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
/
The folding, misfolding and refolding of membrane proteins and the design of a phosphatidylserine-specific membrane sensor
(USC Thesis Other)
The folding, misfolding and refolding of membrane proteins and the design of a phosphatidylserine-specific membrane sensor
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE FOLDING, MISFOLDING AND REFOLDING OF MEMBRANE PROTEINS AND
THE DESIGN OF A PHOSPHATIDYLSERINE-SPECIFIC MEMBRANE SENSOR
by
Yujin E. Kim
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOCHEMISTRY AND MOLECULAR BIOLOGY)
December 2009
Copyright 2009 Yujin E. Kim
ii
Dedication
To Minkyu Scott Lee.
iii
Acknowledgements
Above all others, I would like to express my sincere thanks to my advisor, Dr. Ralf
Langen, for providing me with the opportunity to do the work presented in this thesis,
and also for his support, helpful guidance, and encouragement over the years. I am
also thankful to my guidance and thesis committee members, Dr. Jeannie Chen, Dr.
Bob Chow, Dr. Ian Haworth and Dr. Frank Markland, for their support and
encouragement. I thank the current and past members of the Langen and Chen labs
for helpful suggestions and amiable acquaintanceship which greatly contributed to a
productive and enjoyable graduate experience, particularly Drs. Mario Isas, Sajith
Jayasinghe, and Ani Der-Sarkissian for the technical training I received, Dr. Martin
Margittai for interesting and helpful discussions, and Melania Apostolidou for
friendship. I am grateful to Diana Gegala for taking care of Everything Else. I thank
the members of the Chan lab for technical help and friendly discussions, especially
Dr. Jonah Chan for helpful suggestions and advice. And special thanks are due to
my family, especially my husband, who makes the Best Morning Coffee, for all the
fun, encouragement and moral support.
iv
Table of Contents
List of Figures
Abstract
Preface
Introduction
Chapter 1: Structure of annexin B12 in the Ca
2+
-dependent
membrane bound state
Chapter 2: Annexin B12 Ca
2+
-independent transmembrane state
Chapter 3: An engineered polarity sensitive biosensor for real-
time imaging of apoptosis and degeneration
Chapter 4: Structural Analysis of α-synuclein oligomers by site-
directed spin labeling
Chapter 5: Conclusions and Future Directions
References
v
vii
ix
1
22
41
65
97
120
125
v
List of Figures
Figure I-1: Crystal structure of annexin B12
Figure I-2: Structural organization of α-synuclein
Figure I-3: Structural states of α-synuclein
Figure I-4: Protein misfolding and aggregation
Figure I-5: Site-directed spin labeling
Figure I-6: EPR spectra characteristic of different R1 mobility
Figure I-7: Accessibility parameter Π
Figure 1-1: Membrane binding loops in annexin B12
Figure 1-2: EPR spectra and mobility of spin-labeled loop sites
Figure 1-3: Π(O
2
) and Π(NiEDDA) accessibilities
Figure 1-4: Plot of Φ-values
Figure 1-5: Summary of Ca
2+
- and membrane-binding loops
Figure 2-1: Model of annexin B12 membrane interaction
Figure 2-2: Mobility analysis of annexin B12 in solution and the
membrane-bound state
Figure 2-3: Accessibility of residues 251-273 of annexin B12 in
solution
Figure 2-4: Accessibility of annexin B12 in the membrane-inserted
state
Figure 2-5: R1 immersion depth of membrane-inserted annexin
B12
Figure 2-6: Ca
2+
-independent membrane binding of annexin B12
Figure 3-1: Structure-based design of annexin biosensor
Figure 3-2: Comparison of fluorescence intensities
Figure 3-3: pSIVA controls
5
10
13
14
16
18
20
24
28
31
33
37
43
50
52
56
58
62
68
76
78
vi
Figure 3-4: Application of pSIVA to live-cell imaging of apoptosis
in COS-7 cells
Figure 3-5: Separated channel images of COS-7 cells
Figure 3-6: Application of pSIVA to monitoring degeneration of
DRG neurons
Figure 3-7: Staining pattern of pSIVA on degenerating vs.
permeabilized axons
Figure 3-8: In vivo application of pSIVA
Figure 3-9: Rescue of neuronal degeneration visualized by pSIVA
Figure 3-10: Summary of the progression and/or recovery of early
apoptotic events in the neuron as detected by pSIVA
Figure 4-1: Structural organization of α-synuclein
Figure 4-2: EPR spectra measured for a time course
starting in 10% HFIP-H
2
O
Figure 4-3: Changes in circular dichroism spectra of α-syn∆CT44R1
Figure 4-4: EPR spectra of the transient intermediate at 22 h
Figure 4-5: Structural populations detected by TEM and A11
reactivity
Figure 4-6: Time course in neutral buffer
Figure 4-7: Separation of oligomeric α-syn∆CT by gel filtration
Figure 4-8: Circular dichroism spectra measured after gel filtration
Figure 4-9: Comparison of structures detected by EPR
81
82
84
87
88
91
94
100
106
107
109
111
113
114
115
117
vii
Abstract
The interaction between proteins and membranes underlies many important
biological processes. Proteins influence the structure and fluidity of membranes and
likewise, membranes promote structural reorganization in proteins, mutually
modifying their functions. In addition, protein-membrane interactions can lead to
protein misfolding and the pathogenesis of disease. Despite the importance of these
dynamic protein-membrane interactions, less is known about the structure of
membrane-bound proteins in comparison to proteins in solution. As an alternative to
conventional methods such as X-ray crystallography and NMR, we used electron
paramagnetic resonance (EPR) spectroscopy in combination with site-directed spin
labeling (SDSL) as the principle technique to study how different environmental
conditions can promote the structural transitions that occur in two different
membrane-binding proteins, annexin B12 (anxB12, anxXII) and α-synuclein (α-syn).
AnxB12 is a member of a family of structurally conserved proteins that are
abundantly expressed in most eukaryotic cells. The various proposed functions
include vesicle trafficking, membrane fusion, ion channel formation, and cell
signaling, indicating that membrane interactions are important for all annexin
functions. In order to gain a better understanding of the mechanism of membrane
binding and annexin function, the structures of two different membrane-bound states
of AnxB12 were studied. In addition, using the data obtained from these studies, we
engineered a polarity sensitive annexin-based biosensor applicable to the real-time
detection of apoptotic membrane changes in cultured cells and in vivo. The second
protein of interest, α-syn, is a natively unfolded cytosolic protein which is observed to
bind preferentially to small vesicles (of similar size to synaptic vesicles) and is
viii
localized to the presynaptic nerve terminals. The accumulation of misfolded
cytoplasmic aggregates of α-syn is central to the pathology Parkinson’s disease (PD).
Recent studies have suggested that the soluble oligomers of α-syn are a cause of
cell toxicity and disease pathogenesis. One of the proposed mechanisms by which
these oligomers are toxic is by membrane interaction and disruption. In order to gain
a better understanding of how misfolded α-syn oligomers are assembled and their
role in the disease relevant protein misfolding pathway, we studied the structures of
different oligomeric states of α-syn by EPR techniques.
ix
Preface
Several articles have been published based on the work presented in this thesis.
The data presented in Chapter 1 on the A-B loop regions has been included in
another manuscript currently in preparation for publication. The data presented in on
the D-E loop regions was included in the following publication:
Isas JM, Kim YE, Jao CC, Hegde PB, Haigler HT, Langen R. “Calcium- and
membrane-induced changes in the structure and dynamics of three helical
hairpins in annexin B12.” Biochemistry. 2005 Dec 20;44(50):16435-44.
The work done with phosphatidylserine-containing membranes presented in Chapter
2 on was published in the following article:
Kim YE, Isas JM, Haigler HT, Langen R. “A helical hairpin region of soluble
annexin B12 refolds and forms a continuous transmembrane helix at mildly
acidic pH.” J Biol Chem. 2005 Sep 16;280(37):32398-404.
Chapter 3 was reformatted and has been submitted for publication:
Kim YE, Chen J, Chan JR, Langen R. “Engineering a polarity sensitive
biosensor for time-lapse imaging of apoptosis.”
Chapter 4 is currently a work in progress, and will be prepared for publication:
Kim YE, Langen R. “Structural analysis of α-synuclein oligomers by site-
directed spin labeling”
1
Introduction
One of the more difficult challenges in structural biology is to elucidate the structure, and
consequently the mechanism of function, of membrane proteins. The interaction of
proteins with membranes plays essential roles in a wide variety of cellular processes
important for cell signaling, metabolism and homeostasis. Defects in membrane proteins
are often associated with disease. Additionally, protein-membrane interactions that lead
to membrane disruption have been proposed to play a role in the initiation and
progression of neurodegeneration in protein misfolding and amyloid diseases (Lashuel,
2005; Volles et al., 2001). As such, membrane proteins are often the target of
pharmacological substances, constituting the majority of all drug targets. Despite the
importance and abundance (~30% of all proteins are membrane proteins) of these
interactions, the membrane-bound structures are poorly understood in comparison to
solution structures. Knowing the structure of the membrane-bound protein is important to
the understanding of its physiological or pathological role. Specifically, it provides insight
into the mechanism for membrane binding, how these interactions alter the protein
structure, how protein interactions alter the membrane structure, and the resulting
functional consequences of these interactions. Of more than ~10,000 unique protein
structures which have been solved, only ~200 are membrane protein structures (White,
2009). This underrepresentation of membrane proteins is in part due to the difficulty of
studying membrane proteins by the conventional methods such as X-ray crystallography
due to their dependence on lipids which makes crystallization difficult, or NMR
spectroscopy due to size limitations which exclude the study of larger proteins and
membrane systems. In the examples of membrane protein structures which have been
solved by X-ray crystallography, the structures generally do not include the surrounding
2
membrane except for a few lipids which form part of a crystal unit’s cell. Also, for solution
NMR techniques, detergent micelles are often substituted in place of lipid membranes to
obtain membrane protein structures. However, differences in detergent environments
may alter the protein conformation. Furthermore, changes in the lipid composition and
environmental conditions may trigger structural transitions as well. As an alternative, we
used electron paramagnetic resonance (EPR) in combination with site-directed spin
labeling (SDSL), which allows the study of dynamic conformational changes of proteins
in their native lipid environments and in solution, thus providing a method which
facilitates the study of the various conformational distributions and structural dynamics
related to its function.
My studies were primarily focused on how different environmental conditions trigger the
interaction of proteins with other biological molecules, and how the dynamic structural
transitions which accompany these interactions might alter protein function. EPR
spectroscopy in combination with SDSL was used to study the structural transitions of
two different membrane-binding proteins, annexin B12 (anxB12, annexin XII) and α-
synuclein (α-syn). AnxB12 provides a unique model system for structural studies
because of its ability to interchangeably take up different solution and membrane-bound
states, which can be manipulated in vitro by changes in membrane composition, pH, and
Ca
2+
concentration. To elucidate the mechanisms by which different solution and lipid
environments drive the different modes of membrane binding, I studied the structures of
two of these membrane-bound states of anxB12, the Ca
2+
-dependent surface-bound
trimer (Chapter 1) and the Ca
2+
-independent transmembrane monomer (Chapter 2).
Using the structural information obtained from these studies, I engineered a fluorescent
annexin-based biosensor applicable to the in vivo detection of changing lipid
3
compositions, which are specifically associated with the apoptotic pathway and neuronal
degeneration (Chapter 3). In addition, I studied an example of a membrane protein, α-
syn, for which the membrane interaction of the misfolded oligomeric state has been
proposed to be a mechanism for cell toxicity in Parkinson's disease. With the ultimate
goal of studying the mechanism of membrane interaction of the α-syn oligomer, and to
better understand the mechanism of α-syn misfolding and its role in PD pathogenesis, I
studied the oligomeric state of α-syn formed under different conditions (Chapter 4).
Structural states of anxB12
Several distinct states have been observed for anxB12 including the solution monomer,
the Ca
2+
-dependent surface bound trimer, and the Ca
2+
-independent transmembrane
state. The X-ray crystal structure of anxB12 (Fig. I-1) has been solved (Luecke et al.,
1995) and a comparison with structural data obtained from EPR studies of the protein in
solution indicates that the crystal structure is a representation of the solution state of
anxB12 (Isas et al., 2002; Langen et al., 1998b). In the presence of Ca
2+
, anxB12 binds
peripherally to the surface of negatively charged membranes as trimers (Isas et al.,
2005; Isas et al., 2004; Langen et al., 1998b) and is proposed to form a 2-D crystalline
array as observed for annexin A5 (Seaton, 1996). A second Ca
2+
-independent
peripherally bound state has also been observed (Hegde et al., 2006). In addition to
peripherally binding to membranes, a transmembrane state is observed in the absence
of Ca
2+
at mildly acidic pH (Isas et al., 2000; Isas et al., 2003; Langen et al., 1998a). The
difference in these two binding mechanisms may indicate distinct functional states of the
protein which adapt to changes in the cellular environment.
4
AnxB12 belongs to a family of structurally related proteins, the annexins, defined by their
ability to bind negatively charged membranes in a Ca
2+
-dependent and reversible
manner. The Greek term “annex,” meaning “to hold/bring together” describes the
principal characteristic of annexins, whose functions are related to Ca
2+
- and membrane-
binding events. The various proposed annexin functions include the structuring and
organization of membrane domains, membrane-cytoskeleton linkages,
exocytic/endocytic transport steps, and the regulation of ion fluxes (Gerke et al., 2005;
Gerke and Moss, 2002; Moss and Morgan, 2004).They are abundantly expressed in
eukaryotes, and between 10-20 annexin genes are expressed in most vertebrates (Moss
and Morgan, 2004). All higher eukaryotic cells express multiple annexin types,
suggesting an essential role in cell function (Gerke and Moss, 2002).
Although much is known about the different annexin types, their precise functions have
not been clearly defined.
In addition to the ability to bind negatively charged membranes in a Ca
2+
-dependent and
reversible manner, annexins are defined structurally by the presence of the “annexin
repeat.” All annexins are comprised of a core domain, usually comprised of four repeats,
and a variable N-terminal domain. Analysis of the X-ray crystal structures which have
been solved for most vertebrate annexins (Liemann and Huber, 1997) reveals a highly
conserved core structure, irrespective of the 45-55% conservation in amino acid
sequence identities (Moss and Morgan, 2004). The repeats are ~70 amino acids and
fold into five α-helices, named A-E, which pack into a curved disc-like structure (Fig. I-1).
The loops between the A-B and D-E helices on the convex side of the protein contain
the annexin-type (or type II) Ca
2+
-binding sites which coordinate Ca
2+
ions in order to
form
5
Fig. I-1. Crystal structure of annexin B12
The core domain of anxB12 is comprised of 4 highly conserved homologous repeats
(shown by color) each containing 5 α-helices, labeled A-E (labeled above for repeat I).
The loop regions between the A-B and D-E helices contain calcium-binding sites that
mediate protein binding to negatively charged membranes (Luecke et al., 1995; Seaton,
1996).
6
bridging interactions with negatively charged phospholipids during membrane binding
(Seaton, 1996).
Previous studies indicate that the Ca
2+
-dependent surface-bound trimer of anxB12 (Isas
et al., 2005; Isas et al., 2004; Langen et al., 1998b) maintains a structure similar to the
crystal structure (Fig. I-1). This brings up several questions: 1. Are all of the Ca
2+
-binding
loops which are located on convex side of the protein involved in membrane-binding? 2.
How does a planar membrane accommodate the binding of a convex, disc-shaped
protein? 3. How flexible is anxB12 during membrane interaction, and is the annexin core
subjected to hinge motions to accommodate binding to planar membranes? To address
these questions, residues in the remaining unstudied loops (the A-B loops of repeats I
and II, and the D-E loops in repeats I and IV regions) were tested by EPR and SDSL,
completing the study of all eight Ca
2+
- and membrane-binding loops (Chapter 1).
Under mildly acidic conditions, and in the presence of negatively charged phospholipids,
anxB12 has also been observed to form a Ca
2+
-independent transmembrane monomer
(Isas et al., 2000; Isas et al., 2003; Langen et al., 1998a). Previous EPR studies of the
transmembrane state of anxB12 (encompassing residues 138-159 in repeat II)
confirmed the structural transition of this D-E helix-loop-helix in solution into a single
transmembrane amphipathic helix, with one side facing the hydrophobic lipid
environment and the opposite side facing the aqueous environment, suggesting the
formation of a transmembrane channel (Langen et al., 1998a). We hypothesized that
other transmembrane helices must also form in order to accommodate the
transmembrane structure, which may involve similar structural transitions in other helix-
loop-helix regions. To test this hypothesis, EPR studies were performed on the A-B
7
helix-loop-helix region in repeat IV (residues 251-273). In addition, the formation of the
transmembrane structure was compared in two different lipid environments, one using
phosphatidylserine-containing membranes and the other using cardiolipin-containing
membranes to help elucidate the different effects of membrane charge and pH (Chapter
2).
AnxB12 as an indicator of apoptosis and degeneration
The ability of annexin proteins to sense and bind with high specificity to
phosphatidylserine-containing membranes in the presence of Ca
2+
makes it a useful
indicator of apoptotic processes in cells. Apoptosis is a normal physiological process
which occurs during development and the maintenance of tissue homeostasis. The
apoptotic pathway is characterized by distinct morphological changes in the cell, which
include the condensation of the cytoplasm and nucleus, cleavage of DNA, and loss of
membrane integrity (Kerr et al., 1972; Wyllie et al., 1980). In healthy cells, plasma
membrane asymmetry is highly regulated, and negatively charged lipids such as
phosphatidylserine (PS) are maintained on the inner leaflet of the plasma membrane
(Bretscher, 1972; Verkleij et al., 1973). Loss of plasma membrane asymmetry is one of
the earliest features of the apoptotic pathway, resulting in the externalization of PS
(Fadok et al., 1992; Koopman et al., 1994). Current commercially available kits typically
use a fluorescently labeled annexin (such as fluorescein labeled annexin A5) to bind and
identify cells which have exposed PS to the external leaflet of the plasma membrane. In
order to distinguish between early apoptotic and late-apoptotic or necrotic cells, the
fluorescently labeled annexin is used in conjunction with a membrane-impermeable vital
dye (such as propidium iodide, PI) which will label the nuclei of cells that have lost
8
membrane integrity (which occurs at the end stages of the apoptotic pathway or during
necrosis) (Koopman et al., 1994).
Despite the fact that much is known about the morphological and biological changes
which occur during the apoptotic pathway, there is a lack of methods which allow for the
continuous examination of cells undergoing (or recovering from) apoptosis in cell culture
or in vivo. The current commercially available annexin kits for detecting apoptosis have
been optimized for analysis of cells by FACS (fluorescence-activated cell sorting) or
fluorescence microscopy, both of which accommodate for the removal of unbound
annexin molecules in solution to decrease background fluorescence. An annexin-based
probe that is suitable for live-cell or in vivo imaging would provide a non-perturbative
method to study and monitor cell fate under changing environmental or biological
conditions. Based on our studies of anxB12 in the Ca
2+
-dependent membrane-bound
state (Chapter 1), we determined which residues are involved in lipid interaction and
which residues were not during membrane-binding. Using this information, we designed
a polarity-sensitive annexin-based biosensor, pSIVA (polarity-sensitive indicator of
viability and apoptosis), which emits high fluorescence intensity in the membrane-bound-
state but not in the solution. Using pSIVA, we demonstrate its application to real-time
imaging of the apoptotic pathway in cells. Furthermore, we show the application of
pSIVA for studies of neuronal degeneration and the recovery from degeneration in cell
culture and in vivo (Chapter 3).
Structural states of α-syn and protein misfolding
α-syn is a small, monomeric 140-residue protein that is a natively unfolded monomer in
solution, but depending on its environment it can fold into a helical structure on the
9
surface of vesicles or self-aggregate to form oligomeric and fibrillar structures (Dawson
and Dawson, 2003). α-syn has historically been described as being composed of three
domains: an amphipathic region (residues 1-60), the hydrophobic NAC region (residues
61-95), and a C-terminal acidic region (residues 95-140), shown in Fig. I-2. There are
seven imperfect KTKEGV repeats the the N-terminal region, which aid in the formation
of a helical membrane-bound monomeric state. The NAC domain refers to the segment
of α-syn that is found in the amyloid plaques (non-ß-amyloid component) of Alzheimer's
disease patients.
α-syn and Parkinson’s disease
α-syn is most well known for its involvement in the pathogenesis of Parkinson’s disease
(PD). PD is one of the most common neurodegenerative disorders and is caused by the
loss of dopaminergic neurons in the substantia nigra pars compacta (midbrain). The loss
of these neurons causes progressive motor dysfunctions characterized by resting tremor
and rigidity. The pathological hallmark of PD is the presence of Lewy bodies and Lewy
neurites, which are cytoplasmic inclusions (fibrillar protein aggregates), composed
mainly of α-synuclein, in the cell body and neurites of surviving neurons (Forno, 1996;
Spillantini et al., 1998; Spillantini et al., 1997).
The first gene associated with familial PD (autosomal dominant inheritance) encoded the
α-syn protein with an A53T missense mutation (Polymeropoulos et al., 1997) (Fig. I-2).
Subsequently, two other missense mutations have been identified in α- syn, A30P
(Kruger et al., 1998) and E46K (Zarranz et al., 2004). In addition, a gene triplication of α-
syn has been shown to cause PD (Singleton et al., 2003), which implicates that
increased protein levels of wild-type α-syn can predispose individuals to the disease.
10
Fig. I-2. Structural organization of α-synuclein
α-syn is a small, monomeric 140-residue protein that has seven imperfect KTKEGV
repeats in the N-terminal region, which aid in the formation of a helical membrane-bound
monomeric state. The NAC domain refers to the segment of α-syn that is found in the
amyloid plaques (non-ß-amyloid component) of Alzheimer's disease patients. Its
aggregation is central to the pathogenesis of Parkinson's disease and other
"synucleinopathies."
11
On the basis of the identification of these genetic abnormalities in addition to the
identification of fibrillar α-syn aggregates as a major component of Lewy bodies, α-syn is
considered to have a central role in the pathogenesis of PD.
The “toxic oligomer hypothesis”
The aggregation of α-syn likely reflects changes in the intracellular environment which
promote non-native protein interactions that lead to misfolding of the protein. In the case
of familial PD, the inherited missense mutations in α-syn have been shown to promote
misfolding and aggregation when compared to wild-type protein (Choi et al., 2004;
Conway et al., 2000). The formation of fibrillar α-syn aggregates which are observed in
lewy bodies and lewy neurites are key events in the pathogenesis of PD. However, there
is a lack of association between inclusion formation and neuronal cell death, elucidated
by in vivo and in vitro studies (Volles and Lansbury, 2003), which has raised the
question: what is the primary cause of cell toxicity? Several studies have hypothesized
that the soluble oligomeric intermediates may be the most toxic to cells (Volles and
Lansbury, 2003). This “toxic oligomer hypothesis” is supported by studies in a line of α-
syn transgenic mice, in which motor impairment and the loss of dopaminergic neurons
are observed in the presence of non-fibrillar α-syn inclusions (Masliah et al., 2000).
