril formation process. In all cases, Small Angle Neutron Scattering was the main tool used to investigate the various changes in protein structure.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Controlling the form-dynamics-function relationship of proteins with light illumination
PDF
Photocontrol of protein conformation and enzyme activity in the presence of light-responsive surfactants
PDF
Control the function, dynamics, aggregation of proteins with light illumination
PDF
Enhancement of biofuel enzyme activity and kinetics with azoTAB surfactants
PDF
Photocontrol of DNA compaction using light-responsive surfactants
PDF
The interaction of photo-responsive surfactants with biological macromolecules
PDF
Controlling membrane protein folding with light illumination and catanionic surfactant systems
PDF
Molecular dynamics simulation study of initial protein unfolding induced by the photo-responsive surfactants, azoTAB
PDF
Controlling membrane protein folding using photoresponsive surfactant
Asset Metadata
Creator
Hamill, Andrea C. (author)
Core Title
Photocontrol of protein conformation through the use of photoresponsive surfactants, investigated by small angle neutron scattering
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Chemical Engineering
Publication Date
03/06/2008
Defense Date
02/25/2008
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,protein,small angle neutron scattering,surfactant
Language
English
Advisor
Lee, C. Ted, Jr. (
committee chair
), Bau, Robert (
committee member
), Roberts, Richard W. (
committee member
)
Creator Email
ahamill@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m1043
Unique identifier
UC1119210
Identifier
etd-Hamill-20080306 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-24126 (legacy record id),usctheses-m1043 (legacy record id)
Legacy Identifier
etd-Hamill-20080306.pdf
Dmrecord
24126
Document Type
Dissertation
Rights
Hamill, Andrea C.
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
protein
small angle neutron scattering
surfactant