Additional evidence that the soluble oligomeric structure of α-syn are pathologically
relevant is given by the fact that the soluble oligomers of α-synuclein are present in
human PD affected brains (Sharon et al., 2003). Moreover, in vitro studies of α-syn
aggregation have identified morphologically distinct populations of intermediate
structures that exist before fibril formation, including amorphous aggregates, oligomeric
(spherical or donut-shaped) and protofibrillar structures (Uversky, 2003). Furthermore,
the familial PD mutations A53T and A30P have an increased propensity to form
12
oligomers but not fibrils compared to the wild type (Conway et al., 2000). The oligomeric
structure of α-syn was shown to bind and disrupt vesicles, causing leakage of its
contents (Volles et al., 2001), indicating a possible mechanism of α-syn mediated
neurotoxicity (Fig. I-3).
Mechanisms of α-syn misfolding and aggregation
A central unresolved question is how the unstructured soluble state of α-syn protein is
converted into the insoluble fibrillar state observed in Lewy bodies. In addition, it is
unknown what structural similarities are shared by oligomers and fibrils. Considering that
aggregation is a stepwise process, some similarities could exist between the oligomeric
and fibrillar structures of α-syn. A better understanding of the oligomeric state will
provide information on the mechanism α-syn aggregation, and help elucidate the
possible mechanisms for oligomer mediated neurotoxicity and possible membrane
interactions. For example, the structural details of the aggregation pathway will answer
questions such as whether the oligomer is a structural precursor to the fibril. More
specifically, what are the residue-specific interactions that stabilize the oligomeric state
and how do they differ in the fibril? Which regions are structured and which are
unstructured? What secondary structure elements comprise these regions? Knowledge
of the structural constraints that govern oligomer function is important to the
development of inhibitory molecules to the disease-causing oligomeric precursors,
ultimately for treatment of disease (Fig. I-4). Using SDSL in combination with EPR, we
analyzed the structure of the oligomeric state of α-syn formed under two different
conditions, and provide a comparison between the structural organization of these
oligomers with the fibril structure (Chapter 4).
13
Fig. I-3. Structural states of α-synuclein
α-syn is a soluble, natively unfolded monomer, but depending on its environment it can
fold into different structures which may have different functional consequences. In vitro
studies of α-syn aggregation have identified morphologically distinct populations of
oligomeric structures that exist before fibril formation, which can bind and disrupt
membranes.
14
Fig. I-4. Protein misfolding and aggregation
Numerous studies have shown protein misfolding and aggregation has a central role in
the pathogenesis of several apparently unrelated neurodegenerative diseases
(Alzheimer's disease, Huntington's disease, etc., in addition to Parkinson's Disease).
Even though the proteins involved are unrelated in sequence or structure, each
undergoes protein misfolding and aggregation, eventually resulting in fibrillar aggregates
of the disease causing protein. Recently, the oligomeric structures, rather than the
fibrillar aggregates have been suggested to be the cause of cellular toxicity. Our study
has identified two distinct oligomeric structures, both of which are structurally unrelated
to the fibril, indicating that they are most likely "off-pathway" to fibril formation.
15
Application of EPR spectroscopy to the study of protein structures
SDSL in combination with EPR spectroscopy provides a selective and sensitive
approach for determining structural and dynamic features of membrane proteins, soluble
proteins, and amyloid fibrils (Hubbell et al., 2000; Hubbell et al., 1998; Torok et al., 2002).
Measurement of an EPR signal requires the presence of a paramagnetic species, which
is accomplished by the incorporation of a spin-label at specific residue positions by
SDSL. SDSL involves the substitution of a single cysteine residue at specific position in
a protein by site-directed mutagenesis, which is then reacted with the spin label. If the
protein contains native cysteines, a cysteine-less version of the protein is used, so that
each cysteine mutant contains only one spin-label. The spin label used in these studies
is a nitroxide radical, MTSL (1-oxyl-2,2,5,5-tetramethyl- -3-pyrroline-3-methyl
methanethiosulfonate), which reacts with the cysteine residue to form a disulfide bond
between the protein and spin label, creating the side chain termed R1 (Fig. I-5). The
reaction is performed at physiological pH (~7.4) in solution. Since the size of the spin
label is relatively of similar size to amino acid side chains, is neutral in charge and has
intermediate polarity and hydrophobicity, spin labeling typically has no detectable effect
on the protein.
Electron paramagnetic resonance (EPR) is the resonant absorption of microwave
radiation by electrons exposed to a magnetic field. The spectrum that is obtained by
EPR is actually the first derivative of the absorption intensity. In the presence of an
applied magnetic field, electromagnetic radiation is absorbed by the sample in the form
of microwaves and allows for the transition of an electron from a low to a high energy
state. This absorption of microwave energy is referred to as resonance. Resonance
16
Fig. I-5. Site-directed spin labeling
Native cysteines were converted to alanines, and single cysteine substitutions were
placed at specific sites. The sulfhydryl group reacted with MTSL (1-oxyl-2,2,5,5-
tetramethyl-∆-3-pyrroline-3-methyl methanethiosulfonate) to produce the paramagnetic
R1 nitroxide side chain.
17
occurs when the energy of the photons in the electromagnetic field matches the energy
separation of the spin states [the resonance condition is: ∆E= hυ = gβH, where h is
Planck’s constant, υ is the frequency, g is the Zeeman factor (= 2.0023 for free e-), β is
the Bohr magneton (= eħ/2m) and H is the magnetic field]. Since paired electrons have
spins in the opposite direction, the energy states of paired electrons cancel each other
out. Thus, in order to see absorption spectra, the material must have unpaired electrons
(i.e., be paramagnetic). In addition to the applied magnetic field, the electron can interact
with the neighboring magnetic moments of the surrounding nuclei, which, depending on
the orientation, can add or subtract from the applied magnetic field. The result of these
interactions is the hyperfine splitting (or hyperfine interaction). In the case of the nitroxide
spin label, MTSL, the hyperfine splitting results in the characteristic three line spectrum
shown in Fig. I-6. The spectrum is obtained by monitoring the microwave absorption as
the field is varied to achieve resonance (Knowles et al., 1976).
Analysis of the EPR spectrum (Fig. I-6) provides information on the mobility of spin-
labeled residue, R1, which reflects the conditions of its local environment. The
broadening of the EPR line shape is indicative of the restriction in the motion of the
protein and the degree of conformational restriction of R1. For a more quantitative
analysis, the inverse of the central line width (∆H
0
-1
) is measured as an indicator of
mobility (Fig. I-6), such that high ∆H
0
-1
values correlate to increased mobility and low
∆H
0
-1
values correlate to decreased mobility. R1 mobility is a reliable indicator of the
local structure of the protein (Hubbell et al., 1998; Isas et al., 2002; Mchaourab et al.,
1996) and can be used to monitor conformational changes. The EPR spectra measured
for a library of consecutively spin-labeled positions can be used to determine sequence-
specific secondary structure elements and the overall tertiary organization of the protein.
18
Fig. I-6. EPR spectra characteristic of different R1 mobility
Broadening central line widths and reduced amplitudes correspond to reduced R1
mobility, which is determined by the local structure of the protein. Arrows label spectral
components that indicate an immobilized state of R1.
19
In addition to analysis of the mobility information contained in the EPR spectral line
shape, the EPR power saturation technique provides information on solvent accessibility
and membrane topology. The power saturation method is based on applying partially
saturating microwave radiation to the sample and following the signal with progressing
saturation. Saturation of the sample can be modulated by the presence of paramagnetic
quenchers which enhance the spin-lattice relaxation (a process which returns the spins
from the excited state to the ground state) by collisional or dipole-dipole interactions. For
the study of membrane proteins, paramagnetic quenchers are selected based on their
complementary behavior in membrane-water systems, such as O
2
and Ni (II)
ethylenediaminediacetic acid (NiEDDA) (Fig. I-7). O
2
and NiEDDA accessibility
parameters are in-phase for proteins in solution, and can be used to distinguish between
surface-exposed and buried residues. However, in the presence of membranes,
NiEDDA is preferentially excluded from the hydrophobic core of bilayer while O
2
is
preferentially included (Fig. I-7). The ratio of O
2
and NiEDDA accessibilities measured
for specific R1-labeled residue positions can be used to determine the interaction
between specific sites in a protein and the bilayer. Using spin-labeled lipids of known
structure, the depth of immersion of the spin label can be calibrated to the ratio of the
measured O
2
and NiEDDA accessibilities. This calibration of the EPR parameters to
membrane depth can then be applied to the calculation of the depth of insertion of the
nitroxide side chain, R1 on the protein of interest (Altenbach et al., 1994). Taken
together, a detailed molecular description of the membrane-bound structure of a protein
can be determined (Altenbach et al., 1994; Hubbell et al., 2000; Hubbell et al., 1998). In
addition to mobility and accessibility measurements, we can also detect spin-spin
interactions between single R1 labeled mutants (intermolecular) and double R1 labeled
20
Fig. I-7. Accessibility parameter Π
The accessibility of the R1 side chain to O
2
(which partitions inside the membrane) and
NiEDDA (which partitions outside the membrane) is measured by power saturation.
The accessibility ratios [Π(O
2
) / Π(NiEDDA)] were used to determine which residues
were lipid exposed and which residues were solvent exposed.
21
mutants (intramolecular). Measurement of spin-spin interactions, which are dependent
on proximity, provides information on how different molecules (in the case of proteins
spin labeled at a single position) or different domains (in the case of proteins spin
labeled at two positions) are oriented with respect to each other. For continuous wave
(CW) EPR, spin-spin interactions can be used to calculate distances up to 20 Å (Hubbell
et al., 2000).
22
Chapter 1:
Structure of annexin B12 in the Ca
2+
-dependent membrane-bound state
Abstract:
Annexins are a family of structurally related proteins that bind reversibly to the surface of
lipid membranes in the presence of calcium. Although numerous high-resolution X-ray
crystal structures have been solved for annexins, less is known about the structure of
the membrane-bound state. The transition from the soluble monomer to the membrane-
bound trimer of annexin B12 is mediated by the loop regions located on the convex side
of the protein. In this study, we used site-directed spin labeling (SDSL) to analyze the
dynamic structural changes which occur upon membrane binding and how these
putative calcium and membrane-binding loop regions interact with membranes. We
tested residues in the A-B loops of repeats I and II, and the D-E loops in repeats I and IV.
All of the loops tested showed pronounced differences between the solution and the
membrane-bound forms. Although these sites are highly dynamic in solution (as
expected for loop regions), they became considerably more ordered when bound to the
membrane. Direct membrane interaction was confirmed using O
2
and NiEDDA
accessibility measurements, which also provided information on membrane topography
and structure. Based on this analysis, some loop regions were similar to the X-ray
crystal structure, whereas other loop regions were different, thus indicating the
occurrence of localized changes.
Introduction:
Annexins are a family of structurally conserved proteins that are abundantly expressed
in eukaryotes (Moss and Morgan, 2004). Annexins have been extensively studied as
23
intracellular proteins that localize to the cytosol and undergo reversible binding to
intracellular membranes in response to rising Ca
2+
concentrations (Gerke and Moss,
2002; Seaton, 1996). Changes in the expression and localization pattern of different
annexins have been implicated in several pathological conditions, including cancer and
cardiovascular disease (Hayes and Moss, 2004; Rand, 1999). Although the precise
physiological functions of annexins are not yet fully understood, it is thought that
annexins are likely to be involved in membrane-related events such as vesicle trafficking,
membrane domain organization, membrane fusion, and cell signaling (Gerke and Moss,
2002; Rescher and Gerke, 2004). Since annexins are likely to exert their biological
functions while associated with membranes, a molecular understanding of these
membrane interactions is important. Although high-resolution X-ray crystal structures are
available for the soluble forms of numerous annexins, structural information on
membrane-bound annexins is limited.
The annexin proteins have a variable N-terminal domain and a structurally conserved
core domain that contains the Ca
2+
-binding sites (Moss and Morgan, 2004). The core
domain usually consists of four repeats which are ~70 amino acids in length and
comprised of 5 α-helices, named A-E. The crystal structures of the core domains of
different annexin gene products are nearly superimposable. Helices A-B and D-E form a
bundle with the C-helix sitting on top (Fig. 1-1) (Luecke et al., 1995). The interior of each
bundle contains mainly hydrophobic residues, while the surface of the protein and the
interfaces between the repeats are mainly hydrophilic. The loop regions between the A-
B and D-E helices contain Ca
2+
-binding sites that mediate protein binding to
phospholipid bilayers (Seaton, 1996). Site-directed spin labeling studies of solution and
Ca
2+
-dependent membrane-bound annexin B12 (anxB12, also known as annexin
24
Fig. 1-1. Membrane-binding loops in annexin B12
Side and bottom views of a single anxB12 molecule from the crystal structure (Luecke et
al., 1995) are shown. The core domain contains four repeats (shown by color), each
containing 5 α-helices, named A-E. The loops between the A-B and D-E helices contain
the Ca
2+
- and membrane-binding sites. The location of R1 sites studied are shown by
spheres on the C
α
.
25
XII) showed that both forms had structures similar to the X-ray crystal structure (Isas et
al., 2002; Isas et al., 2004; Luecke et al., 1995). Electron microscopy (Oling et al., 2001;
Voges et al., 1994) and atomic force microscopy (Reviakine et al., 1998; Reviakine et al.,
2000) images of annexins bound to the surface of bilayers in the presence of Ca
2+
also
indicated that the membrane-bound proteins had the same general shape as observed
in X-ray crystals.
Previous studies of the D-E helical hairpin in repeat II of anxB12 (residues 134-163)
(Isas et al., 2004) demonstrated that major structural changes were limited to the loop
residues during membrane-binding. To further study the conformational changes that
occur in anxB12 in the Ca
2+
-dependent membrane-bound state, spin-labeled residues in
the other Ca
2+
- and membrane-binding loops were analyzed by EPR. Single cys mutants
were made at residues in the A-B loops in repeats I (residues 26-32) and II (residues 98-
104) and in the D-E loops in repeats I (residues 68-74) and IV (residues 300-305) and
reacted with the MTSL spin label to form the side chain R1 (see introduction, Fig. I-5).
The mobility and accessibility of R1 in each of these derivatives were then determined
by EPR spectroscopy in order to obtain structural information for soluble and membrane-
bound forms of the protein. Combined with previous EPR studies (Isas et al., 2005; Isas
et al., 2004), this work completes the study of all of the eight Ca
2+
-and membrane-
binding loop regions in the Ca
2+
-dependent membrane-bound state of anxB12.
Methods:
Protein expression and purification
Single-cysteine mutations were introduced into the appropriate loop sites in the annexin
B12 pSE420-mrp33H plasmid (Mailliard et al., 1997) by site-directed mutagenesis
26
(QuikChange, Stratagene), and were verified by sequencing. The annexin B12 mutants
were expressed in DH5α Escherichia coli and purified by reversible Ca
2+
-dependent
binding to phospholipid vesicles followed by size-exclusion chromatography, as
previously described (Langen et al., 1998b; Mailliard et al., 1997). The proteins were
stored in 20 mM Hepes containing 100 mM NaCl at pH 7.4 (Hepes-NaCl), supplemented
with 1 mM dithiothreitol (DTT). Final protein
concentrations were determined by
absorbance at 280 nm and use of the extinction coefficient ε = 12,288 M
-1
cm
-1
.
Spin labeling and EPR spectroscopy
Immediately prior to spin labeling, DTT was removed from the buffer by size exclusion
chromatography with PD-10 columns (Amersham Biosciences). The proteins were
eluted with Hepes-NaCl buffer and were concentrated with centrifugal filter devices
(Amicon Ultra-4 10,000 MWCO). The introduced cysteines were reacted (~1 hr, 22
o
C)
with a three-fold molar excess of 1-oxyl-2,2,5,5-tetramethyl-∆-pyrroline-3-methyl
methanethiosulfonate (MTSL) to create the spin labeled side chain R1 (See Introduction,
Fig. I-5). Unreacted spin label was removed by gel filtration with a PD-10 column. Spin-
labeling did not appear to structurally or functionally alter the anxB12 mutants, as
evidenced by the fact that they underwent reversible Ca
2+
-dependent binding to
phospholipid vesicles similar to native anxB12 (data not shown). Large phospholipid
vesicles composed of a 2:1 molar ratio of palmitoyl-oleoyl-phosphatidylserine to
palmitoyl-oleoyl-phosphatidylcholine (Avanti Polar Lipids) were prepared according to
the Reeves Dowben protocol (Reeves and Dowben, 1969). A total of 30 µg of protein
was mixed with 600 µg phospholipid vesicles in 20 mM Hepes containing 100 mM NaCl
(Hepes-NaCl buffer) at pH 7.4 and was incubated (15 min, 22
o
C) in the presence of
1mM CaCl
2
in order to induce binding. The mixture was centrifuged for 10 minutes in a
27
tabletop centrifuge (15,000 rcf, 4°C). The lipid pellets containing bound anxB12 were
loaded into either glass (VitroCom Inc.) or TPX capillaries for EPR analysis. EPR spectra
of lipid-free anxB12 were recorded in Hepes-NaCl buffer (pH 7.4).
EPR spectra were recorded on a Bruker EMX spectrophotometer fitted with an ER
4119HS resonator at 12 mW incident microwave power. The scan width for EPR spectra
was 100 gauss. The accessibility parameters Π(O
2
) and Π(NiEDDA) were measured by
the power saturation method using a dielectric resonator (~22°C). The oxygen
concentration used to equilibrate the protein solutions was equal to that of oxygen in air,
and the concentration of NiEDDA was 3 mM.
Results:
Structural analysis of anxB12 in solution
Previous structural studies of anxB12 have indicated that during the transition from the
solution to the Ca
2+
-dependent membrane-bound state, major conformational changes
were limited to the loop regions while the helices in the core domain maintained their
structure (Isas et al., 2005; Isas et al., 2004). In order to complete the study of the Ca
2+
-
and membrane- binding loops, EPR studies were performed on the remaining unstudied
loop regions. Several variants of anxB12 containing a single cysteine were made and
labeled with MTSL to place the R1 side chain at positions the A-B loops in repeats I
(residues 26-32) and II (residues 98-104) and in the D-E loops in repeats I (residues 68-
74) and IV (residues 300-305) (Figs. 1-1 and 1-2a).
28
Fig. 1-2. EPR spectra and mobility of spin-labeled loop sites
a, The location of R1 in the different spin-labeled anxB12 variants which were studied
are shown with spheres on the C
α
. b, EPR spectra measured for each R1 variant in
solution (Hepes-NaCl buffer, pH 7.4, containing 30% sucrose) are shown in black, and
spectra measured for the Ca
2+
-dependent membrane-bound form are shown in red.
Arrows indicate the broadening of the hyperfine component in the spectra upon
membrane-binding. c, The inverse of the central line width (∆H
0
-1
) is plotted as a
measure of R1 mobility for each residue position in solution (black circles) and in the
Ca
2+
-dependent membrane-bound state (red diamonds). Higher ∆H
0
-1
values correlate to
higher R1 mobilities, and lower ∆H
0
-1
values correlate to increased R1 immobilization.
29
In order to detect residue-specific conformational changes, the EPR spectra were
recorded for each R1 variant in solution (Fig. 1-2, black EPR spectra) and in the Ca
2+
-
dependent membrane-bound state (Fig. 1-2, red EPR spectra). In order to provide a
more accurate comparison of structural differences between the two states, the EPR
spectra for the solution state was measured in the presence of 30% sucrose, which
slows the tumbling rates of the proteins in solution to make them comparable to the
tumbling rates of membrane-bound proteins (Isas et al., 2002; Isas et al., 2004; Langen
et al., 1998b).
Changes in R1 mobilities
In solution, the EPR spectra measured for residues in the center of the A-B loop regions
(residues 28-30 and 100-102) were characterized by high amplitudes and sharp line
shapes, indicative of high mobility, as expected for unstructured loop residues. In
comparison, the spectra measured for residues located at (residues 26 and 98) or near
structured helix sites (residues 27, 31-32 and 99, 103-104) were characterized by
broader line shapes and lower amplitudes (Fig. 1-2a-b). To provide a quantitative
comparison between the EPR spectra measured for the solution (Fig. 1-2b, black
spectra) and the Ca
2+
-dependent membrane-bound states (Fig. 1-2, red spectra), the
inverse of the central line width (∆H
0
-1
) was calculated and plotted as an indicator of R1
mobility (Fig. 1-2c). The plot of ∆H
0
-1
for residues 26-32 and 98-104 indicated a gradual
decrease in mobility in the transition toward the structured helix sites (Fig. 1-2c). In the
Ca
2+
-dependent membrane-bound state, the loop residues became more ordered.
Large-scale changes in the spectra were observed for residues located in the central
loop regions (residues 28-30, 100-102), which included a broadening of the line shape
and decreases in amplitude (Fig. 1-2b, red spectra), indicating a transition from a
30
completely mobile to immobile environment. Spectra measured for residues near or at
the structured helix sites indicated further R1 immobilization upon membrane binding, as
evidenced by slightly lower amplitudes and the broadening of the hyperfine components
(indicated by red arrows in Fig. 1-2b).
In comparison to the A-B loop residues, residues in the center of the D-E loops were
characterized by lower mobilities in solution (Fig. 1-2b-c), as expected for shorter loop
regions (Figs. 1-1, 1-2a). The EPR spectra measured for residues 72 and 303-304
located in the center of the loops were characterized by sharp lines and high amplitudes.
However, spectra measured for the loop residues 71 and 304 had lower amplitudes and
broader line shapes uncharacteristic of loop residues. Inspection of the crystal structure
revealed that the side chains for residues 71 and 304 project toward the core of the
protein, which explains the increased side chain rigidity. In the Ca
2+
-dependent
membrane-bound state, the largest changes in spectral line shape was observed for
residues 72 and 303-304, which indicated the transition of R1 from a mobile to immobile
environment. In addition, further ordering of residues at or near the helix sites were
observed by the broadening of the hyperfine components of the spectrum (indicated by
arrows) and by decreases in the amplitude.
Membrane interaction probed by NiEDDA and O
2
accessibilities
The plot of ∆H
0
-1
in Fig. 1-2c indicates that the mobility of all R1 residues tested
decreased in the Ca
2+
-dependent membrane-bound state, indicating membrane-induced
ordering effects. However, to determine which residues interacted directly with the
membrane, the accessibilities of the R1 side chain to paramagnetic reagents O
2
and
NiEDDA were measured by the power saturation method. Hydrophobic O
2
preferentially
31
Fig. 1-3. Π(O
2
) and Π(NiEDDA) accessibilities
The accessibility parameters Π(O
2
) and Π(NiEDDA) of the different spin-labeled sites
were plotted for anxB12 in solution (a, top panel) and Ca
2+
-dependent membrane-bound
state (b, bottom panel). The solution accessibility values for Π(O
2
) and Π(NiEDDA) are
in phase and provide an indication of solvent accessibility of R1 at each site. In the
presence of membranes, O
2
preferentially partitions into the hydrophobic lipid
environment while NiEDDA preferentially partitions to the aqueous environment. For the
Ca
2+
-dependent membrane-bound form (bottom panel), high Π(O
2
) values and low
Π(NiEDDA) are observed for residues 29-30, 69-73, 99-103, and 303-304, indicating
that these residues are exposed to the lipid environment.
32
partitions into the hydrophobic core of lipid bilayer, while hydrophilic NiEDDA
preferentially partitions into the aqueous environment and is relatively excluded from the
core of bilayers (Hubbell, Gross et al. 1998). Thus, membrane-exposed sites have low
accessibilities to NiEDDA and relatively high accessibilities to O
2.
Fig. 1-3 shows the plot
of the accessibility parameters Π(O
2
) (black circles) and Π(NiEDDA) (red diamonds) for
the solution (Fig. 1-3a) and membrane-bound states (Fig. 1-3b). As expected, the
accessibility parameters Π(O
2
) and Π(NiEDDA) were comparatively high and in-phase
for the solution state (Fig. 1-3a). For the Ca
2+
-dependent membrane-bound state,
Π(NiEDDA) values were lower for all residues tested, while Π(O
2
) values varied with the
position of the R1 side chain (Fig. 1-3b). In addition, Π(O
2
) and Π(NiEDDA) values were
generally out of phase, particularly for residues in the center of the loops (residues 29-32,
68-74, 98-104, and 302-305), indicative of the differential partitioning of the O
2
and
NiEDDA in the lipid environment. Π(O
2
) and Π(NiEDDA) values for some residues near
the helix sites (residues 26-28 and 300-301) were in phase, indicative of solution
exposure (Hubbell, Gross et al. 1998).
The parameter Φ (Altenbach, Greenhalgh et al. 1994; Hubbell, Gross et al. 1998)
provides an indication of the degree of membrane exposure of a given site. The value of
Φ increases with increasing membrane exposure. To determine which residues are
exposed to the lipid environment in the membrane-bound state, the Φ-values of each
residue position were plotted for the membrane-bound (Fig. 1-4, red diamonds) state. As
a point of reference, the Φ-values for the solution state were also plotted (Fig. 1-4, black
circles).
33
Fig. 1-4. Plot of Φ-values
The parameter Φ, which is the ln[Π(O
2
)/ Π(NiEDDA)], provides an indication of
membrane exposure and membrane immersion depth for lipid-exposed residue positions.
The Φ-values (calculated from the accessibility values in Fig. 1-3) of the Ca
2+
-dependent
membrane-bound state are shown in red and the Φ-values for the solution state are
shown in black.
34
For the A-B loop in repeat I (Fig. 1-4a), Φ-values indicate that residues 29 and 30 in the
center of the loop region are immersed in the lipid environment. Φ-values for residues
26-28 and 31-32 in the membrane-bound state were similar to values calculated for the
solution state, indicating that these sites remained exposed to the aqueous environment
during Ca
2+
-dependent membrane-binding. In the A-B loop region of repeat II (Fig. 1-4c),
the comparison of Φ-values from the solution and Ca
2+
-dependent membrane-bound
state indicate that residues 99-103 are immersed in the lipid environment while only the
flanking residues 98 and 104 remain exposed to the aqueous solution. For the D-E loops
region in repeat I (Fig. 1-4b), the comparison of Φ-values indicate that residues 69-73
are exposed to the lipid environment while the flanking residues 68 and 74 are exposed
to the aqueous environment. In repeat IV, D-E loop residues 303 and 304 are exposed
to the lipid environment, while residues 300-302 and 305 remain exposed to the
aqueous solution (Fig. 1-4d).
Taken together, these data complete the study of all eight Ca
2+
- and membrane-binding
loops in anxB12. A summary of the loop residues which interact directly with the
membrane from all four repeats is shown in Fig. 1-5a.
Discussion
The range of amplitudes and broad line shape of the EPR spectra measured for the
Ca
2+
-dependent membrane-bound state of anxB12 (Fig. 1-2b, red spectra) indicates that
all residues in the loop region (including adjacent helix sites) are in similar immobilized
states. A comparison of the mobility data obtained from the spectra (Fig. 1-2c) to the
accessibility data obtained from power saturation experiments (Figs. 1-3, 1-4) indicates
that the observed changes in mobility are not solely due to the direct interaction of these
35
residues with the lipid environment, but are a result of the immobilization of the protein
and surrounding lipids. For example, although major decreases in mobility (Fig. 1-2b-c)
were observed for loop region in repeat I (residues 26-32), accessibility measurements
(Figs. 1-3, 1-4a) indicate that only residues 29-30 are in direct contact with the
hydrophobic lipid environment. Also, in the D-E loop region in repeat IV (residues 300-
305), we observed a decrease in mobility for the entire region (including a moderate
decrease in mobility for residue 302, Fig. 1-2b-c), however accessibility measurements
determined that only residues 303-304 are exposed to the lipid environment (Figs. 1-3,
1-4d). Therefore, the changes in R1 mobility in the loop regions observed between the
solution and Ca
2+
-dependent membrane-bound state indicate a general membrane-
induced immobilization, in addition to the transition (for particular residue positions) from
an aqueous environment to an environment in which the residue is in direct interaction
with the lipid. This is further supported by previous studies which have described the
unique ability of annexins to reduce the rate of lateral lipid diffusion, and immobilize
membranes upon Ca
2+
-dependent binding (Gilmanshin et al., 1994; Megli et al., 1998;
Peng et al., 2004; Saurel et al., 1998). In the physiological context, these data provide a
mechanism of function which further support the previous observation that annexins
serve as molecular fences that create or maintain domains of heterogeneous
distributions of lipids and/or proteins in cellular membranes (Gerke et al., 2005).
According to the crystal structure, residues 71 and 302 project toward the protein core.
This is corroborated by the EPR spectra and mobility values (∆H
0
-1
) measured for the
solution state, which indicate an immobilized state for R1 at these positions in contrast to
the higher mobility values of the surrounding loop residues. Consistent with previous
studies of the D-E loop region (Isas et al., 2005; Isas et al., 2004), this orientation and
36
immobilized state is maintained for residue 302 in the Ca
2+
-dependent membrane-bound
state. For residue 71 (in the D-E loop of repeat I), although the residue remains in an
immobilized state in the solution and membrane-bound states, the accessibility
measurements indicate that residue 71 is exposed the hydrophobic lipid environment,
indicating a change in side chain orientation and a change from the crystal structure.
Furthermore, flanking residues (69-70, 72-73) also directly interact with the lipid, which is
inconsistent with the membrane-bound structure of other D-E loop regions, such as in
repeat IV (residues 300-305) and previously studied D-E loop regions (Isas et al., 2005;
Isas et al., 2004). In the D-E loop of repeat IV, only two residues (303 and 304) are
exposed to the lipid environment. In the D-E loop of repeat I, more residues are involved
in lipid interaction, including residues 69 and 70 which are helix sites according to the
crystal structure.
The A-B loop of repeat I (residues 26-32) is also inconsistent with other A-B loop sites
(residues 98-104 of A-B loop in repeat II, and unpublished data). In contrast to repeat II,
in which nearly all of the loop residues (99-103) are exposed to the lipid environment,
only two residues (29 and 30) are exposed to the lipid in repeat I. These inconsistencies
which are observed for these loop regions in repeat I may be explained by their location
in the crystal structure (Fig. 1-5a-b), which places repeat I in the outer region during
trimer formation, with the D-E helix placed closer to center than the A-B loop. In addition,
the arrangement of anxB12 into trimers in the Ca
2+
-dependent membrane-bound state
might cause the loop regions (such as the A-B loops in repeats I and III) which are
located toward the exterior to tilt away from the membrane surface, resulting in fewer
residues which interact directly with the membrane. Consistent with this, the A-B loop in
37
Fig. 1-5. Summary of Ca
2+
- and membrane-binding loops
a, The crystal structure of a single AnxB12 molecule is shown with the loop residues
which were exposed to lipid environment (determined by EPR studies) in the Ca
2+
-
dependent membrane-bound state highlighted by red spheres in the C
α
. b, The
organization of three anxB12 molecules into a trimer is shown, (bottom view of the loop
regions taken from the crystal structure of the anxB12 hexamer). (Luecke et al., 1995) c,
Π(O
2
) values were plotted against Π(NiEDDA) values for the Ca
2+
-dependent
membrane-bound state of anxB12. Residues are color coded by repeat (repeats I in
magenta, II in blue, III in purple and IV in green). Circles plot residues located in the A-B
helices and loops, and open squares plot residues located in the D-E helices and loops.
Residues sites which remain in solution show a linear relationship between the Π(O
2
)
and Π(NiEDDA) values (gray shaded area), which applies to sites within helices (plotted
for helices A and B in repeats III, IV, and helices D and E in repeat III) . Some loop
residues are exposed to the lipid membrane and fall into the blue and red shaded
regions. The A-B loop residues of repeat II are more accessible to oxygen and penetrate
furthest into the membrane.
38
39
repeat II is located in the center during trimer formation in the Ca
2+
-dependent
membrane-bound state (Fig. 1-5b) and also has Φ-values indicative of greater
membrane depth in comparison to residues in other loop regions (Fig. 1-4c, 1-5c),
suggesting the possibility that the loop regions which are centrally located in the trimer
are tilted toward the membrane interior.
A summary plot of the O
2
and NiEDDA accessibilities of all the loops tested (Fig. 1-5c),
illustrates the degree of membrane interaction of the different loop residues (and helix
sites) in reference to each other. With the exception of loop residues located in repeat I,
the A-B loops generally have a greater membrane immersion depth (indicated by higher
O
2
accessibilities) than residues in the D-E loops (Fig. 1-5c). This is in accordance with
the structure of the loop regions in the crystal structure, which shows the D-E loops are
shorter in length (~3-4 residues) in comparison to the A-B loops (~5-6 residues). In
addition, residue 101 located in the center of the A-B loop in repeat II had the highest Φ
value, indicating the greatest degree of membrane depth (Figs. 1-4c, 1-5c).
Based on our analysis of the EPR data reported here and in previous studies, we
conclude that residues 29-30 (A-B loop, repeat I), 99-103 (A-B loop, repeat II), 185-186
(A-B loop, repeat III), 259-262 (A-B loop, repeat IV), 69-73 (D-E loop, repeat I), 142,
144,145 (D-E loop, repeat II), 228, 229 (D-E loop, repeat III), 303 and 304 (D-E loop,
repeat IV) are exposed to the membrane during Ca
2+
-dependent membrane-binding (Fig.
1-5a). With the exception of some localized changes in some loop regions (D-E loop
residues in repeat I), the structure of the membrane-bound loops remain similar to that
shown in the crystal structure. Furthermore, although the structural information obtained
by EPR methods for the solution and membrane-bound structures were both consistent
40
with the crystal structure, EPR mobility data revealed a strong immobilization of the all
loop regions. Thus, considerable changes in protein and lipid dynamics were observed,
suggesting a functional mechanism for anxB12.
41
Chapter 2:
Annexin B12 in the Ca
2+
-independent transmembrane state
Abstract:
In addition to the Ca
2+
-dependent binding of annexins to the surface of negatively
charged phospholipid bilayers, annexins also bind to membranes in a Ca
2+
-independent
manner at mildly acidic pH. The structural changes that occur during this Ca
2+
-
independent membrane binding were investigated by performing a site-directed spin
labeling study in combination with EPR spectroscopy on the helical hairpin
encompassing helices A and B in repeat IV of annexin B12. Single cysteine mutants
were made from residue positions 251 through 273 of annexin B12 which were then
labeled with a nitroxide spin label (MTSL). As expected, EPR mobility and accessibility
analyses of annexin B12 free in solution indicated that the structure was in agreement
with the crystal structure. However, EPR studies of annexin B12 bound to membranes at
mildly acidic pH indicated a major structural rearrangement in the scanned region. The
helix-loop-helix structure present in the solution state refolded into a continuous
transmembrane α-helix in which one face of the helix was exposed to the hydrophobic
core of the bilayer and the opposite face was exposed to an aqueous pore. The
hydrophobic membrane-exposed face of the amphipathic transmembrane helix
contained an Asp residue (264), thereby suggesting that protonation of its carboxylate
group stabilized the transmembrane form. Analysis of the amino acid sequence of
annexin B12 revealed several other helical hairpin regions that might refold and form
continuous amphipathic transmembrane helices in response to protonation of asp or glu
residues on or near the hydrophobic face of the helix.
42
Introduction:
Although one of the defining properties of annexins is the ability to reversibly bind to
negatively charged membranes in the presence of calcium, a number of annexins have
been shown to bind to membranes in the absence of calcium, at mildly acidic pH (Fig. 2-
1a) (Faure et al., 2002; Golczak et al., 2001; Isas et al., 2000; Isas et al., 2003; Kohler et
al., 1997; Langen et al., 1998a; Rosengarth et al., 1998). The conditions required for this
Ca
2+
-independent membrane interaction varies between different annexins and is
modulated by pH and the properties of the phospholipid membranes (Fischer et al.,
2007; Hegde et al., 2006; Isas et al., 2003). Biochemical studies on annexins A5 (anxA5,
anxV) and B12 (anxB12, anxXII) have demonstrated that at acidic pH the overall
hydrophobicity of these proteins increases and therefore promotes membrane insertion
(Isas et al., 2000; Kohler et al., 1997). These conclusions are further supported by the
observation that matrix vesicle annexins selectively partition into lipophilic organic
solvents at acidic but not at neutral pH (Genge et al., 1991). Moreover, single-channel
conductance studies done on the Ca
2+
-independent (pH-dependent) membrane-bound
state have demonstrated that anxA5 and anxB12 mediate ion flux across phospholipid
bilayers, suggesting the formation of a transmembrane ion channel (Isas et al., 2000).
Although the physiological significance of this channel-like activity is not known, a
structural model of Ca
2+
-independent binding of annexins to membranes at mildly acidic
pH will further elucidate the potential function.
Previous studies of anxB12 using EPR in combination with site-directed spin labeling
(SDSL) have shown that the Ca
2+
-independent binding to membranes at mildly acidic pH
results in a global refolding of the protein (Isas et al., 2003). Analysis of the region
43
Fig. 2-1. Model of annexin B12 membrane interaction
a, In the presence of calcium, anxB12 form trimers on the surface of negatively charged
membranes. At mildly acidic pH, anxB12 forms a transmembrane structure. Shown in
blue is a transmembrane helix region which was determined by previous studies
(Langen et al., 1998a). We hypothesized that similar changes occur in other regions of
the protein. b, The crystal structure of the anxB12 monomer (Luecke et al., 1995), pdb
1AEI, as viewed from the convex face that contains the Ca
2+
-binding loops. The four
repeats of the core domain are labeled with Roman numerals and are shown in different
colors. Each repeat is composed of five helices, labeled A-E. SDSL and EPR studies for
the Ca
2+
-independent membrane-bound state were performed on residues highlighted
with spheres on the α-carbons, which include a previously published study of residues
138-163 in the D-E helix region in repeat II (Langen et al., 1998a; Langen et al., 1998b)
and the current study of residues 251-273 in the A-B helix region of repeat IV. The α-
carbons of residues in the 251-273 region that are exposed to either aqueous solution
(red) or to lipid (yellow) in the membrane-inserted form at pH 4.0 were determined by Φ
values, see Results. Residues that had intermediate Φ values are shown in white.
Residue 257, shown in gray, was not tested.
44
45
encompassing the D-E helix-loop-helix region of repeat II of anxB12 (Fig. 2-1) has
provided insights into the mechanism of refolding (Langen et al., 1998a). This D-E
helical hairpin was observed to refold into a single continuous transmembrane helix. In
addition, the transmembrane helix was asymmetrically solvated, with the hydrophobic
face exposed to the lipid bilayer and the hydrophilic face exposed to an aqueous
environment, supporting its possible arrangement in a transmembrane channel (Langen
et al., 1998a). This structural reorganization was mediated by an inside-out refolding that
exposes hydrophobic residues which are buried in the protein core in the solution state
to the hydrophobic lipid bilayer in the membrane-bound state. The protonation of two glu
residues in the hydrophobic face of the transmembrane helix was speculated to act as a
pH-dependent trigger to induce refolding and membrane insertion (Langen et al., 1998a).
A second experimental approach using fluorescence spectroscopy confirmed the
formation of this transmembrane helix in the D-E helix region in repeat II of anxB12
(Ladokhin et al., 2002).
In the present study we hypothesized that, in order for anxB12 to form a transmembrane
channel containing an aqueous pore (Fig. 2-1a), other regions would also have to refold
to form transmembrane helices. Additionally, because the membrane-inserted form of
anxB12 at mildly acidic pH is monomeric (Ladokhin and Haigler, 2005), multiple
transmembrane regions are likely to be present in the same protein molecule. Analysis
of the amino acid sequence of anxB12 reveals that seven of the eight helical hairpin
regions (including both A-B and D-E helices) shown in the crystal structure are well-
suited to form transmembrane amphipathic helices that place glu or asp residues on the
hydrophobic face (data not shown). To determine whether an A-B helical hairpin forms a
transmembrane structure, we selected the A-B helix region in repeat IV (residues 251-
46
273, see Fig. 2-1b) for analysis using EPR. The primary goal of this study was to
determine whether this region might also undergo a pH- and lipid-dependent
reorganization from a helical hairpin in solution to a continuous transmembrane helix. In
addition, we wanted to determine whether this putative transmembrane region would
also participate in lining the aqueous cavity of transmembrane anxB12. Site-directed
spin-labeling (SDSL) in combination with EPR spectroscopy was used to investigate the
structural changes that occur in the A-B helix region of repeat IV upon pH-dependent
membrane insertion. The mobility analysis of single R1-labeled cysteine mutants at
consecutive residue positions from 251-273 was used to determine local secondary and
tertiary structure as well as conformational changes. EPR analysis of consecutive spin-
labeled residue positions revealed periodic changes in EPR parameters which were
used to determine secondary structure. In order to determine which residues were
involved in membrane interaction and the membrane depth of the lipid exposed residues,
the R1 accessibility to collision with paramagnetic molecules, such as O
2
or Ni (II)
ethylenediaminediacetic acid (NiEDDA) were analyzed.
EPR data obtained for residues 251-273 of anxB12 in solution were consistent with their
location in a helical hairpin structure, as indicated by the crystal structure. Upon pH-
dependent membrane interaction, this helical hairpin region refolded to form a
continuous transmembrane helix that lined an aqueous pore. Taken together with
previous studies of the D-E helical hairpin in repeat II, these data provide further
evidence for the mechanism of transmembrane insertion of anxB12 at acidic pH.
47
Methods:
Protein expression and purification
A series of 23 single-cysteine mutations was introduced into the appropriate sites in the
anxB12 pSE420-mrp33H plasmid (Mailliard et al., 1997) by site-directed mutagenesis
(QuikChange, Stratagene). All mutants were verified by sequencing and the proteins
were expressed and purified as described above (See “Methods,” Chapter 1).
Spin labeling and EPR spectroscopy
23 single cysteine mutants (residue positions 251-273) were made and spin-labeled as
described above. Attachment of the spin label did not appear to structurally or
functionally alter 22 of the 23 anxB12 mutants. Residue position 257 was unstable and
was excluded from EPR analysis.
Large phospholipid vesicles composed of a 2:1 molar ratio of palmitoyl-oleoyl-
phosphatidylserine to palmitoyl-oleoyl-phosphatidylcholine and a 1.2:1 molar ratio of
1,1',2,2'-tetratetradecanoyl cardiolipin to palmitoyl-oleoyl-phosphatidylcholine (Avanti
Polar Lipids) were prepared according to the Reeves Dowben protocol (Reeves and
Dowben, 1969). A total of 30 µg of protein was mixed with 600 µg phospholipid vesicles
in 100 mM sodium acetate buffer at mildly acidic pH (pH 4.0 in the presence
phosphatidylserine-containing vesicles and pH 5.0 in the presence of cardiolipin-
containing vesicles) and incubated (30 min., ~22
o
C) in order to induce binding. The
mixture was centrifuged for 10 minutes in a tabletop centrifuge (15,000 rcf, 4°C). The
lipid pellets containing bound anxB12 were loaded into either glass (VitroCom Inc.) or
TPX capillaries for EPR analysis. EPR spectra of lipid-free anxB12 were recorded in
Hepes-NaCl buffer (pH 7.4), or in 100 mM sodium acetate buffer (pH 4.0 and pH 5.0).
48
EPR spectra and accessibility parameters Π(O
2
) and Π(NiEDDA) were measured as
described above (see Chapter 1). The concentration of NiEDDA was 3 mM for anxB12 in
aqueous solution and 50 mM for membrane-bound protein. The membrane immersion
depth for lipid-exposed residues was calculated from the parameter Φ (Altenbach et al.,
1994; Hubbell et al., 1998). The calibration of Φ in terms of depth was obtained with the
use of 1-palmitoyl-2-DOXYL-stearoyl-sn-glycero-3-phosphocholine (Avanti Polar Lipids)
spin-labeled on the acyl chains at positions 5, 7, 10 and 12 (Altenbach et al., 1994).
Under the conditions used in the experiments reported herein, depth Å = 3.8 (Φ) + 10.2.
Results:
Structural analysis of anxB12 in solution
To study the conformational changes that occur in anxB12 in the Ca
2+
-independent
membrane-bound state at mildly acidic pH, we performed a nitroxide scanning
experiment. Single cys mutants were made at consecutive residue positions 251 to 273
and reacted with the MTSL spin label to form the side chain R1 (Fig. I-5). All sites except
257 tolerated spin labeling (see Methods). The mobility and accessibility of R1 in each of
these derivatives were then determined by EPR spectroscopy in order to obtain
structural information for soluble and membrane-bound forms of the protein.
The EPR spectra of all sites analyzed in aqueous solution at pH 7.4 (Fig. 2-2a, red
traces) were consistent with the crystal structure of anxB12. For example, the high
amplitude and sharp line shapes of the EPR spectra measured for residues 258R1 to
263R1 in the loop (Fig. 2-1) indicate a high degree of mobility, which is characteristic of
loop or unfolded regions (Isas et al., 2002; Mchaourab et al., 1996). The inverse of the
49
central line width, ∆H
o
-1
, was used as a parameter to quantify R1 mobility, with higher
values indicating higher mobility (Mchaourab et al., 1996). The plot of ∆H
o
-1
as a function
of amino acid sequence number highlights the high mobility associated with the 258-263
loop region (Fig. 2-2b, red circles). In contrast, the spectra measured for residues
264R1-273R1, which correspond to helix B which is largely buried in the core of the
protein, exhibited broad lines of lower amplitude and lower mobility (∆H
o
-1
) characteristic
of buried and immobilized sites (Isas et al., 2002; Mchaourab et al., 1996) (Fig. 2-2a red
traces and 2-2b, red circles). Strongly immobilized spectral components are highlighted
by red arrows in Fig 2-2a. Generally, residues which correspond to positions at the
beginning of helix B and face the outer surface of the protein were more mobile than
residues in helix B which were buried in the center of the protein (Fig. 2-2a and b). The
EPR spectrum of residue 264, though located at the beginning of helix B near the protein
surface, exhibited a more immobilized component (red arrow) that indicated tertiary
contacts with the side chains of residues 302 and 74 (Luecke et al., 1995). In the crystal
structure, helix A is partially exposed to the solvent (Fig. 2-1). The EPR spectra
measured for residues 251R1-256R1 (helix A) are consistent with this structure. Strong
immobilization was observed at residues that are buried in the crystal structure (Fig. 2-
2a, red traces, see arrows; Fig. 2-2b), the EPR spectra for the surface-exposed residues
indicated higher mobility, especially for residue 255R1.
In order to determine whether a change in pH alone causes changes in the solution
structure of residues 251-273, EPR spectral measurements were repeated at pH 4.0 in
the absence of membranes. The EPR mobility parameter at pH 4.0 (data not shown)
were nearly identical to the results obtained at pH 7.4, confirming that structure
50
Fig. 2-2. Mobility of annexin B12 in solution and membrane-bound state
a, Spectra for anxB12 containing R1 side chains at the indicated positions were
recorded in buffer (red traces, pH 7.4) or following binding to phospholipid vesicles at pH
4.0 in the absence of Ca
2+
(black traces, see Experimental Procedures). All spectra were
normalized to the same number of spins. Red arrows highlight highly immobilized
components in the spectra of the soluble protein. b, ∆H
o
-1
, the inverse of the central line
width, of R1-labeled anxB12 in the soluble (red) and membrane-bound (black) state
were measured from the data presented in panel a and plotted as a function of
sequence position. The gray shaded boxes on the left and right correspond to residues
in the A and B helices, respectively, in the crystal structure of anxB12.
51
remained the same. Previous studies of other regions in anxB12 have also indicated that
the backbone fold of the protein in solution did not change as a function of pH in the
range of 4.0 – 7.4, in the absence of phospholipids (Isas et al., 2003; Langen et al.,
1998a).
To further characterize the structure of anxB12 in solution, the local accessibilities of the
R1-labeled residues to collision with the paramagnetic reagents O
2
and NiEDDA were
measured. As expected, the accessibility parameters Π(O
2
) and Π(NiEDDA) were in-
phase throughout the scanned region (Fig. 2-3a). Both reagents had higher
accessibilities to residues on the surface of the protein and lower accessibilities to
residues located in the protein interior (Hubbell et al., 1998). Residues in the loop region
of anxB12 (258R1-263R1) had the highest accessibility values, while buried residues
(251R1-253R1 in helix A, 266R1-273R1 in helix B) were inaccessible to both colliders.
Helix sites which are exposed on the protein surface according to the crystal structure
(254R1, 255R1 in helix A and 264R1, 265R1 in helix B) had intermediate Π(O
2
) and
Π(NiEDDA) values (Fig. 2-3a).
To make a better comparison between the EPR accessibility data of anxB12 in solution
and the crystal structure, the fractional solvent accessibility (FSA) was calculated for
residues 251-273 from the crystal structure using the program MOLMOL (Koradi et al.,
1996) (Fig. 2-3b). The solvent accessible surfaces of the residues in the crystal structure
were in agreement with O
2
and NiEDDA accessibilities measured for R1 labeled
residues of anxB12 in solution. Overall, there was a high correlation between mobility
and accessibility parameters (Figs. 2-2b and 2-3a) of anxB12 in solution. The EPR data
were entirely consistent with the X-ray crystal structure of the soluble form of anxB12.
52
Fig. 2-3. Accessibility of residues 251-273 of annexin B12 in solution
a, Accessibility parameters Π(O
2
) (black circles) and Π(NiEDDA) (red diamonds) of R1
side chains at the indicated positions of anxB12 measured in solution at pH 7.4. b, The
FSA was calculated from the crystal structure of Annexin 12 using the program
MOLMOL (Koradi et al., 1996). Residues with high (pink), intermediate (green) and low
(yellow) solvent accessible surface are illustrated as spheres on the C
α
.
53
Structural analysis of anxB12 bound to membranes at mildly acidic pH.
To investigate the dynamic structural changes which occur during Ca
2+
-independent
membrane-binding at mildly acidic pH, EPR spectra were measured for R1-labeled
anxB12 derivatives in the presence of large vesicles composed of phosphatidylserine-
phosphatidylcholine (2:1 w/w) at pH 4.0. As shown in Fig. 2-2a, most of the spectra for
the membrane-bound form (black traces) were notably different from those of the soluble
form (red traces). To quantify these changes the mobility parameter, ∆H
o
-1
, was plotted
as a function of amino acid sequence position (Fig. 2-2b, black circles). The most
pronounced changes were observed for residues in the loop region (residues 258R1-
263R1) which transitioned from a highly mobile to immobile state upon membrane
binding. The broader line shape and lower amplitude of the EPR spectra of residues
258R1-263R1 (Fig. 2-2a, black traces) were not consistent with a loop structure and
indicated the formation of a folded structure in this region.
The EPR spectra (Fig. 2-2a, black traces) for sites in the regions corresponding to helix
A and helix B were consistent with an ordered structure in the membrane-bound form. A
comparison of the spectra for the solution state and the membrane-bound state revealed
conformational changes at most R1-labeled sites. Sites with low mobility in solution that
were buried in the crystal structure showed significant increases in mobility in the
membrane-bound form (Fig. 2-2). These increases in mobility indicated that packing
interactions in the core of the helical bundle in repeat IV were disrupted and indicated a
conformational change. Conversely, 255R1, which had intermediate mobility and was
exposed to the surface of the protein in solution, became less mobile upon membrane
binding (Fig. 2-2), also revealing a structural change. Thus, analysis of all the R1-labeled
sites indicated that both the helical and loop regions underwent conformational changes
54
in response to binding to membranes at mildly acidic pH. The plot of ∆H
o
-1
of the
membrane bound state as a function of sequence revealed intermediate mobility
parameters that did not show any obvious periodic changes through this region (Fig. 2-
2b, black circles). Furthermore, individual spectra of the membrane-bound derivatives
did not contain well-resolved outer peaks which signify hyperfine interactions (Fig. 2-2a,
black traces), thereby indicating that the membrane-bound structure was more loosely
packed.
In order to determine if the conformational changes which occurred during Ca
2+
-
independent membrane binding at mildly acidic pH was specific to phosphatidylserine
containing membranes, EPR measurements were repeated with anxB12 in the presence
of large vesicles composed of cardiolipin-phosphatidylcholine (2:1 w/w) at pH 5.0. The
EPR spectra and mobility parameters were similar to those measured for anxB12 bound
to phosphatidylserine-containing membranes (data not shown), indicating that similar
conformational changes had occurred in the 251-273 region upon Ca
2+
-independent
membrane binding cardiolipin-containing membranes.
To obtain additional structural information and to determine which residues were
exposed to the hydrophobic core of the lipid bilayer in the Ca
2+
-independent membrane-
bound state, we determined the accessibility of the R1-labeled derivatives to O
2
and
NiEDDA [Π(O
2
) and Π(NiEDDA), respectively]. As shown in Fig. 2-4a and b, the pattern
of O
2
and NiEDDA accessibility data for the membrane-bound anxB12 derivatives were
different from those obtained for the soluble form. While the accessibility data of the
soluble protein clearly revealed three separate regions (helix A, the loop and helix B),
the O
2
and NiEDDA accessibilities of membrane-bound state exhibited a continuous
55
periodic oscillation throughout the entire scanned region with both phosphatidylserine-
and cardiolipin-containing membranes. The maxima (or minima) of the accessibility
values were spaced ~3 to 4 amino acids apart, which is a highly characteristic periodicity
observed for α -helical structures, which contain an average of 3.6 amino acids/turn.
Another difference between the soluble and membrane-bound forms was that the O
2
and
NiEDDA accessibilities were out-of-phase in the membrane-bound form, while they were
in-phase in the soluble protein. Previous studies have shown that the out-of-phase
periodicity is caused by a differential partitioning of O
2
and NiEDDA in membranes
(Hubbell et al., 1998). Nonpolar O
2
preferentially partitions into the hydrophobic core of
lipid bilayer, while polar NiEDDA preferentially partitions into the aqueous environment
and is relatively excluded from the core of bilayers (Hubbell et al., 1998). As a
consequence, membrane-exposed residues exhibit high O
2
and low NiEDDA
accessibilities. In contrast, the solvent-exposed sites of a membrane protein have high
accessibilities to NiEDDA and relatively low accessibilities to O
2.
A convenient measure
of the membrane or solvent exposure of a given site can therefore be obtained through
use of the parameter Φ = ln [Π(O
2
)/ Π(NiEDDA)] (Altenbach et al., 1994; Hubbell et al.,
1998). The value of Φ increases with increasing membrane exposure and, Φ increases
linearly with the immersion depth of a given site for lipid exposed residues (see
Experimental Procedures). As illustrated in Fig. 2-4c, the plot of Φ values for consecutive
residue positions revealed a 3-4 residue periodicity, which corresponds to that of an α-
helix. Furthermore, the overlay of the Φ plot for anxB12 bound to phosphatidylserine-
containing residues (black line) with the plot for anxB12 bound to cardiolipin-containing
vesicles (gray line) showed that the Φ values were nearly identical. These data confirm
56
Fig 2-4. Accessibility of annexin B12 in the membrane-inserted state
a, Accessibility parameters Π(O
2
) (black circles) and Π(NiEDDA) (red diamonds) of R1
side chains at the indicated positions of anxB12 were measured following binding to
phosphatidylserine-containing vesicles at pH 4.0 in the absence of Ca
2+
. b, Accessibility
parameters Π(O
2
) (black circles) and Π(NiEDDA) (red diamonds) of R1 side chains at
the indicated positions of anxB12 were measured following binding to cardiolipin-
containing vesicles at pH 5.0 in the absence of Ca
2+
. c, The function Φ = ln [Π(O2) /
Π(NiEDDA)] at the indicated sequence positions were calculated from the data in panel
a. Residues with side chains exposed to lipid are highlighted in yellow, and those with
side chains exposed to aqueous solution are highlighted in red. The plot with black lines
and circles indicate positions that were exposed to phosphatidylserine-containing
membranes and the plot with gray lines and diamonds indicate positions which were
exposed to cardiolipin-containing membranes. The amino acid sequence of the scanned
region is presented in the single letter code under the graph. d, A helical wheel
representation of the scanned region with the individual positions color-coded based on
the data presented in panel c. The asp at position 264 on the hydrophobic face of the
amphipathic helix is highlighted with a blue box. Met257 (gray) was not analyzed.
57
that the formation of a continuous helix by residues 251 to 273 is not specific to one lipid
type, but is likely a general property of Ca
2+
-independent membrane binding to
negatively charged membranes at mildly acidic pH.
Plotting the accessibility data for the scanned sequence on a helical wheel (Fig. 2-4d)
showed that this α-helix must be asymmetrically solvated because it contained a face
with membrane-exposed residues (yellow circles) and another face with solvent-exposed
residues (red circles). The solvent-exposed face contained predominantly hydrophilic
residues, including several positively charged amino acids (Fig. 2-4d). With the
exception of Asp264 (see Discussion), the face that was exposed to lipids primarily
contained hydrophobic amino acids (Fig. 2-4d). The nitroxide side chain at position 254
was exposed to lipid but was located toward the interface between the hydrophobic and
hydrophilic faces of the amphipathic helix (Fig. 2-4d). If position 254 were occupied by
the native His residue, the charged imidazole side chain might assume side chain
rotamers that could take up more hydrophilic orientations.
The asymmetrically solvated α-helix could fit into two different structural models. The
helix could be peripherally bound, parallel to the membrane in the interfacial region of
the membrane or it could be transmembrane helix lining a water-filled pore (see
Discussion). Both types of structures have been reported for other proteins (Altenbach et
al., 1994; Gross et al., 1999; Hubbell et al., 1998; Jao et al., 2004). In order to
distinguish the orientation of the helix formed by residues 251-273, the immersion depth
of the lipid-exposed residues were calculated (Hubbell et al., 1998). The membrane-
exposed residues in a peripherally bound helix will be at comparable immersion depth,
58
Fig. 2-5. R1 immersion depth of membrane-inserted annexin B12
The immersion depths of R1-labeled side chains following anxB12 binding to
phosphatidylserine-containing vesicles at pH 4.0 (black circles) or cardiolipin-containing
vesicles at pH 5.0 (red diamonds) were calculated for residues with maximal Φ values in
Figure 2-4c (see Experimental Procedures).
59
while the immersion depths of lipid-exposed increases linearly as residues in a
transmembrane helix approach the center of the bilayer. The immersion depth of the
residues with maximal lipid exposure in the scanned region of anxB12 was determined
from the parameter Φ and a calibration curve as described in Experimental Procedures.
As shown in Fig. 2-5 the immersion depths for both lipid types decreased linearly on
either side of 264R1, which, at an immersion depth of about 20 Å, represented the
deepest residue. These data were inconsistent with a peripherally bound α-helix, but
were in excellent agreement with the formation of a transmembrane helix.
Discussion:
Previous studies have described different structural states for anxB12 including a soluble
monomer (Langen et al., 1998b; Mailliard et al., 1997), a Ca
2+
-dependent peripheral
membrane-bound trimer (Langen et al., 1998b), and a Ca
2+
-independent transmembrane
monomer (Golczak et al., 2001; Isas et al., 2000; Ladokhin and Haigler, 2005; Ladokhin
et al., 2002; Langen et al., 1998a). These three forms undergo reversible inter-
conversion, with the equilibrium modulated by phospholipid, Ca
2+
and H
+
(Fig. 2-6). The
primary goal of the current study was to use an SDSL approach combined with EPR
spectroscopy to compare the structure of the A-B helical hairpin in repeat IV of anxB12
in the soluble and Ca
2+
-independent membrane-inserted forms. In solution, the EPR
mobility (Fig. 2-2) and accessibility (Fig. 2-3) parameters for the scanned region were
consistent with the crystal structure of the protein. Studies of this A-B helical hairpin in
repeat IV of anxB12 in the Ca
2+
-dependent membrane-bound form indicated that the
surface bound structure is representative of the crystal structure and no large structural
rearrangements were detected (Isas, Kim, Haigler & Langen, unpublished). However,
significant changes in both mobility (Fig. 2-2) and accessibility (Fig. 2-4) were observed
60
at most sites in this region when these EPR parameters were measured following pH-
induced association with membranes. The helix-loop-helix structure observed in solution
refolded into a continuous α-helix in the membrane-inserted state. Accessibility data
clearly showed that the membrane-associated helical structure was asymmetrically
solvated, with one side exposed to the lipid and the other side exposed to an aqueous
environment (Fig. 2-4). The depth profile of the lipid-exposed residues demonstrated
that the helix was transmembrane (Fig. 2-5). A positively charged residue, Arg274, is
located one residue beyond this transmembrane helix, which may stabilize the proposed
structure by interacting with the negatively charged head groups of the phospholipid
bilayer.
A structural model that is consistent with these data is presented in Fig. 2-6. In this
model, the hydrophilic face of the scanned α-helix faces an aqueous pore formed by
several other amphipathic transmembrane helices. This pore structure also
accommodates the in vitro ion channel-like activity of anxB12 (Isas et al., 2000). The
hydrophobic face of the transmembrane helix faces the core of the bilayer.
Most of the residues on the lipid-facing side of the transmembrane α-helix are
hydrophobic amino acids that were buried in the core of the crystal structure of soluble
anxB12 (Figs. 2-1b and 2-4d, color coded in yellow). However, Asp264 is also located
on the center of the hydrophobic face of the helix (Fig. 2-4d) at the site that is most
deeply buried in the bilayer (Fig. 2-5). The interaction of charged carboxylate asp side
chains with the hydrophobic core of bilayers is very energetically unfavorable, but
uncharged protonated side chains (as expected at pH 4.0-5.0) can be embedded in
bilayers without penalty (Wimley and White, 1996). It is interesting to note that Asp264 is
likely to be a Ca
2+
-coordination site that participates in Ca
2+
-dependent binding of
61
anxB12 to the surface of bilayers. Thus, a switch between the interaction of Asp264 with
Ca
2+
and H
+
may be a trigger for refolding and membrane insertion of this region of
anxB12.
There are a number of structural similarities and a few differences between the pH-
induced membrane insertions of the D-E hairpin that was previously reported (Langen et
al., 1998a) and the A-B hairpin described herein. The common elements are that both
helical hairpins undergo membrane and H
+
-induced refolding to form continuous
transmembrane amphipathic helices with glu or asp residues on or near the hydrophobic
faces. Both transmembrane helices are approximately 20 amino acids long and are
flanked by positively charged amino acids that may interact favorably with the negatively
charged head groups of the bilayer phospholipids. Both transmembrane helices are
amphipathic and are formed by inside-out refolding of helices in the soluble protein. In
the soluble form of anxB12, the hydrophilic face of the A, D, and E helices of repeat IV
(Fig. 2-6) largely face the solvent-exposed external surface of the protein. In contrast,
the hydrophilic face of the B-helix is buried in the center of the protein. It is unusual for
soluble proteins to pack hydrophilic side chains in their core. In anxB12, this feature is
accomplished by a mechanism in which the buried hydrophilic face of the B-helix
interacts with hydrophilic residues across the interface of different repeat domains in the
center of the protein.
A major difference between the studied A-B and D-E helical hairpins in anxB12 is that
the loop between the D and E helices contains three residues, while the loop between
the A and B helices contains six residues. These loops refold and contribute
62
Fig. 2-6. Ca
2+
-independent membrane binding of annexin B12.
The crystal structure of anxB12 is shown with the D-E helix region in repeat II highlighted
in dark blue and the A-B helix region in repeat IV highlighted in green and yellow (the
yellow faces of the helices correspond to the yellow spheres in the lower panel). Both
helical hairpins refold and form continuous transmembrane amphipathic helices upon
Ca
2+
-independent binding to membranes at mildly acidic pH (bottom panel). In the
bottom panel, the lipid-exposed residues (yellow) that were analyzed in the current study
are positioned according to their immersion depths (see Results). The gray cylinders
represent other putative transmembrane helices (see Discussion). The number of these
putative helices is arbitrary.
63
approximately one or two turns, respectively, to the transmembrane helix. In both cases,
the length of the loop is such that the hydrophobic faces of the original helix-loop-helix
regions line up on the same side of the transmembrane helix.
The model shown in Fig. 2-6 depicts an aqueous pore formed by several amphipathic
helices. Since the Ca
2+
-independent membrane-inserted form of anxB12 is a monomer
(Ladokhin and Haigler, 2005), several other regions capable of forming transmembrane
helices are thought to be present within the amino acid sequence of this protein.
Inspection of the amino acid sequence of anxB12 reveals five additional helical hairpins
with structural similarities to the transmembrane helices formed by the A-B helical
hairpins in this study and the D-E region in the previous study (Langen et al., 1998a).
These transmembrane helices (five putative plus the two subjected to nitroxide scans)
correspond to seven of the eight A-B and D-E helical hairpins on the convex face of
anxB12. The A-B helical hairpin of repeat III was not included as a putative
transmembrane helix since it did not contain a region that could form a continuous α-
helix, perhaps because it contains one more amino acid in its loop than are present in
the other three A-B loops. The seven putative transmembrane helices are reasonably
amphipathic, and also place glu and asp residues on or near the hydrophobic face. Each
of these helices also contains ~20 amino acids and is therefore of sufficient length to
span the bilayers. In addition, each helix is flanked by either positively charged residues
or good turn-forming residues, a feature commonly observed in transmembrane helices.
Thus, these regions also might be able to refold and form continuous transmembrane
helices as indicated by the gray cylinders shown in Fig.6a. Future studies are needed to
evaluate the validity of those additional transmembrane regions of anxB12, and to
64
determine the physiological significance of Ca
2+
-independent membrane insertion of this
protein.
65
Chapter 3:
An engineered polarity sensitive biosensor for real-time imaging of
apoptosis and degeneration
Abstract:
Apoptosis is of central importance to many areas of biological research, including cell
homeostasis, development, cancer and neurodegeneration. Currently there is a lack of
methods which permit continuous monitoring of apoptosis or cell viability in a non-toxic
and non-invasive manner. Here we report the development of a tool applicable to live
cell imaging that facilitates the visualization of real-time apoptotic changes without
perturbing the cellular environment. We designed a polarity sensitive annexin-based
biosensor (pSIVA) with switchable fluorescence states, allowing detection only when
bound to apoptotic cells. Using pSIVA in combination with live-cell imaging we observe
dynamic local changes in individual neurons throughout the progression of apoptosis.
Furthermore, we observe that pSIVA binding is reversible and clearly define the critical
period for neurons to be rescued. We anticipate pSIVA can be widely applied to address
questions concerning spatiotemporal events in apoptosis, its reversibility and the general
viability of cells in culture.
Introduction:
Live-cell imaging has become a valuable technique for studying dynamic biological
processes in real-time. The ability to visualize and track active processes in a single
living cell has provided new insights into cellular architecture, membrane organization,
dynamic protein assemblies, molecular organization, and cellular responses to external
stimuli.
66
Central to our understanding of cellular and pathological processes is the knowledge of
the apoptotic state of the cells of interest. While the morphological (Kerr et al., 1972;
Wyllie et al., 1980) and biochemical changes that occur at different stages of apoptosis
are well understood, imaging these changes in living cells has been difficult. Several
assays are available, which are aimed at detecting the specific biochemical changes that
occur at different stages of apoptosis, such as, phosphatidylserine (PS) exposure to the
outer leaflet of the plasma membrane (Fadok et al., 1992; Koopman et al., 1994),
mitochondrial dysfunction (Gottlieb and Granville, 2002; Green and Reed, 1998),
activation of caspases (Cohen, 1997; Li and Yuan, 2008; Nicholson and Thornberry,
1997), DNA fragmentation (Gavrieli et al., 1992; Wyllie, 1980), and loss of membrane
integrity (Koopman et al., 1994; Martin et al., 1995). Current methods for these assays
are limiting and generally disruptive to the cellular environment. In most cases, these
assays are toxic to the cells or require fixation.
A number of reasons make the detection of PS on the extracellular face of the plasma
membrane an attractive target for live-cell imaging. In healthy cells, plasma membrane
asymmetry is closely regulated, and PS is restricted to the inner leaflet (Bretscher, 1972;
Verkleij et al., 1973). Exposure of PS has been well established as a near universal
indicator of early apoptotic processes (Fadok et al., 1992; Martin et al., 1995). Also, PS
provides abundant and easily accessible binding targets that can be detected without the
need to penetrate into the cell. Moreover, it is an early event (Chan et al., 1998; Martin et
al., 1995; Rimon et al., 1997), thus monitoring PS exposure provides a way to observe
the initiation of the apoptotic pathway before other changes are present. This is
particularly useful for the detection of apoptotic processes in which progression into cell
death does not occur (i.e., neuronal pruning or Wallerian degeneration).
67
The current method for PS detection involves using a fluorescently tagged annexin
protein to detect PS exposed on the plasma membrane of apoptotic cells (Koopman et
al., 1994). The annexins are widely used in apoptosis assays because of their ability to
bind specifically to negatively charged lipids such as PS (Gerke and Moss, 2002; Seaton,
1996). The different annexin types have a variable N-terminal tail and a structurally
conserved core domain. The core domain is comprised of the annexin repeat, which
contains five α-helices, designated A-E. The calcium binding loops are located between
the A-B and D-E helices on the convex side of the protein and mediate the protein
binding to membranes (Fig. 3-1a) (Gerke and Moss, 2002; Isas et al., 2005; Isas et al.,
2004; Seaton, 1996).
The annexin-based probes described to date are impractical for live-cell imaging
experiments since separate steps are required for binding of the fluorescent annexin
probe to the apoptotic cells and subsequent removal of the unbound protein in order to
reduce the background fluorescence before analysis by microscopy.
In order to circumvent these problems, we engineered an annexin-based fluorescent
biosensor with built in “on” (membrane-bound) and “off” (solution) fluorescent states.
Based on our previous studies on the solution and Ca
2+
-dependent membrane-bound
structures of annexin B12 (anxB12, also annexin XII) (Isas et al., 2005; Isas et al., 2004;
Langen et al., 1998b), we conjugated polarity sensitive thiol-reactive fluorophores to
cysteines introduced at specific sites in annexin. This coupled membrane-binding to a
measurable increase in fluorescence emission intensity due to their ideal location in the
membrane-binding loops (Fig. 3-1a). In addition to AnxB12 labeled with a single polarity
68
Fig. 3-1: Structure-based design of annexin biosensor
a, Top: The crystal structure of anxB12 (PDB entry 1AEI) (Luecke et al., 1995) is shown
with the four annexin repeats represented in different colors. Each repeat contains two
calcium and membrane binding loops, the A-B loop and the D-E loop (indicated by
arrows for repeat I). Bottom: A representation of anxB12 interaction with negatively
charged membranes in the presence of Ca
2+
is shown with the location of tested sites
(red spheres on C
α
) in the membrane binding loops (101C, 260C) used to create pSIVA.
Residue 4C was also tested as a negative control since its location in the N-terminal tail
on the concave side of the protein is expected to stay fully exposed to the aqueous
environment in both the solution and membrane-bound states. b, Structures of polarity
sensitive labels N,N'-Dimethyl-N-(iodoacetyl)-N'-(7-nitrobenz-2-oxa-1,3-diazol-4-
yl)ethylenediamine (IANBD) and c, 6-Bromoacetyl-2- dimethylaminonaphthalene
(BADAN) (molecular probes). The thiol-reactive labels were attached to annexin via
cysteines which were placed at the specified residue positions by site-directed
mutagenesis.
69
sensitive fluorophore, we generated a double-labeled AnxB12 molecule with
fluorophores attached at both 101 and 260, in order to increase the intensity of the probe.
Because annexin A5 (anxA5, also known as annexin V) has already been widely used
and characterized for apoptosis assays (Koopman et al., 1994; Vermes et al., 1995), we
also tested position 262 in AnxA5, a site homologous to residue 260 in AnxB12 (Seaton,
1996). As a negative control, residue position 4 in AnxB12 was labeled to confirm that
detected changes in fluorescence were directly a result of membrane interaction (Fig. 3-
1a).We screened for polarity-sensitive molecules that emit increased fluorescence
intensity in nonpolar environments and chose two thiol-reactive labels: N,N'-Dimethyl-N-
(iodoacetyl)-N'-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)ethylenediamine (IANBD, Fig. 3-1b)
and 6-Bromoacetyl-2- dimethylaminonaphthalene (BADAN, Fig. 3-1c).
We hypothesized that this polarity sensitive biosensor, pSIVA (polarity-sensitive indicator
of viability and apoptosis), was applicable to the investigation of real-time, chronological
and dynamic events occurring in apoptosis by live-cell imaging. One of the advantages
of live-cell imaging methods is the ability to observe the distinct cell-to-cell variations in
the responses to the apoptotic stimulus, directly revealing the different vulnerabilities of
individual cells. A better understanding of where the first cellular responses occur, the
timing and the severity of the response is fundamental to the understanding of the
pathophysiology of apoptotic processes. Moreover, pSIVA can be used as a tool to
measure the viability of cells under different experimental conditions in a non-
perturbative way.
We tested pSIVA in combination with time-lapse microscopy, to visualize and monitor
the progression of apoptosis in living cells, from early stages to complete cell death at
70
the single cell level. The efficacy of pSIVA was first tested in COS-7 cells as a proof of
principle by the addition of etoposide. To better approximate the physiological condition,
pSIVA was applied to primary dorsal root ganglion (DRG) sensory neurons which were
induced to undergo cell death by nerve growth factor (NGF) deprivation. When pSIVA
was applied to the visualization of neuronal degeneration, we observed a dynamic
progression of apoptosis, first in the axons and then in the cell bodies. Also, we
determined the critical period necessary for pSIVA positive neurons to be rescued by
replenishing NGF in the culture medium at various time points after deprivation, and
observed that pSIVA binding is reversible as neurons recover and regain health.
Methods:
Protein purification and labeling
Cysteine mutations were placed in the appropriate sites in Cys-less variants of AnxA5
(C316A) and AnxB12 (C113A-C302A) (Mailliard et al., 1997) plasmids by site-directed
mutagenesis (QuickChange, Stratagene) and expressed and purified as described
previously (Langen et al., 1998b; Mailliard et al., 1997).
The proteins were reacted with a 10-fold molar excess of N,N'-Dimethyl-N-(iodoacetyl)-
N'-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)ethylenediamine (IANBD, Fig. 3-1b) or 6-
Bromoacetyl-2- dimethylaminonaphthalene (BADAN, Fig. 3-1c) at the introduced
cysteine sites (~2h at room temperature or overnight at 4°C). Labeling was quenched
with a 2-fold molar excess of ß-mercaptoethanol and labeled proteins were eluted with a
PD-10 column in 20 mM Hepes buffer containing 100 mM NaCl at pH 7.4 (Hepes-Nacl).
71
In vitro membrane binding assay
Large unilamellar vesicles (LUVs) containing 100% 1-palmitoyl-2-oleoyl-sn-glycero-3-
phosphocholine (POPC) or 25% 1-palmitoyl-2-oleoyl-sn-glycero-3-[phospho-L-serine]
(POPS) - 75% POPC (Avanti Polar Lipids) were prepared as described previously
(Fischer et al., 2007; Reeves and Dowben, 1969). To induce binding, 1 µM annexin
protein was mixed with 1 mM lipid in Hepes-NaCl buffer containing 1 mM Ca
2+
. The
fluorescence emission was measured using a 1 cm path-length quartz cuvette in a Jasco
FP-6500 spectrofluorometer. The excitation wavelength for IANBD was set to 478 nm,
and fluorescence emission was monitored from 480-650 nm. For BADAN, the excitation
wavelength was set to 380 nm, and fluorescence emission was monitored from 400-650
nm. For the comparison of pSIVA to AnxV-FITC, 1 µM annexin protein was incubated
with lipid vesicles at various molar ratios (1:1000, 1:500, 1:250, 1:100, 1:50. 1:10, 1:1)
and the fluorescence emission intensity was measured at the different protein:lipid ratios
for anxB12-101C-,260C-IANBD and compared with AnxV-FITC (excitation max = 485
nm, emission max = 520 nm).
COS-7 cell culture
COS-7 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM)
supplemented with 10% fetal bovine serum (FBS) and 100 U/ml penicillin and 100 µg/ml
streptomycin and grown on tissue culture plates 24 h-48h prior to live-imaging
experiments. Apoptosis was induced in COS-7 cells by100 µm etoposide and compared
with healthy cells grown without the addition of etoposide (DMSO only).
72
Primary neuron cultures
Dorsal root ganglion (DRG) neurons were isolated and purified as previously described
(Rosenberg et al., 2008). Briefly DRG neurons from E13-15 Sprague-Dawley rats were
dissociated, plated, and purified on collagen-coated tissue culture plates in the presence
of NGF (100 ng/ml) for 7 days prior to imaging. Neurons were grown in etched wells in
order to limit the orientation of axon growth along a single axis. Apoptosis was induced
by removal of NGF from the culture media. For rescue of neuronal degeneration, NGF
was replenished (100 ng/ml) after 7, 10 and 15hrs of deprivation.
Time-lapse microscopy and live-cell imaging
Time-lapse microscopy was performed on an Axiovert 200 motorized inverted
microscope equipped with a complete incubation system (Zeiss). Time-lapse images of
COS-7 cells were taken with the AxioCam MRm digital camera from Zeiss. In order to
minimize phototoxicity, neurons were imaged with a Cascade: 1K camera from
Photometrics, which reduced exposure times. All images were processed using
Axiovision 4.7 software (Zeiss). Cells were imaged in the presence of pSIVA (5-10
µg/ml) and propidium iodide (1 µM).
Co-sedimentation assay
5µg of the various IANBD- or BADAN-labeled annexin proteins were incubated at a
1:1000 molar ratio with sucrose-filled large multilamellar vesicles composed of 100%
POPC or 25% POPS-75%POPC as described previously (Isas et al., 2005; Kim et al.,
2005) in Hepes-NaCl buffer containing 1 mM Ca
2+
in order to induce binding. The
mixture was centrifuged for 10 minutes in a tabletop centrifuge (15,000 rcf, 4°C). The
supernatant was concentrated with 10,000 MWCO centrifugal filters prior to loading on
73
to a gel for SDS-PAGE as the solution fraction “S”. The lipid pellets were incubated in 10
µl of buffer containing 3 mM EDTA to reverse the binding of the annexin proteins to
membranes. The mixture was centrifuged and the supernatant was loaded on to a gel
for SDS-PAGE as the lipid pellet fraction “P”.
Immunostaining
Immunostaining of DRG cultures was performed as previously described (Chan et al.,
2004). Briefly, cultures were deprived of NGF for 20 h in order to induce degeneration.
Cultures were then fixed in 4 % paraformaldehyde and then permeabilized and blocked
by incubation with 20% goat serum and 0.2% Triton X-100 in PBS. Cells were then
washed and incubated (1 h) with pSIVA and the neurofilament antibody (ATCC) before
washing and application of the secondary antibody for neurofilament. For live-cell
staining, DRG neurons were deprived of NGF for 20 h and then subsequently incubated
with pSIVA and neurofilament antibody (ATCC) in D-PBS for 15 min, prior to fixation in
4% cold paraformaldehyde (10 min). The cells were permeabilized in 0.02% digitonin (30
min) before blocking and staining with the secondary antibody.
In vivo transection and imaging of the rat sciatic nerve
Sprague Dawley rats (one month postnatal) were anaesthetised with an IP injection of a
premixed solution containing ketamine (50 mg/kg) plus xylazine (5 mg/kg). After shaving
and preparing the skin, the sciatic nerve from one side of the rat was exposed and
transected just below the greater trocanter using surgical Moria spring scissors with a 5
mm blade (Fine Science Tools). The contralateral sciatic nerve was exposed but was not
transected. The wound was closed with 50 Vicryl (Ethicon) in the muscle fascia and 50
74
Dermalon (Ethicon) sutures in the skin. The rats were given subcutaneous injections of
buprenex/buprenorphine at 0.05 mg/kg daily and housed in standard cages and fed
laboratory chow and water. Three days after transection, pSIVA was applied by an
intramuscular injection below the greater trocanter on the left (uninjured control) and
right (injured) sides of the animal. The nerves were exposed for imaging.
Results:
Structure-based design of a fluorescent polarity sensitive biosensor
In order to design a probe more suited for live-cell imaging applications, we engineered
pSIVA based on the structure of the Ca
2+
-dependent membrane-bound state. Polarity
sensitive labels were placed in the loop regions which mediate Ca
2+
-dependent
membrane interactions, transitioning from a polar (aqueous solution) to a nonpolar (lipid
membrane) environment upon membrane binding (Fig. 3-1a) (Isas et al., 2005; Langen
et al., 1998b). We chose residues at positions 101 and 260 as the labeling sites, due to
their ideal location in the membrane-binding loops (Fig. 3-1a). In addition to AnxB12
labeled with a single polarity sensitive fluorophore, we generated a double-labeled
AnxB12 molecule with fluorophores attached at both 101 and 260, in order to increase
the intensity of the probe. Because annexin A5 (anxA5, also known as annexin V) has
already been widely used and characterized for apoptosis assays (Koopman et al.,
1994; Vermes et al., 1995), we also tested position 262 in AnxA5, a site homologous to
residue 260 in AnxB12 (Seaton, 1996). As a negative control, residue position 4 in
AnxB12 was labeled to confirm that detected changes in fluorescence were directly a
result of membrane interaction (Fig. 3-1a).We screened for polarity-sensitive molecules
that emit increased fluorescence intensity in nonpolar environments and chose two thiol-
reactive labels: N,N'-Dimethyl-N-(iodoacetyl)-N'-(7-nitrobenz-2-oxa-1,3-diazol-4-
75
yl)ethylenediamine (IANBD, Fig. 3-1b) and 6-Bromoacetyl-2- dimethylaminonaphthalene
(BADAN, Fig. 3-1c).
Validation by in vitro fluorescence assays
In order to determine the exact differences in fluorescence emission between the
solution and membrane-bound states we measured the fluorescence intensities of these
annexin-based polarity sensitive biosensors in an in vitro binding assay. Membrane
binding was induced by the presence of PS-containing vesicles in neutral buffer
containing 1 mM Ca
2+
. As expected, the fluorescence intensity was negligible for all the
labeled annexins free in solution (Fig. 3-2a-j, dotted lines), and for AnxB12 4C-IANBD
(Fig. 3-2a) and AnxB12 4C-BADAN (Fig. 3-2f) in both the solution and membrane-bound
states. For AnxB12 101C-IANBD, AnxB12 260C-IANBD, AnxA5 262C-IANBD (Fig. 3-2b-
d) considerable increases in fluorescence intensities were measured along with a slight
blue shift from an emission maximum at 540 nm in the solution state to 525 nm in the
membrane-bound state. Fluorescence emission measured for the double-labeled
AnxB12 101C-, 260C- IANBD in the membrane-bound state was substantially brighter,
with only a negligible increase in the background fluorescence of the solution state (Fig.
3-2e). Based on the typical emission profiles of conventional filter sets for FITC /green
fluorescence, the fluorescence emissions between 500–550 nm (green bar, Fig. 3-2e)
were quantified, revealing a ~45-fold increase in the membrane-bound AnxB12 101C-,
260C- IANBD when compared to the solution state. Slightly lower fluorescence
intensities were measured for the BADAN labeled annexin molecules at corresponding
sites (Fig. 3-2g-j). Using the typical emission profile for blue fluorescence filter sets, the
fluorescence emissions between 400-650 nm were quantified, resulting in a ~10-fold
76
Fig. 3-2. Comparison of fluorescence intensities
Top panel: the excitation wavelength was set to 478 nm and fluorescence emission
intensities were measured for IANBD labeled annexins in the solution state (grey lines)
and the membrane-bound state (green lines): a) AnxB12 4C-IANBD, b) AnxB12 101C-
IANBD, c) AnxB12 260C-IANBD, d) AnxA5 262C-IANBD and e)AnxB12 101C-, 260C-
IANBD (the emission profile of a typical filter set for green fluorescence is shown by the
green box). Bottom panel: the excitation wavelength was set to 380 nm and the
fluorescence emission intensities were measured for BADAN labeled annexins in the
solution state (grey lines) and the membrane-bound state (blue lines): f) AnxB12 4C-
BADAN, g) AnxB12 101C- BADAN, h) AnxB12 260C- BADAN, i) AnxA5 262C- BADAN
and j) AnxB12 101C-, 260C- BADAN (the blue box denotes the ideal emission profile,
420-470 nm, which may be used for the design of custom filters).
77
increase in fluorescence intensity for the membrane-bound AnxB12-101C-, 260C-
BADAN compared to the solution state. A large blue shift was observed from an
emission maximum of 530 nm in the solution state to 450 nm in the membrane-bound
state, which may be used to design custom filters, to cut off most of the background
fluorescence of the solution state while maximizing the emitted fluorescence of the
membrane-bound state. For example, an analysis of the fluorescence emission
intensities between 420-470 nm (green bar, Fig. 3-2j) resulted in a ~50-fold increase in
fluorescence of the membrane-bound AnxB12-101C-, 260C-BADAN compared to the
solution state. For the various pSIVA probes tested in Fig. 3-2, there was no detectable
loss in membrane binding ability of the annexin protein, as judged by co-sedimentation
with PS-containing vesicles (Fig. 3-3a-b). No binding or increase in fluorescence
intensity were observed in the presence of vesicles comprised of phosphatidylcholine
(PC), indicating that specificity to PS is maintained (Fig. 3-3a-j). The advantage of pSIVA
over the conventional FITC-labeled anx A5 is further illustrated by measuring the
respective fluorescence intensities in the presence of PS-containing vesicles (Fig. 3-3k).
As expected, the fluorescence intensities measured for the conventional FITC-labeled
anxA5 remained the same regardless of whether it was in the solution or membrane-
bound state whereas the pSIVA fluorescence strongly increased with increasing
amounts of PS-containing membranes.
Thus, attachment of polarity sensitive labels IANBD and BADAN to residues in the
membrane binding loops provided an effective way to generate annexin derivatives with
built-in “on” and “off” fluorescence states, in a range of excitation and emission
wavelengths. In addition both IANBD- (green fluorescence) and BADAN- (blue
78
Fig. 3-3. pSIVA controls
A co-sedimentation assay of the various IANBD (a) and BADAN (b) labeled annexin
proteins indicate that labeling at the various sites has no effects on membrane-binding.
All of the proteins were associated with the lipid pellet (P) when incubated with PS-
containing vesicles. When incubated with vesicles composed of 100% PC all of the
proteins remained in the supernatant (S). c-j shows the fluorescence intensities of the
different pSIVA variants in solution with PS vesicles (dotted lines) and the fluorescence
intensities in the presence of 100% PC vesicles (solid lines). The excitation wavelength
was set to 478 nm and fluorescence emission intensities were measured for IANBD
labeled annexins: c) AnxB12 101C-IANBD, d) AnxB12 260C-IANBD, e) AnxA5 262C-
IANBD and f) AnxB12 101C-, 260C-IANBD. The excitation wavelength was set to 380
nm and the fluorescence emission intensities were measured for BADAN labeled
annexins: g) AnxB12 101C- BADAN, h) AnxB12 260C- BADAN, i) AnxA5 262C- BADAN
and j) AnxB12 101C-, 260C- BADAN. The only measurable changes observed between
the comparison of fluorescence intensities in the absence or presence of PC vesicles
was due to lipid scatter. k, To provide a comparison of pSIVA (Anx B12 101C-, 260C-
IANBD, solid line) to AnxV-FITC (dotted line), 1 µM annexin protein was incubated with
varying amounts of PS-containing vesicles (1:1000, 1:500, 1:250, 1:100, 1:50. 1:10, and
1:1 molar ratios of protein:lipid). The maximum fluorescence emission intensities
measured at the different protein:lipid ratios for anxB12-101C-,260C-IANBD (excitation
max. = 478 nm, emission max. = 525) was compared with AnxV-FITC (excitation max. =
485 nm, emission max. = 520 nm). As expected for pSIVA, the fluorescence intensity
was directly proportional to membrane-binding, as increasing fluorescence intensities
were observed with increasing amounts of PS-containing vesicles. No change in the
79
overall fluorescence intensity was observed for AnxV-FITC when increasing the amount
of PS-containing vesicles.
80
fluorescence) labeled annexins may be used with conventional filter sets equipped on
most fluorescence microscopes.
Application to live-cell imaging
Next we tested whether pSIVA was suitable for live-cell imaging. In order to test its
capacity to specifically highlight cells undergoing apoptosis, pSIVA was added directly to
the culture medium of COS-7 cells induced to undergo apoptosis by etoposide, a known
apoptotic factor (Karpinich et al., 2002). The cells were monitored under physiological
conditions (37°C, 5% CO
2
) by time-lapse microscopy. We tested both IANBD- and
BADAN- labeled variants of pSIVA and observed similar results (data not shown).
Therefore, we used AnxB12 101C-, 260C-IANBD (referred to as pSIVA) for all
subsequent cell culture experiments based on its enhanced brightness (Fig. 3-2e). In
addition to higher fluorescence intensities, IANBD has the advantage of being excitable
in the visible light spectrum, thereby avoiding the potentially harmful UV spectrum. As
expected, we observed bright pSIVA staining of COS-7 cells in the early stages of
apoptosis (Figs. 3-4, and 3-5b) and a gradual increase in staining concurrent with
progression into late stage cell death, marked by propidium iodide (PI) staining. In
comparison, no annexin or PI staining was observed in COS-7 cells grown under normal
conditions (Fig. 3-5a), confirming that pSIVA binding and fluorescence emission was
specific to apoptotic cells. Furthermore the background fluorescence from the solution
state was close to undetectable. To confirm that the presence of pSIVA in the culture
medium did not perturb the cellular environment, we cultured COS-7 cells in the
presence and absence of pSIVA and did not observe any differences in the cell growth
rate (data not shown). Thus, the use of pSIVA in combination with live-cell imaging
81
Fig. 3-4. Application of pSIVA to live-cell imaging of apoptosis in COS-7 cells
The same fields of COS-7 cells were monitored for two days under physiological
conditions without the presence of an apoptosis inducing factor (DMSO, negative
control) and in the presence of etoposide by time-lapse microscopy. Shown are merged
images of phase contrast, green and red fluorescence channels. Green fluorescence
indicated pSIVA binding to the PS exposed on the outer leaflet of the plasma membrane,
and red fluorescence indicated propidium iodide (PI) staining of nuclei in cells in late
apoptosis, with loss of plasma membrane integrity. * Marks the time point when PI
staining was first seen in the cell. (Scale bar, 20 µm)
82
Fig. 3-5. Separated channel images of COS-7 cells.
Separated channel images are shown for the merged images presented in Fig. 3. COS-7
cells monitored under physiological conditions are shown in (a) and cells monitored in
the presence of the apoptotic factor, etoposide, are shown in (b). Green fluorescence
indicates pSIVA binding to the PS exposed on the outer leaflet of the plasma membrane,
and red fluorescence indicate propidium iodide (PI) staining of nuclei during later stages
of apoptosis, with loss of plasma membrane integrity.
83
provides a means to continuously monitor the progression of apoptosis in living cells
without perturbing the cellular environment.
Application to neuronal degeneration
Having established the utility of pSIVA in live-cell imaging of a simple model system, we
next asked if it could be used to provide insights in a more complex apoptotic process,
such as in neuronal apoptosis and degeneration. One of the interesting features is that
under different conditions, axonal degeneration and death can occur at different times
and independently from each other (Raff et al., 2002). We used pSIVA to study apoptotic
processes in purified DRG sensory neurons. These peripheral neurons are unique in
that they only possess a single axon without dendrites.
Because DRG neurons are dependent on trophic factor support for survival, we induced
apoptosis by deprivation of NGF and monitored cell death via time-lapse microscopy.
Similar to what was observed in COS-7 cells, a time lag of several hours was observed
between initial PS exposures in the axons and complete cell death, indicated by PI
staining of the nuclei. pSIVA staining was observed in both axons and cell bodies of
NGF-deprived neurons (Fig. 3-6b) while pSIVA fluorescence was largely absent in
neurons grown in the presence of NGF (Fig. 3-6a). A gradual increase in fluorescence in
NGF-deprived neurons was observed, corresponding to both a gradual increase in
amount of PS exposure in an individual neuron and also the number of degenerating
neurons present over longer periods of NGF deprivation (Fig. 3-6b). Furthermore,
annexin binding occurred in a specific spatiotemporal order, indicating that PS exposure
84
Fig. 3-6. Application of pSIVA to monitoring the degeneration of DRG neurons
Time-lapse microscopy was used to image DRG neurons in normal physiological
conditions (a) and under NGF deprivation (b); left panels show pSIVA fluorescence and
right panels show merged images of phase contrast, green (pSIVA) and red (PI)
fluorescence. c, Time-lapse images (3 frames/h) showing the progressive movement of
apoptotic PS exposure along single axons to cell bodies from 10 – 14 h (left) and 13 –
17 h (right). * Marks the time point when PI staining was first seen in the cell body. d,
Localized progression of PS-exposure along a single axon shown at 20 minute intervals
revealing a punctate staining pattern of pSIVA. The arrowhead indicates the starting
point and arrows indicate the progression of PS exposure. Merged images in c-d show
the overlay of phase contrast, green (pSIVA) and red (PI) fluorescence of the last time
point shown. The times shown indicate the time after NGF was removed (b-d) or
replaced (a) in fresh media. (Scale bars, 100 µm)
85
86
occurs successively, originating from a particular location in the axon and progressing
toward the cell body or the axon terminal (Fig. 3-6c). A closer look at PS exposure on a
single axon revealed a dynamic, sequential punctate staining pattern (Fig. 3-6d), which
may be an indication of the underlying localized cellular processes involved in the initial
stages of axonal degeneration (Goldstein et al., 2000; Mirnikjoo et al., 2009). The
punctate staining pattern of pSIVA in the axons is characteristic of binding to the PS
exposed on the outer leaflet of the plasma membrane rather than binding PS from the
intracellular side where PS is more abundant and uniformly distributed (Fig. 3-7). The
loss of membrane integrity and late stage cell death, indicated by PI incorporation (Fig.
3-6c left panel) was typically observed ~1h after PS-exposure was observed on the cell
bodies.
The advantages of pSIVA also extend to imaging apoptotic processes in vivo. To
demonstrate the utility of pSIVA for in vivo applications, we imaged degenerating
neurons in the rat sciatic nerve three days after nerve transection. An intramuscular
injection of pSIVA along the sciatic nerve was performed to administer the biosensor. Fig.
3-8 shows the specific staining of degenerating neurons by pSIVA after transection.
Except for some minor staining of tissue damaged while exposing the sciatic nerve for
imaging purposes, pSIVA exclusively stained axons in the sciatic nerve distal to the site
of injury (Fig. 3-8c)
(Saxena and Caroni, 2007; Vargas and Barres, 2007), while staining
in the contralateral control (uninjured) sciatic nerve was undetectable (Fig. 3-8b and e).
Punctate staining was observed in the sciatic nerve axons, reminiscent of the staining
observed in degenerating axons of DRG neurons in vitro (Fig. 3-6). Additionally,
experiments were performed with conventional FITC-labeled anxA5 and pSIVA along
uninjured sciatic nerves to demonstrate the advantage of pSIVA for in vivo and live-
87
Fig. 3-7. Staining pattern of pSIVA on degenerating vs. permeabilized axons.
The staining pattern of pSIVA and neurofilament on NGF-deprived (20 h) DRG neurons.
Live-staining (unpermeabilized) of DRG neurons was compared with staining of
permeabilized neurons. Similar to what was observed in the live-cell imaging
experiments, pSIVA staining of axons which were not permeabilized resulted in a
punctate staining pattern (a). Neurofilament staining was excluded from the majority of
axons (b), indicating that the pSIVA signal (a) arises from binding to exposed PS on
axons with membrane integrity still intact. In contrast the staining pattern of
permeabilized cells with pSIVA was uniform throughout the axon (c), indicative of the
more abundant and uniform distribution of PS on the inner leaflet of the plasma
membrane. For comparison, neurofilament staining of permeabilized axons is shown in
d.
88
Fig. 3-8. In vivo application of pSIVA
In vivo imaging of axonal degeneration after sciatic nerve transection. The sciatic nerve
was transected just below the greater trocanter as indicated by the asterisk (*) in a and c.
Images in b-e display the overlay of fluorescence (green) with bright field images (black
and white). b, Aside from some minor staining of surrounding tissue damaged during the
exposure of the nerve for imaging, pSIVA fluorescence was not observed in the
untransected sciatic nerve. c, Intense staining of pSIVA is seen on axons in the sciatic
nerve distal from the site of injury (marked by *). To demonstrate the advantage of
pSIVA over conventional Annexin dyes, AnxV-FITC (10 µg) was applied to the exposed
nerve on one side of an uninjured animal (d) and compared with pSIVA (100 µg) applied
to the uninjured nerve on the contralateral side (e). The images display a striking
difference in the backgound signal from pSIVA, making it more a suitable tool for
imaging in vivo.
89
90
imaging applications. The striking difference in background fluorescence illustrates the
greater sensitivity of pSIVA for detection of axonal degeneration.
Rescue of neuronal degeneration
Previous reports have indicated that the initiation of apoptosis is not necessarily
indicative of a terminal commitment to cell death (Geske et al., 2001; Hammill et al.,
1999; Kiprianova et al., 1999; Maiese and Vincent, 2000; Tang et al., 2009). In neurons,
axonal degeneration is a highly regulated and dynamic process, which involves
apoptotic mechanisms without necessarily resulting in cell death (Saxena and Caroni,
2007). Understanding the temporal dynamics of if and when rescue is possible in
degenerating neurons will provide an indication of the severity of the cellular response
and the time window in which neuronal survival mechanisms can still be effective. While
these studies have been difficult to address in the past, using time-lapse microscopy and
pSIVA, we set out to define the critical period necessary to reverse the apoptosis
process in sensory neurons.
We first initiated apoptosis in DRG neurons by NGF deprivation and subsequently added
back NGF once PS exposure was detected on the axons (at 7, 10 and 15 h after initial
NGF removal). Complete cell death was blocked by the re-administration of NGF in
some cells but not all (Fig. 3-9). Concurrent with neuronal survival, we observed a
decrease in pSIVA fluorescence indicating that binding decreased as PS was restored to
the inner leaflet of the plasma membrane (Fig. 3-9a-c, e).In general rescue was
observed in neurons with pSIVA stained axons, but not in neurons with pSIVA stained
91
Fig. 3-9. Rescue of neuronal degeneration visualized by pSIVA.
DRG neurons dependent on tropic support were induced to undergo apoptosis by
deprivation of NGF for a) 7, b) 10, or c) 15 hours before NGF was replenished in the
culture media. Time lapse images showed pSIVA binding reversed in some but not all
axons, while pSIVA staining of cell bodies were retained, indicating that rescue was
possible in certain neurons in which PS exposure had not progressed to the cell bodies
(late-stage). Left panels show pSIVA fluorescence and right panels show merged
images of phase contrast, green (pSIVA) and red (PI) fluorescence. The times shown
indicate the time after initial NGF removal. (Scale bars, 100 µm) Graphs in d-e show the
quantification of total fluorescence measured for three fields of view, indicated by
different colors, during a time-lapse imaging experiment using a 2.5X objective. The total
fluorescence measured for pSIVA in the presence of NGF-deprived neurons (d), and in
the presence of neurons which were deprived of NGF for 15 h prior to induction of
rescue by the re-addition of NGF to the culture medium (e) are shown.
92
cell bodies (Fig. 3-9a-c), which was generally observed at later stages (Fig 3-6). From
this pattern, we deduced that rescue was observed in neurons which were still in an
early stage of apoptosis. Some axons retained pSIVA staining after the addition of NGF,
indicating that in intermediate stages of apoptosis, no rescue was possible. We
observed that the rescue of single neurons from apoptosis when NGF was re-
administered within 7–15 h after initial NGF withdrawal. This relatively large time window
may be attributed to the heterogeneous timing of the responses due to the different
vulnerabilities between individual neurons, in addition to the critical period after the
initiation of apoptotic mechanisms in which rescue is still possible (after PS exposure in
the axon but before PS exposure in the cell body).
In order to quantify this process, the total fluorescence was measured for different fields
of view at low magnification. For neurons which were deprived of NGF, low levels of
fluorescence intensity were measured for the first 13 – 15 h, indicating that plasma
membrane asymmetry was initially maintained. This was followed by a period of
increasing fluorescence intensity, indicative of the amount of PS exposure (Fig. 3-3k), in
the time period up to 24 h after NGF withdrawal, after which maximum fluorescence
levels were reached and maintained for the remainder of the time course (Fig. 3-9d).
Similarly, for neurons which were deprived of NGF for 15 h prior to the initiation of
rescue by replenishing NGF to the culture medium, maximum fluorescence intensities
were reached at 20 – 24 h, followed by a period of decreasing fluorescence intensity
which continued to the end of the 40 h time course (Fig 3-9e). Although variations
between the different fields exist, we observed a clear decrease in the fluorescence
intensities after rescue was initiated (24 – 40 h, Fig. 3-9e), compared to the fluorescence
intensities measured for neurons which were continuously deprived of NGF (24 – 40 h,
93
Fig. 3-9d). The variation in the amount of decreasing fluorescence measured for different
fields after the induction of rescue (24 – 40 h, Fig. 3-9e) is a likely representation of the
different vulnerabilities and timings of the neurons in each particular field. The
fluorescence intensities measured for neurons cultured under normal conditions
remained low throughout the time course of 40 h (data not shown).
Discussion:
Using a structure-based design strategy, we developed a polarity sensitive annexin-
based biosensor, pSIVA, applicable to real-time imaging of apoptotic membrane
changes in living cells. The main technical novelty of this approach is in the design of
pSIVA, which couples binding to PS-containing membranes directly to a switching “on”
fluorescence. This makes pSIVA an ideal tool for live-cell imaging since it can be used in
excess, and its continuous presence in the cell culture medium is undetectable until
membrane-binding occurs. Therefore, pSIVA allows for high sensitivity and
instantaneous visualization of PS-exposure on cells from early to late stages of the
apoptotic pathway. In addition, we observed that pSIVA binding is reversible, as the cells
regain health and PS is restored to the inner leaflet of the plasma membrane when
survival factors are replenished prior to a critical time window when a commitment to cell
death is established (Fig. 3-10) These experiments demonstrate the advantage of using
pSIVA as a tool to monitor the fate of a single cell from the initiation of the apoptotic
program to recovery or cell death. This is particularly useful for studies of neuronal
degeneration and (axonal) regeneration.
In our application of pSIVA to studying degeneration in purified neuronal cultures, we
demonstrate the strength of pSIVA as a tool to follow the spatial progression of apoptotic
94
Fig. 3-10. Summary of the progression and/or recovery of early apoptotic events in
the neuron as detected by pSIVA
Recovery from early apoptotic processes which are limited to the axonal compartment
was observed. However, rescue was not observed in the later stages in which pSIVA
staining has progressed to the cell body. Loss of membrane integrity (observed by red
propidium iodide staining of the nucleus) and final cell death was observed soon after
PS exposure is detected on the cell body (within ~1h).
95
mechanisms during the time course of axonal degeneration, using PS exposure as an
indicator (Figs. 3-6, 3-9). Sequential punctate staining at various localized areas of PS-
exposure (Fig. 3-6) may be an indication of the movement of localized apoptotic
signaling processes through the neuron (Goldstein et al., 2000; Mirnikjoo et al., 2009). A
recent study showed that caspase 6 was activated in a
similar punctate staining pattern during axonal degeneration (Nikolaev et al., 2009). The
similarities in the appearance of punctate PS exposure and punctate caspase 6
activation in axonal degeneration further supports pSIVA as an indicator to visualize
apoptotic changes occurring in real time at localized areas in axons.
In summary, pSIVA provides several advantages over previous annexin-based probes
(Figs. 3-3k and 3-8d-e) because it is optimized for live-cell and in vivo imaging methods.
The ability to immediately visualize local membrane changes in the initial stages of
apoptosis is important because it provides an indication of where the first cellular events
are occurring, which may be of pathological or physiological importance. One of the key
strengths of pSIVA is that it provides the means to investigate the time course of
apoptotic mechanisms at the single cell level, regardless of cell-to-cell variations in the
response of neighboring cells to the same environment. Furthermore, in combination
with current and future technical advances in in vivo imaging methods we anticipate the
utility of pSIVA for examining the spatiotemporal progression of cell death and
degeneration in model systems and after injury or disease. pSIVA is particularly suitable
for the study of apoptotic processes in which progression into cell death is not obligatory,
such as in Wallerian degeneration, axonal pruning, and axonal degeneration. Combined
with other cellular markers, pSIVA may be used to directly investigate how signals for
96
survival might counteract the apoptotic pathway in disease, development and cell
homeostasis.
97
Chapter 4:
Structural analysis of α-synuclein oligomers by site-directed spin labeling
Abstract:
The accumulation of misfolded α-synuclein aggregates is associated with the pathology
of Parkinson’s disease (PD) and other neurodegenerative disorders termed
synucleinopathies. Numerous studies on the aggregation pathway of α-synuclein have
described different types of aggregates, which include spherical oligomers, rings, chains,
amorphous structures and fibrils. Increasing evidence indicates that the primary cause of
cell toxicity in disease may be the non-fibrillar oligomeric intermediates rather than
mature fibril. Although various biophysical and biochemical techniques have been used
to examine the structure of these oligomeric states of α-synuclein, a direct comparison of
the structural conformation between these oligomer types and the fibril has not been
shown. In this study, we used electron paramagnetic resonance (EPR) spectroscopy
combined with site-directed spin labeling to obtain residue-specific structural information
for the oligomeric state of α-synuclein. Using this method, we identified two different
types of structures, a transient intermediate and a stable oligomer. In addition, we
observe that the EPR spectra from both types of structures are different from that
obtained for the fibril, suggesting that they are conformationally unrelated to the fibril.
This suggests that these intermediates are not structural precursors to the fibril, but
instead “off-pathway” to fibril formation, providing further support for the suggestion that
fibrils may be protective and that the oligomers are the toxic species. This has important
implications for the understanding of PD pathology and for the development of
therapeutics for the treatment of PD and other α-synuclein related diseases.
98
Introduction:
The pathological hallmark of Parkinson’s disease (PD) is the presence of Lewy bodies
and Lewy neurites, which are cytoplasmic inclusions, composed mainly of fibrillar α-
synuclein. The aggregation of α-synuclein occurs by a misfolding pathway and is a key
factor in the pathogenesis of the disease. α-synuclein protein aggregation is thought to
occur by a stepwise process that includes unstructured monomers, oligomeric and
protofibrillar species, and insoluble fibrils (Uversky, 2003). Recent studies have
suggested that the soluble oligomeric species contributes to the neurodegeneration seen
in affected PD brains, and that the fibrillar form of α-synuclein present in Lewy bodies is
the end product of an aggregation pathway. It is thought that neurotoxicity is mediated
by the well known ability of the oligomers to bind and disrupt vesicles causing leakage of
its contents (Volles et al., 2001), indicating a possible disease mechanism. The primary
objective of this study is to determine the structure of the neurotoxic oligomeric species
in order to gain a better understanding of the mechanism of α-synuclein mediated
toxicity. Despite its biological and pathological relevance, high-resolution structures of
the soluble and membrane-bound α-synuclein oligomers have not been defined, partly
due to the limits of structure determination by conventional methods such as X-ray
crystallography or NMR spectroscopy. Recently, some of these limitations have been
overcome by the application of EPR spectroscopy combined with site-directed spin
labeling (SDSL) to the study of protein structures. This is the principle technique that I
will be using to study the dynamic structural changes that occur in α-synuclein during
misfolding and aggregation.
It was previously reported that full-length α-synuclein undergoes proteolytic processing
in cells and neurons, resulting in a truncated version containing the N-terminal and the
99
NAC domains, but not the C-terminal domain (residues 1-119). The C-terminally
truncated α-synuclein was enriched in both soluble and aggregated forms in brains with
α-synuclein pathology (Li, West et al. 2005). This truncation, which removes the
negatively charged residues in the C-terminus, might facilitate its aggregation and
contribute to the pathogenesis of PD. In addition, aggregation experiments done in the
presence of polyamines and other polycations (Uversky, Li et al. 2001; Antony, Hoyer et
al. 2003; Hoyer, Cherny et al. 2004) indicate that factors which interact with the negative
charges in the C-terminus aids in the aggregation process. Furthermore, the
pathophysiological relevance of the C-terminally truncated α-synuclein was
demonstrated in vivo, by a study in which conditional transgenic mice expressing the C-
terminally truncated α-synuclein (residues 1-119) in nigral dopaminergic neurons were
observed to have reduced striatal dopamine (Daher et al., 2009).
For our studies, we used a C-terminally truncated form of α-synuclein (α-syn∆CT,
residues 1-115), which excludes the negatively charged C-terminus that is unfavorable
to the aggregation process (Fig. 4-1). The aggregation conditions (buffer, pH,
temperature, protein concentration) were optimized to favor the formation of oligomeric
populations. Oligomers were prepared by two different methods and compared. In the
first method, the organic solvent HFIP was used to aid the initial structuring of the
natively unfolded proteins. Fluorinated alcohols such as HFIP have been observed to
stabilize α-helical secondary structure in peptides, by a mechanism which involves
clustering around a protein to exclude water and promote the formation of local
interactions (Hirota et al., 1997; Roccatano et al., 2005). In the second
Fig. 4-1 Structural organization of α-synuclein
100
α-synuclein (α-syn) is a natively unfolded protein, which is observed to undergo different
structural transitions. It binds to small vesicles as a helical membrane-bound monomer
(Jao et al., 2004), which has been suggested to be related to its physiological function.
α-synuclein is also observed to misfold and aggregate, forming ß-sheet-rich, parallel-
arranged fibrils (Chen et al., 2007; Der-Sarkissian et al., 2003) which have been
associated with Parkinson’s disease pathology. To optimize the protein for aggregation,
we used a C-terminally truncated variant of α-synuclein which excludes most of the
negatively charged residues in the C-terminus that are unfavorable to aggregation.
101
method, oligomers were prepared in a more physiologically relevant condition using
neutral buffer. Aggregation was induced in both methods by agitating an initially
monomeric population of α-synuclein at high protein concentrations. Both methods
resulted in heterogeneous structural populations, which require further isolation of the
oligomeric sub-species by gel filtration or centrifugation. Structural information was
obtained by performing EPR measurements on a library of single cysteine mutants spin-
labeled at consecutive positions at different time points in the aggregation pathway.
Concurrently, circular dichroism (CD) and transmission electron microscopy (TEM)
measurements were performed to confirm the presence of soluble oligomers.
Methods:
Protein expression and purification:
A library of α-synuclein constructs with a C-terminal truncation (α-syn∆CT, containing
residues 1-115) and containing a single cysteine mutatation at the indicated residue
position were expressed and purified as previously described (Chen et al., 2007). Briefly,
the proteins were expressed in Escherichia coli
BL21(DE3)pLys-S cells (Novagen)
induced with 0.5
mM isopropyl-beta-D-thiogalactopyranoside at 25 °C overnight.
The
bacterial cells were harvested by centrifugation at 4,000 x
g for 10 min and lysed in 100
mM Tris-HCl (pH 8.0) containing 500 mM NaCl, and supplemented with 1 mM EDTA and 1
mM phenylmethylsulfonyl fluoride. The
cell lysate was boiled for 30 min and
subsequently centrifuged
for 30 min at 13,000 x g. This was followed by acid
precipitation of the resulting supernatant at pH 3.5 and subsequent centrifugation for an
additional 30 min at
15,000 x g. The supernatant was dialyzed against 20 mM MES at
pH 5.5 containing 1mM EDTA and 1mM DTT (MES buffer) overnight at 4C. After dialysis,
the solution was loaded on to a HiTrap SPXL column (Amersham Biosciences,
GE
102
Healthcare) for ion-exchange chromatography. α-synuclein proteins were eluted with a
gradient of 0-1 M NaCl in MES buffer, and further purified by gel filtration on a Superdex
75 column (Amersham Biosciences,
GE Healthcare) in order to isolate the peak
corresponding to the monomeric form of the purified α-synuclein protein. The purified
monomeric proteins were stored in buffer containing DTT at -80C.
Spin-labeling
DTT was removed from the buffer by gel filtration using a PD-10 column, and the α-
syn∆CT cysteine mutants were placed in 10 mM NH
4
HCO
3
buffer at pH 7.4 for the spin-
labeling reaction. A 10-fold excess of the spin label MTSL, 1-oxyl-2,2,5,5-tetramethyl-∆-
3-pyrroline-3-methyl methanethiosulfonate, (or the diamagnetic analog, 1-acetyl-2,2,5,5-
tetramethyl-3-pyrroline-3-methyl-methanethiosulfonate) was reacted with the α-syn∆CT
variants (~1h, 25°C) to place the nitroxide side chain R1 (or nonparamagnetic R'1) at the
specified residue positions. Unreacted label was removed by gel filtration using a PD-10
column. The spin-labeled proteins were then lyophilized for storage prior to oligomer
preparation.
Oligomer preparations
Two different methods were used to prepare the oligomeric state of α-synuclein:
1. (Using an organic solvent) Purified, lyophilized monomeric α-syn∆CT protein was
dissolved in HFIP (1,1,1,3,3,3-hexafluoro-2-propanol) and subsequently diluted
to 10-20% in H
2
O (~1-3mg/ml protein concentration). The HFIP was slowly
evaporated while agitating the sample by stirring over a time course of 0 h to
several days.
103
2. (Using aqueous solution) Purified, lyophilyzed monomeric α-syn∆CT protein was
dissolved in neutral buffer (10 mM PO4, pH 7.4) and stirred for 20 h in order to
induce aggregation. The sample was then filtered through a membrane (0.2 µm
pore size) to remove large fibrillar aggregates, and loaded onto a Superdex 75
column for gel filtration in order to separate monomeric and oligomeric fractions.
EPR spectroscopy
Changes in the structure were monitored at specific residue positions by EPR
spectroscopy at different time points during the aggregation pathway or for the
separated gel filtration fractions. EPR spectra were recorded on a Bruker EMX
spectrophotometer fitted with an ER 4119HS resonator at 12 mW incident microwave
power. The scan width for EPR spectra was 150 gauss.
A11 dot blot
A11 dot blots were prepared and developed as described previously (Kayed et al., 2003).
Briefly, 2 µl of each sample were dotted on a nitrocellulose membrane and allowed to
air-dry. The membrane was blocked for 1 hour with 10% nonfat dried milk dissolved in
Tris-buffered saline (20 mM Tris, 0.8% NaCl, pH 7.4) containing 0.001% Tween-20
(TBS-T), incubated (1h) with the oligomer specific A11 antibody (1:5,000 in TBS-T
containing 3% BSA), washed, incubated (1h) with HRP-conjugated secondary anti-rabbit
antibody (1:10,000 in TBS-T containing 3% BSA), and washed. The membrane was
developed using enhanced chemiluminescence reagents (Amersham Biosciences;
Piscataway, NJ) and exposed to Hyperfilm (Amersham Biosciences; Piscataway, NJ).
The same procedure was performed for dot blot with α-synuclein antibody (1:10,000; BD
Biosciences) using 0.1% Tween-20 in TBS.
104
Circular dichroism (CD) spectroscopy
A Jasco circular dichroism J-810 spectrometer (Easton,
MD) was used to determine the
overall secondary structure content
for the spin-labeled α-syn∆CT variants at the
different time points and for the separated gel filtration fractions. Spectra were recorded
in 0.1 cm cells from 260 to 190 nm. For all spectra, an average of 10-25 scans were
obtained, and the CD spectra of the appropriate buffers were recorded and subtracted
from the protein spectra.
Transmission electron microscopy (TEM)
10 µl of sample was applied to a 150-mesh formvar-carbon coated copper grid (Electron
Microscopy Sciences, Hatfield, PA) and negatively stained with 6% uranyl acetate. The
grids were examined using a Jeol JEM1400 microscope with an accelerating voltage of
100 kV.
Results:
Oligomer preparation using an organic solvent
In order to first prepare a homogeneous structural population, lyophilized α-syn∆CT
(residues 1-115) was dissolved in 100% HFIP, which induces monomeric α-helical
structuring of the proteins (Hirota et al., 1997; Roccatano et al., 2005). The sample was
then diluted with water to a 10-20% HFIP-H
2
O solution, and agitated at 25°C in order to
induce aggregation and oligomer formation, as the HFIP was evaporated slowly.
Residue-specific changes in the structure were monitored over time, by EPR
spectroscopy. Fig. 4-2 shows the EPR spectra measured for α-syn∆CT spin-labeled at
position 81 (81R1), over a time course started in an initial 10% HFIP-H
2
O solution. The
105
spectrum measured at 0 h is representative of the HFIP-induced artificial structuring of
the protein into a monomeric α-helical state (Hirota et al., 1997; Roccatano et al., 2005).
Although there is a slight decrease in the amplitude and broadening of the spectra at 5 h,
at 22 h and later time points we observed the appearance of a sharp mobile component
and increasing amplitudes. The sharp, mobile component of the spectrum is
representative of the unfolded monomeric state of α-syn∆CT (Fig. 4-2, red spectrum).
The same time course was repeated for α-syn∆CT spin-labeled at other residues
positions (5R1, 23R1, 30R1, 44R1, 56R1, 81R1, 108R1), and similar changes in the
EPR spectra were observed (data not shown). The observed increases in the mobile
component at all tested residue positions were indicative of the transition from a
structured to unstructured state. Measurement of the CD spectra of the sample at these
time points indicated a shift in the secondary structure content from α-helical to random
coil, indicative of the increasing population of unstructured, monomeric α-syn∆CT (Fig.
4-3). Taken together, these data indicate that the increases in the mobile component of
the spectra observed between 0-48 h is a result of the transition of α-helical to unfolded
monomers as HFIP is lost through evaporation.
Inspection of the line shape at 22 h reveals an additional broader, immobile component
in the spectrum, representative of a population of structured intermediates, which is
observed transiently at 22 h. In addition, the CD spectrum of the sample at 22 h was
indicative of structural heterogeneity, while CD spectra measured at later time points
were indicative of a mostly unfolded population of proteins (Fig. 4-3). The mobile,
monomeric component of the spectrum was subtracted from the spectrum measured at
106
Fig. 4-2. EPR spectra measured for a time course starting in 10% HFIP-H
2
O
EPR spectra were measured at the indicated times during a time course of α-syn∆CT (1-
115) 81R1 dissolved in 10% HFIP-H
2
O. At the 22 h time point, there is an immobile
spectral component (red arrow) that is not present at earlier or later time points. For
reference, the EPR spectrum of α-syn∆CT-81R1 monomer in aqueous solution is shown
in red.
107
Fig. 4-3. Changes in circular dichroism spectra of α-syn∆CT-44R1
CD spectra were measured at the indicated time points during a time course of α-
syn∆CT-44R1 aggregation. At intermediate time points between 5-48 h, the CD spectra
are indicative of heterogeneous structural populations. From 22-72 h, there is an
increase in the population of random coil structures.
108
22 h (Fig. 4-4a), revealing the immobile component of the spectrum. This immobile
component was characterized by low amplitude and broad line shape, compared to the
spectrum measured for the monomeric α-helical state, indicating a more ordered
structure. Furthermore, spin-exchange or spin-spin interaction was not detected in the
spectrum and a comparison with the EPR spectrum measured for 81R1 in the fibril
(which is characterized by spin-exchange) indicates that this transiently structured
intermediate is less ordered, and in a different conformation from the fibril state.
Strong immobilization was also observed (transiently, 22 h) in the EPR spectra
measured for α-syn∆CT spin-labeled at other residue positions (Fig. 4-4b), including N-
and C-terminal residues. Residues 44, 56 and 81 correspond to residues located in the
fibril core which is highly ordered and arranged in a parallel, in register manner (Chen et
al., 2007). Residues 5, 23, and 30 correspond to the structurally heterogeneous N-
terminal region in the fibril structure, however, these sites are strongly immobilized in the
transient intermediate observed at ~22h. Residue 108, which corresponds to the
unstructured C-terminal region of the fibril state, was also observed to be strongly
immobilized, indicating that (unlike the fibril) regions throughout the protein were
structured in this transient intermediate. Furthermore, we deduced that this transiently
structured state was dependent on the presence of low concentrations of HFIP, which
accounts for the loss of these structures at the later time points (Fig. 4-2). In addition,
attempts to isolate this structured population by gel filtration were unsuccessful, likely
due to its instability in different HFIP or buffer conditions (data not shown).
In order to study the size and overall morphology of the aggregated α-syn∆CT structures
present at the different time points, we analyzed the samples by transmission electron
109
Fig. 4-4. EPR spectra of the transient intermediate at 22 h
a, The sharp mobile component of the spectrum (indicative of the presence of
unstructured monomeric state) at the 22hr time point was subtracted from the spectrum,
in order to obtain spectra for the “transient intermediate" structure. b, Preparations of α-
syn∆CT (1-115) spin labeled at other sites gave similar results. The spectra for this
"transient intermediate" structure indicates that residues throughout the protein
(including N- and C- terminal sites) were immobile.
110
microscopy (TEM) (Fig. 4-5a). We observed large spherical aggregates (~10-30 nm
diameter) and long fibrils (~ 5 nm diameter) throughout the time course, revealing the
presence of other smaller populations of structures which were not represented by the
EPR spectra. Spherical oligomeric structures were observed at earlier time points (~0 –
20 h), compared to fibrillar structures, however, these oligomeric structures were also
present at later time points (~ 100 h), indicating that they are not related to the transient
structures observed by EPR at ~22 h. In agreement with previous reports, fibrillar
structures were observed later (starting at ~20-48 h), at increasing amounts over time
(Fig. 4-5a, and observed by centrifugation of fibril precipitates, data not shown).
In order to determine if the structures which we observed by EPR and TEM are related
to previously described cytotoxic oligomers, we performed a dot blot with the A11
antibody. A11 is a structure-specific antibody which has been shown to specifically
recognize the oligomeric state of several different proteins, including α-syn, amyloid-ß,
IAPP and tau (Kayed et al., 2003). The dot blot (Fig. 4-5b) shows that A11-positive
structures were present at all time points throughout the time course, indicating that they
are unrelated to the transiently stable structures at 22 h. Thus, multiple structural
populations are present in the sample, including α-helical monomers, unfolded
monomers, the HFIP-induced transient intermediates observed at ~22 h, A11-positive
oligomers, large spherical oligomers, and fibrils (Fig. 4-5b).
In order to prepare a less heterogeneous population of structures, HFIP was eliminated
from the protocol and the oligomers were prepared in neutral buffer (10 mMPO
4
, pH 7.4).
The samples were incubated at room temperature with stirring in order to induce
111
Fig. 4-5. Structural populations detected by TEM and A11 reactivity
a, Electron micrographs taken of samples from a time course (also shown in b) at the
indicated time points for 56R1, demonstrate that populations of spherical oligomers and
fibrils coexist. b, The A11 antibody, a structural antibody which has been shown to
detect toxic oligomeric structures of several different amyloidogenic proteins, (Kayed et
al., 2003), was used to detect the formation of a specific subset of oligomeric structures.
These A11-positive oligomeric structures were present throughout the time course, and
showed no correlation to the transient intermediates present at 22 h. Furthermore, as
indicated by the up-down arrows, multiple structural populations were present in the
sample.
112
aggregation of the protein. The EPR spectra measured for the different time points were
indicative of the monomeric unfolded state (Fig. 4-6a). However, inspection of the
samples by TEM revealed (Fig. 4-6b) the presence of spherical oligomers and fibrils,
indicating that these structures were present, though not in high enough proportions to
be detected over the monomer signal by EPR. Therefore, in order to analyze the
oligomeric structures by EPR, the sample was separated further at the 22 h time point
(Fig. 4-7a). The larger fibrillar aggregates were removed by filtering through a membrane
(0.2 µm pore size) and monomers were separated from oligomers by gel filtration
(Superdex 75, GE Healthcare). The EPR spectra of the resulting monomeric and
oligomeric fractions are shown in Fig. 4-7b. The monomer spectrum was characterized
by the typical 3 sharp lines, while the oligomer spectrum was characterized by lower
amplitudes and broad line shape, indicating that position 81 is structured in the oligomer.
Inspection of the oligomer fraction by TEM (Fig. 4-7c) revealed a more homogeneous
structural population which consisted of spherical oligomers (~20 nm). Furthermore,
these oligomeric structures were stable during gel filtration. Analysis of the CD spectra of
the separated monomer and oligomer fractions revealed that the oligomer fraction
contained mostly ß-sheet structure, while the monomer fraction was mostly random coil,
as expected (Fig. 4-8).
A comparison of the EPR spectra measured for the oligomer formed in neutral buffer
with the transiently stable HFIP-dependent intermediate demonstrates differences in
mobility and amplitude, which indicate that these structures are arranged in different
conformations (Fig. 4-9). The EPR spectra measured for positions 23 and 44 in the
oligomer formed by aggregation in neutral buffer is characterized by higher amplitude,
113
Fig. 4-6. Time course in neutral buffer
a, EPR spectra were measured during a time course of α-syn∆CT-81R1 dissolved in
buffer (pH 7). No EPR spectral changes were observed, though electron micrographs of
the sample at 48hrs showed that oligomeric and fibrillar aggregates were present (b).
114
Fig. 4-7. Separation of oligomeric α-syn∆CT by gel filtration
a, α-syn∆CT-44R1 was dissolved in 10mM PO
4
buffer at pH7, stirred for 20hrs. Large
fibrillar aggregates were first removed by filter through a membrane (0.2 µm pore size)
before loading the sample on a Superdex 75 gel filtration column to separate monomeric
and oligomeric structures. b, EPR spectra were measured for the sample prior to gel
filtration, and afterwards for the oligomeric fraction #13 and monomeric fraction #20. c, A
typical EM micrograph of the separated oligomeric fraction is shown (wild-type α-
syn∆CT).
115
Fig. 4-8. Circular dichroism spectra measured after gel filtration
The CD spectra were measured for the separated gel filtration fractions of an oligomer
preparation (α-syn∆CT-44R1 stirred 20 h in 10 mM PO
4
, pH 7.4). The spectrum for the
oligomer peak indicates that the structures are mostly ß-sheet. As expected, the
spectrum measured for the monomer peak indicates that the proteins are unfolded.
116
while the spectra measured for the same sites in the HFIP-dependent transient
intermediate is characterized by lower amplitude and broader line shape, indicative of
greater immobilization. The fibril spectrum for position 44 is shown for comparison, and
indicates that all three structures are different. The EPR spectrum for the fibril state is
indicative of spin exchange, while the spectra measured for the oligomeric state formed
in neutral buffer and the transient HFIP-dependent intermediate structure show no
indications of spin-spin interaction, indicating that these structures are less ordered.
Discussion:
In summary, different types of structural populations are readily formed in vitro, and
depending on the method of preparation and the solvent conditions, different sub-
populations can be enriched. Since only certain structures may be pathologically
relevant, choosing the right method is essential. In the case of using HFIP to induce
structure formation, we observed that A11-positive oligomers, spherical oligomers, and
fibrils form readily (Fig. 4-5). However, we also observed the formation of HFIP-
dependent structures, such as the initial α-helical monomer and the transient
intermediate (observed at ~22 h), which might not be physiologically relevant (Figs. 4-2,
4-4). In addition, HFIP is known to cluster around the proteins in a HFIP-H
2
O mixture
and form local interactions with the protein surface (Roccatano et al., 2005). Although
these interactions favor the formation of secondary structure, it also raises the
question of whether the HFIP is completely removed from the protein structure at a
particular time in the aggregation or protein folding process. This uncertainty in whether
HFIP is still present in the sample may be problematic when these structures are tested
in assays to assess cell toxicity, since HFIP itself is a cause of membrane
permeabilization and toxicity.
117
Fig. 4-9. Comparison of structures detected by EPR
Stable oligomers separated by gel filtration in neutral buffer have a different structure
than the HFIP-dependent transient intermediate (~22 h), indicating that different
methods of oligomer preparation favor the formation of different types of oligomers. EPR
spectra of both oligomer types did not show spin exchange (shown for position 44 in the
fibril state), which indicated that these oligomeric structures are different from the fibril
structure.
118
In addition to the method of preparation, it is important to note that different methods of
detection were necessary to identify the different types of structures which were present
in the sample. For example, in the case of structural heterogeneity EPR and CD
measurements are both biased toward detecting larger structural populations. In addition,
EPR measurements are strongly biased toward detecting mobile structures (such as the
unfolded monomer) while methods such as EM excludes detection of monomers, and
only detect larger aggregates. Thus, it is important to consider the potentially different
effects of different structural populations which are present in the sample.
In this study, we demonstrate that different types of intermediates are formed in vitro (in
the α-syn aggregation pathway), which vary in structure and stability. A large number
population of intermediates formed in HFIP-H
2
O are transiently stable (~22 h, “transient
intermediate”) while a smaller subpopulation of A11-positive oligomers are stable for
several months (Figs. 4-2, 4-4 and 4-5). Stable oligomeric intermediates formed in
neutral buffer without organic solvents were isolated by gel filtration (Fig. 4-7). EPR
spectral analysis indicated that two structurally distinct intermediates were generated,
both of which have a structure that is different and more loosely packed than the fibril
(Fig. 4-9). In the case of the "early intermediate" formed in the presence of HFIP, both N-
and C- terminal residues were immobile, indicating that the domain organization is also
different from the fibril (Fig. 4-4). None of intermediates prepared in the present study
showed structural resemblance to fibrils, which suggests that during the transition to
fibrils, major structural reorganizations are necessary. Considering the vast extent of the
conformational changes, it is likely that α-syn molecules within these oligomers
dissociate before incorporation into the fibril ("off pathway"), rather than for
reorganization to occur within the oligomeric structures "on pathway" to fibril formation.
119
This analysis of (in vitro) aggregation inducing conditions sets the stage for the study of
in vivo cellular influences on aggregation. Furthermore, knowing the structure of the toxic
oligomeric state of α-synuclein will provide a specific target for the design of drugs that
can block its pathogenic properties. In addition, an understanding of the disease-causing
structure is important to the understanding of its mechanism and function.
120
Chapter 5:
Conclusions and future directions
These studies demonstrate the importance of environmental conditions to the
modulation of structural transitions which occur in membrane-binding proteins and, as a
result, direct protein function. This principle was examined in two examples of
membrane-binding proteins, annexin B12 and α-synuclein. Both are able to take up
different solution and membrane-bound states.
Structural studies of the Ca
2+
-dependent membrane-bound state of annexin B12
The studies presented in Chapter 1 together with previous studies in the lab provides a
comprehensive analysis of the Ca
2+
-dependent membrane-bound state of anxB12 (Isas
et al., 2005; Isas et al., 2002; Isas et al., 2004; Isas et al., 2003; Langen et al., 1998b). In
addition to the identification of loop residues which are directly involved in membrane
interaction (shown in Chapter 1), EPR studies have also demonstrated that the Ca
2+
-
dependent membrane-bound state of anxB12 is consistent with the crystal structure, and
that the overall backbone fold remains the same (Isas et al., 2002; Isas et al., 2004). For
the Ca
2+
-dependent membrane-bound state (surface-bound trimer), the binding and
sequestering of Ca
2+
ions in membrane interactions could be a mechanism of regulating
local Ca
2+
concentrations in the cell. In addition, the strong immobilization observed for
all loop regions indicated that the surrounding phospholipid molecules were also
immobilized, also supporting the idea that annexins can act as molecular fences that
prevent or hinder membrane diffusion.
121
Ca
2+
-independent transmembrane state of anxB12 at acidic pH.
In the absence of Ca
2+
and at low pH, anxB12 refolds into a transmembrane monomer,
which has different functional implications compared to the Ca
2+
-dependent surface-
bound trimer. For the transmembrane state of anxB12, the presence of two membrane-
spanning amphipathic helices supports the proposed annexin function as an ion channel,
possibly to regulate Ca
2+
. Sequence analysis of other helical hairpin regions suggests
that these other regions would also capable of refolding into continuous amphipathic
helices. Additional studies are needed to confirm the presence of these helices in the
transmembrane state, their arrangement in the membrane, and to elucidate the exact
function and physiological relevance of the transmembrane state.
Although annexins are primarily cytosolic proteins, their versatility in adapting different
membrane-bound structural conformations suggests that they can adjust readily to
localized changes in the cellular environment to alter their function accordingly. The
ability of anxB12 (and the human homolog, anxA5) to reversibly and interchangeably
take up different membrane-bound (i.e., surface-bound trimer, transmembrane
monomer) and solutions states has been demonstrated in vitro with vesicles and
different buffer conditions (Isas et al., 2000), however this has not been demonstrated
for the cellular environment. In addition anxB12 has been shown to bind membranes in a
curvature-dependent manner (Fischer et al., 2007), suggesting that they may have a
physiological role in regulating or maintaining membrane curvature. Further studies are
needed to describe the interactions of annexins in the cellular environment, and the role
of these interactions in their biological function.
122
Applications of polarity sensitive indicator of cell viability and apoptosis (pSIVA)
Using the data obtained from the structural studies of anxB12 in the Ca
2+
-dependent
membrane-bound state I showed the design and application of a polarity-sensitive
annexin-based biosensor (pSIVA) for live-cell imaging of apoptosis and degeneration of
cells in culture and in vivo. Although the primary goal of this study (Chapter 3) was the
development of the method, pSIVA can be used in future studies to investigate the
underlying biological mechanisms which regulate cell death, particularly in the early
stages of apoptotic signaling before the cell has been committed to death. This allows
pSIVA to be applicable to studying processes that do not necessarily result in cell death,
such as axonal pruning and degeneration. Additionally, pSIVA is a suitable biosensor for
predicting cell fate, and can be used to test conditions which promote apoptotic or
survival signaling in disease, development and cell homeostasis. For example, in the
studies of rescue from neuronal degeneration (Chapter 3, Fig. 3-9), I showed that pSIVA
binds to neurons induced to undergo apoptosis, and binding and fluorescence reverses
as the neurons regain health. To further investigate the interplay between the factors
which promote degeneration and survival in neurons pSIVA can be used in combination
with other cellular markers to elucidate the underlying biological processes which are
involved, and their timing. Furthermore, the critical period in which rescue from cell death
is possible can be better described in terms of the signaling processes that have been
activated, reversed or carried out. Additional studies are also needed to clarify the
biological mechanisms which control PS exposure during apoptosis and degeneration.
During neuronal degeneration (Chapter 3, Figs. 3-6, 3-7), PS exposure on the axons
appeared progressively along the axon (toward to cell body or axon terminals) and in a
punctate pattern, which suggests that PS exposure is a regulated process. Further
123
studies are required to determine the cellular mechanisms that coordinate this process
and relate PS exposure to the other processes involved in neuronal degeneration.
α-synuclein oligomers
α-synuclein is primarily described as a natively unfolded cytosolic protein that is
predominantly localized to the presynaptic nerve terminals. The monomeric, helical
membrane-bound state of α-synuclein is thought to be correlated with its physiological
vesicle-associated function. However, aggregation of α-synuclein into oligomeric and
fibrillar structures has been implicated in the pathogenesis of Parkinson’s disease.
Furthermore the oligomeric state of α-synuclein has been shown to interact with and
disrupt membranes, indicating a possible mechanism for oligomer-mediated cell toxicity.
In Chapter 4, I described the structures of different populations of oligomeric α-synuclein
aggregates, which were induced by different environmental conditions (pH, buffer,
protein concentration, etc.), which were characterized by EPR and other biochemical
techniques. Additional studies are still needed to determine the exact 3-D structures for
the oligomeric state, which will involve measuring EPR spectra of other residue positions,
accessibility measurements to determine surface-exposed vs. buried residues, and
distance measurements with double-labeled mutants to determine the distances
between key residues. Furthermore, additional studies are needed to determine
membrane interaction, and to describe the structural conformation(s) that are involved in
membrane disruption. This can be tested using vesicles of known lipid compositions to
determine which lipid environments are most vulnerable to interaction with the α-
synuclein oligomers. Moreover, further studies are needed to determine the
physiological and pathological relevance of the structures described in these studies,
124
and relate how changes in the local cellular environment might affect the misfolding and
aggregation pathway of α-synuclein.
125
References
Altenbach, C., Greenhalgh, D.A., Khorana, H.G., and Hubbell, W.L. (1994). A Collision
Gradient-Method to Determine the Immersion Depth of Nitroxides in Lipid Bilayers -
Application to Spin-Labeled Mutants of Bacteriorhodopsin. P Natl Acad Sci USA 91,
1667-1671.
Bretscher, M.S. (1972). Asymmetrical lipid bilayer structure for biological membranes.
Nat New Biol 236, 11-12.
Chan, A., Reiter, R., Wiese, S., Fertig, G., and Gold, R. (1998). Plasma membrane
phospholipid asymmetry precedes DNA fragmentation in different apoptotic cell models.
Histochem Cell Biol 110, 553-558.
Chan, J.R., Watkins, T.A., Cosgaya, J.M., Zhang, C., Chen, L., Reichardt, L.F., Shooter,
E.M., and Barres, B.A. (2004). NGF controls axonal receptivity to myelination by
Schwann cells or oligodendrocytes. Neuron 43, 183-191.
Chen, M., Margittai, M., Chen, J., and Langen, R. (2007). Investigation of alpha-
synuclein fibril structure by site-directed spin labeling. J Biol Chem 282, 24970-24979.
Choi, W., Zibaee, S., Jakes, R., Serpell, L.C., Davletov, B., Crowther, R.A., and Goedert,
M. (2004). Mutation E46K increases phospholipid binding and assembly into filaments of
human alpha-synuclein. FEBS Lett 576, 363-368.
Cohen, G.M. (1997). Caspases: the executioners of apoptosis. Biochem J 326, 1-16.
Conway, K.A., Lee, S.J., Rochet, J.C., Ding, T.T., Williamson, R.E., and Lansbury, P.T.,
Jr. (2000). Acceleration of oligomerization, not fibrillization, is a shared property of both
alpha-synuclein mutations linked to early-onset Parkinson's disease: implications for
pathogenesis and therapy. Proc Natl Acad Sci U S A 97, 571-576.
Daher, J., Ying, M., Banerjee, R., McDonald, R., Hahn, M., Yang, L., Flint Beal, M.,
Thomas, B., Dawson, V., Dawson, T., et al. (2009). Conditional transgenic mice
expressing C-terminally truncated human alpha-synuclein (alphaSyn119) exhibit
reduced striatal dopamine without loss of nigrostriatal pathway dopaminergic neurons.
Molecular Neurodegeneration 4, 34.
Dawson, T.M., and Dawson, V.L. (2003). Molecular pathways of neurodegeneration in
Parkinson's disease. Science 302, 819-822.
126
Der-Sarkissian, A., Jao, C.C., Chen, J., and Langen, R. (2003). Structural organization
of alpha-synuclein fibrils studied by site-directed spin labeling. J Biol Chem 278, 37530-
37535.
Fadok, V.A., Voelker, D.R., Campbell, P.A., Cohen, J.J., Bratton, D.L., and Henson, P.M.
(1992). Exposure of phosphatidylserine on the surface of apoptotic lymphocytes triggers
specific recognition and removal by macrophages. J Immunol 148, 2207-2216.
Faure, A.V., Migne, C., Devilliers, G., and Ayala-Sanmartin, J. (2002). Annexin 2
"secretion" accompanying exocytosis of chromaffin cells: possible mechanisms of
annexin release. Exp Cell Res 276, 79-89.
Fischer, T., Lu, L., Haigler, H.T., and Langen, R. (2007). Annexin B12 is a sensor of
membrane curvature and undergoes major curvature-dependent structural changes. J
Biol Chem 282, 9996-10004.
Forno, L.S. (1996). Neuropathology of Parkinson's disease. J Neuropathol Exp Neurol
55, 259-272.
Gavrieli, Y., Sherman, Y., and Ben-Sasson, S.A. (1992). Identification of programmed
cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol 119, 493-
501.
Genge, B.R., Wu, L.N., Adkisson, H.D.t., and Wuthier, R.E. (1991). Matrix vesicle
annexins exhibit proteolipid-like properties. Selective partitioning into lipophilic solvents
under acidic conditions. Journal of Biological Chemistry 266, 10678-10685.
Gerke, V., Creutz, C.E., and Moss, S.E. (2005). Annexins: linking Ca2+ signalling to
membrane dynamics. Nat Rev Mol Cell Biol 6, 449-461.
Gerke, V., and Moss, S.E. (2002). Annexins: From Structure to Function. Physiol Rev 82,
331-371.
Geske, F.J., Lieberman, R., Strange, R., and Gerschenson, L.E. (2001). Early stages of
p53-induced apoptosis are reversible. Cell Death Differ 8, 182-191.
Gilmanshin, R., Creutz, C.E., and Tamm, L.K. (1994). Annexin IV reduces the rate of
lateral lipid diffusion and changes the fluid phase structure of the lipid bilayer when it
binds to negatively charged membranes in the presence of calcium. Biochemistry 33,
8225-8232.
127
Golczak, M., Kicinska, A., Bandorowicz-Pikula, J., Buchet, R., Szewczyk, A., and Pikula,
S. (2001). Acidic pH-induced folding of annexin VI is a prerequisite for its insertion into
lipid bilayers and formation of ion channels by the protein molecules. Faseb J 15, 1083-
1085.
Goldstein, J.C., Waterhouse, N.J., Juin, P., Evan, G.I., and Green, D.R. (2000). The
coordinate release of cytochrome c during apoptosis is rapid, complete and kinetically
invariant. Nat Cell Biol 2, 156-162.
Gottlieb, R.A., and Granville, D.J. (2002). Analyzing mitochondrial changes during
apoptosis. Methods 26, 341-347.
Green, D.R., and Reed, J.C. (1998). Mitochondria and Apoptosis. Science 281, 1309-
1312.
Gross, A., Columbus, L., Hideg, K., Altenbach, C., and Hubbell, W.L. (1999). Structure of
the KcsA potassium channel from Streptomyces lividans: A site-directed spin labeling
study of the second transmembrane segment. Biochemistry 38, 10324-10335.
Hammill, A.K., Uhr, J.W., and Scheuermann, R.H. (1999). Annexin V staining due to loss
of membrane asymmetry can be reversible and precede commitment to apoptotic death.
Exp Cell Res 251, 16-21.
Hayes, M.J., and Moss, S.E. (2004). Annexins and disease. Biochemical and
Biophysical Research Communications 322, 1166-1170.
Hegde, B.G., Isas, J.M., Zampighi, G., Haigler, H.T., and Langen, R. (2006). A novel
calcium-independent peripheral membrane-bound form of annexin B12. Biochemistry 45,
934-942.
Hirota, N., Mizuno, K., and Goto, Y. (1997). Cooperative alpha-helix formation of beta-
lactoglobulin and melittin induced by hexafluoroisopropanol. Protein Sci 6, 416-421.
Hubbell, W.L., Cafiso, D.S., and Altenbach, C. (2000). Identifying conformational
changes with site-directed spin labeling. Nat Struct Biol 7, 735-739.
Hubbell, W.L., Gross, A., Langen, R., and Lietzow, M.A. (1998). Recent advances in
site-directed spin labeling of proteins. Curr Opin Struct Biol 8, 649-656.
128
Isas, J.M., Cartailler, J.P., Sokolov, Y., Patel, D.R., Langen, R., Luecke, H., Hall, J.E.,
and Haigler, H.T. (2000). Annexins V and XII insert into bilayers at mildly acidic pH and
form ion channels. Biochemistry 39, 3015-3022.
Isas, J.M., Kim, Y.E., Jao, C.C., Hegde, P.B., Haigler, H.T., and Langen, R. (2005).
Calcium- and membrane-induced changes in the structure and dynamics of three helical
hairpins in Annexin B12. Biochemistry 44, 16435-16444.
Isas, J.M., Langen, R., Haigler, H.T., and Hubbell, W.L. (2002). Structure and dynamics
of a helical hairpin and loop region in annexin 12: A site-directed spin labeling study.
Biochemistry 41, 1464-1473.
Isas, J.M., Langen, R., Hubbell, W.L., and Haigler, H.T. (2004). Structure and dynamics
of a helical hairpin that mediates calcium-dependent membrane binding of Annexin B12.
J Biol Chem 279, 32492-32498.
Isas, J.M., Patel, D.R., Jao, C., Jayasinghe, S., Cartailler, J.P., Haigler, H.T., and
Langen, R. (2003). Global structural changes in annexin 12. The roles of phospholipid,
Ca2+, and pH. J Biol Chem 278, 30227-30234.
Jao, C.C., Der-Sarkissian, A., Chen, J., and Langen, R. (2004). Structure of membrane-
bound alpha-synuclein studied by site-directed spin labeling. P Natl Acad Sci USA 101,
8331-8336.
Karpinich, N.O., Tafani, M., Rothman, R.J., Russo, M.A., and Farber, J.L. (2002). The
course of etoposide-induced apoptosis from damage to DNA and p53 activation to
mtochondrial release of cytochrome c. J Biol Chem 277, 16547-16552.
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.
Kerr, J.F., Wyllie, A.H., and Currie, A.R. (1972). Apoptosis: a basic biological
phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 26, 239-257.
Kim, Y.E., Isas, J.M., Haigler, H.T., and Langen, R. (2005). A helical hairpin region of
soluble annexin B12 refolds and forms a continuous transmembrane helix at mildly
acidic pH. J Biol Chem 280, 32398-32404.
129
Kiprianova, I., Freiman, T.M., Desiderato, S., Schwab, S., Galmbacher, R., Gillardon, F.,
and Spranger, M. (1999). Brain-derived neurotrophic factor prevents neuronal death and
glial activation after global ischemia in the rat. J Neurosci Res 56, 21-27.
Knowles, P.F., Marsh, D., and Rattle, H.W.E. (1976). Magnetic resonance of
biomolecules : an introduction to the theory and practice of NMR and ESR in biological
systems (London ; New York, Wiley).
Kohler, G., Hering, U., Zschoring, O., and Arnold, K. (1997). Annexin V interaction with
phosphatidylserine-containing vesicles at low and neutral pH. Biochemistry 36, 8189-
8194.
Koopman, G., Reutelingsperger, C.P., Kuijten, G.A., Keehnen, R.M., Pals, S.T., and van
Oers, M.H. (1994). Annexin V for flow cytometric detection of phosphatidylserine
expression on B cells undergoing apoptosis. Blood 84, 1415-1420.
Koradi, R., Billeter, M., and Wuthrich, K. (1996). MOLMOL: a program for display and
analysis of macromolecular structures. J Mol Graph 14, 51-55, 29-32.
Kruger, R., Kuhn, W., Muller, T., Woitalla, D., Graeber, M., Kosel, S., Przuntek, H.,
Epplen, J.T., Schols, L., and Riess, O. (1998). Ala30Pro mutation in the gene encoding
alpha-synuclein in Parkinson's disease. Nat Genet 18, 106-108.
Ladokhin, A.S., and Haigler, H.T. (2005). Reversible transition between the surface
trimer and membrane-inserted monomer of annexin 12. Biochemistry 44, 3402-3409.
Ladokhin, A.S., Isas, J.M., Haigler, H.T., and White, S.H. (2002). Determining the
membrane topology of proteins: Insertion pathway of a transmembrane helix of annexin
12. Biochemistry 41, 13617-13626.
Langen, R., Isas, J.M., Hubbell, W.L., and Haigler, H.T. (1998a). A transmembrane form
of annexin XII detected by site-directed spin labeling. P Natl Acad Sci USA 95, 14060-
14065.
Langen, R., Isas, J.M., Luecke, H., Haigler, H.T., and Hubbell, W.L. (1998b). Membrane-
mediated assembly of annexins studied by site-directed spin labeling. J Biol Chem 273,
22453-22457.
Lashuel, H.A. (2005). Membrane permeabilization: a common mechanism in protein-
misfolding diseases. Sci Aging Knowledge Environ 2005, pe28.
130
Li, J., and Yuan, J. (2008). Caspases in apoptosis and beyond. Oncogene 27, 6194-
6206.
Liemann, S., and Huber, R. (1997). Three-dimensional structure of annexins. Cell Mol
Life Sci 53, 516-521.
Luecke, H., Chang, B.T., Mailliard, W.S., Schlaepfer, D.D., and Haigler, H.T. (1995).
Crystal structure of the annexin XII hexamer and implications for bilayer insertion. Nature
378, 512-515.
Maiese, K., and Vincent, A.M. (2000). Membrane asymmetry and DNA degradation:
functionally distinct determinants of neuronal programmed cell death. J Neurosci Res 59,
568-580.
Mailliard, W.S., Leucke, H., and Haigler, H.T. (1997). Annexin XII forms calcium-
dependent multimers in solution and on phospholipid bilayers: A chemical cross-linking
study. Biochemistry 36, 9045-9050.
Martin, S.J., Reutelingsperger, C.P., McGahon, A.J., Rader, J.A., van Schie, R.C.,
LaFace, D.M., and Green, D.R. (1995). Early redistribution of plasma membrane
phosphatidylserine is a general feature of apoptosis regardless of the initiating stimulus:
inhibition by overexpression of Bcl-2 and Abl. J Exp Med 182, 1545-1556.
Masliah, E., Rockenstein, E., Veinbergs, I., Mallory, M., Hashimoto, M., Takeda, A.,
Sagara, Y., Sisk, A., and Mucke, L. (2000). Dopaminergic loss and inclusion body
formation in alpha-synuclein mice: implications for neurodegenerative disorders. Science
287, 1265-1269.
Mchaourab, H.S., Lietzow, M.A., Hideg, K., and Hubbell, W.L. (1996). Motion of spin-
labeled side chains in T4 lysozyme, correlation with protein structure and dynamics.
Biochemistry 35, 7692-7704.
Megli, F.M., Selvaggi, M., Liemann, S., Quagliariello, E., and Huber, R. (1998). The
calcium-dependent binding of annexin V to phospholipid vesicles influences the bilayer
inner fluidity gradient. Biochemistry 37, 10540-10546.
Mirnikjoo, B., Balasubramanian, K., and Schroit, A.J. (2009). Mobilization of Lysosomal
Calcium Regulates the Externalization of Phosphatidylserine during Apoptosis. J Biol
Chem 284, 6918-6923.
Moss, S., and Morgan, R. (2004). The annexins. Genome Biol 5, 219-225.
131
Nicholson, D.W., and Thornberry, N.A. (1997). Caspases: killer proteases. Trends
Biochem Sci 22, 299-306.
Nikolaev, A., McLaughlin, T., O'Leary, D.D., and Tessier-Lavigne, M. (2009). APP binds
DR6 to trigger axon pruning and neuron death via distinct caspases. Nature 457, 981-
989.
Oling, F., Bergsma-Schutter, W., and Brisson, A. (2001). Trimers, dimers of trimers, and
trimers of trimers are common building blocks of annexin a5 two-dimensional crystals. J
Struct Biol 133, 55-63.
Peng, S., Publicover, N.G., Airey, J.A., Hall, J.E., Haigler, H.T., Jiang, D., Chen, S.R.,
and Sutko, J.L. (2004). Diffusion of single cardiac ryanodine receptors in lipid bilayers is
decreased by annexin 12. Biophys J 86, 145-151.
Polymeropoulos, M.H., Lavedan, C., Leroy, E., Ide, S.E., Dehejia, A., Dutra, A., Pike, B.,
Root, H., Rubenstein, J., Boyer, R., et al. (1997). Mutation in the alpha-synuclein gene
identified in families with Parkinson's disease. Science 276, 2045-2047.
Raff, M.C., Whitmore, A.V., and Finn, J.T. (2002). Axonal self-destruction and
neurodegeneration. Science 296, 868-871.
Rand, J.H. (1999). "Annexinopathies" -- A New Class of Diseases. N Engl J Med 340,
1035-1036.
Reeves, J.P., and Dowben, R.M. (1969). Formation and properties of thin-walled
phospholipid vesicles. J Cell Physiol 73, 49-60.
Rescher, U., and Gerke, V. (2004). Annexins - unique membrane binding proteins with
diverse functions. J Cell Sci 117, 2631-2639.
Reviakine, I., Bergsma-Schutter, W., and Brisson, A. (1998). Growth of Protein 2-D
Crystals on Supported Planar Lipid Bilayers Imagedin Situby AFM. Journal of Structural
Biology 121, 356-361.
Reviakine, I., Bergsma-Schutter, W., Mazeres-Dubut, C., Govorukhina, N., and Brisson,
A. (2000). Surface Topography of the p3 and p6 Annexin V Crystal Forms Determined
by Atomic Force Microscopy. Journal of Structural Biology 131, 234-239.
132
Rimon, G., Bazenet, C.E., Philpott, K.L., and Rubin, L.L. (1997). Increased surface
phosphatidylserine is an early marker of neuronal apoptosis. J Neurosci Res 48, 563-
570.
Roccatano, D., Fioroni, M., Zacharias, M., and Colombo, G. (2005). Effect of
hexafluoroisopropanol alcohol on the structure of melittin: a molecular dynamics
simulation study. Protein Sci 14, 2582-2589.
Rosenberg, S.S., Kelland, E.E., Tokar, E., De la Torre, A.R., and Chan, J.R. (2008). The
geometric and spatial constraints of the microenvironment induce oligodendrocyte
differentiation. Proc Natl Acad Sci U S A 105, 14662-14667.
Rosengarth, A., Wintergalen, A., Galla, H.J., Hinz, H.J., and Gerke, V. (1998). Ca2+-
independent interaction of annexin I with phospholipid monolayers. FEBS Lett 438, 279-
284.
Saurel, O., Cezanne, L., Milon, A., Tocanne, J.F., and Demange, P. (1998). Influence of
annexin V on the structure and dynamics of phosphatidylcholine/phosphatidylserine
bilayers: a fluorescence and NMR study. Biochemistry 37, 1403-1410.
Saxena, S., and Caroni, P. (2007). Mechanisms of axon degeneration: from
development to disease. Prog Neurobiol 83, 174-191.
Seaton, B.A. (1996). Annexins: Molecular Structure to Cellular Function Seaton, B. A.
(ed ) (Austin,Texas, R.G. Landes Company).
Sharon, R., Bar-Joseph, I., Frosch, M.P., Walsh, D.M., Hamilton, J.A., and Selkoe, D.J.
(2003). The formation of highly soluble oligomers of alpha-synuclein is regulated by fatty
acids and enhanced in Parkinson's disease. Neuron 37, 583-595.
Singleton, A.B., Farrer, M., Johnson, J., Singleton, A., Hague, S., Kachergus, J., Hulihan,
M., Peuralinna, T., Dutra, A., Nussbaum, R., et al. (2003). alpha-Synuclein locus
triplication causes Parkinson's disease. Science 302, 841.
Spillantini, M.G., Crowther, R.A., Jakes, R., Hasegawa, M., and Goedert, M. (1998).
alpha-Synuclein in filamentous inclusions of Lewy bodies from Parkinson's disease and
dementia with lewy bodies. Proc Natl Acad Sci U S A 95, 6469-6473.
Spillantini, M.G., Schmidt, M.L., Lee, V.M., Trojanowski, J.Q., Jakes, R., and Goedert, M.
(1997). Alpha-synuclein in Lewy bodies. Nature 388, 839-840.
133
Tang, H.L., Yuen, K.L., Tang, H.M., and Fung, M.C. (2009). Reversibility of apoptosis in
cancer cells. Br J Cancer 100, 118-122.
Torok, M., Milton, S., Kayed, R., Wu, P., McIntire, T., Glabe, C.G., and Langen, R.
(2002). Structural and Dynamic Features of Alzheimer's Abeta Peptide in Amyloid Fibrils
Studied by Site-directed Spin Labeling. J Biol Chem 277, 40810-40815.
Uversky, V.N. (2003). A protein-chameleon: conformational plasticity of alpha-synuclein,
a disordered protein involved in neurodegenerative disorders. J Biomol Struct Dyn 21,
211-234.
Vargas, M.E., and Barres, B.A. (2007). Why is Wallerian degeneration in the CNS so
slow? Annu Rev Neurosci 30, 153-179.
Verkleij, A.J., Zwaal, R.F., Roelofsen, B., Comfurius, P., Kastelijn, D., and van Deenen,
L.L. (1973). The asymmetric distribution of phospholipids in the human red cell
membrane. A combined study using phospholipases and freeze-etch electron
microscopy. Biochim Biophys Acta 323, 178-193.
Vermes, I., Haanen, C., Steffens-Nakken, H., and Reutelingsperger, C. (1995). A novel
assay for apoptosis. Flow cytometric detection of phosphatidylserine expression on early
apoptotic cells using fluorescein labelled Annexin V. J Immunol Methods 184, 39-51.
Voges, D., Berendes, R., Burger, A., Demange, P., Baumeister, W., and Huber, R.
(1994). Three-dimensional Structure of Membrane-bound Annexin V : A Correlative
Electron Microscopy-X-ray Crystallography Study. Journal of Molecular Biology 238,
199-213.
Volles, M.J., and Lansbury, P.T., Jr. (2003). Zeroing in on the pathogenic form of alpha-
synuclein and its mechanism of neurotoxicity in Parkinson's disease. Biochemistry 42,
7871-7878.
Volles, M.J., Lee, S.J., Rochet, J.C., Shtilerman, M.D., Ding, T.T., Kessler, J.C., and
Lansbury, P.T., Jr. (2001). Vesicle permeabilization by protofibrillar alpha-synuclein:
implications for the pathogenesis and treatment of Parkinson's disease. Biochemistry 40,
7812-7819.
White, S.H. (2009). Biophysical dissection of membrane proteins. Nature 459, 344-346.
Wimley, W.C., and White, S.H. (1996). Experimentally determined hydrophobicity scale
for proteins at membrane interfaces. Nat Struct Biol 3, 842-848.
134
Wyllie, A.H. (1980). Glucocorticoid-induced thymocyte apoptosis is associated with
endogenous endonuclease activation. Nature 284, 555-556.
Wyllie, A.H., Kerr, J.F., and Currie, A.R. (1980). Cell death: the significance of apoptosis.
Int Rev Cytol 68, 251-306.
Zarranz, J.J., Alegre, J., Gomez-Esteban, J.C., Lezcano, E., Ros, R., Ampuero, I., Vidal,
L., Hoenicka, J., Rodriguez, O., Atares, B., et al. (2004). The new mutation, E46K, of
alpha-synuclein causes Parkinson and Lewy body dementia. Ann Neurol 55, 164-173.
Abstract (if available)
Abstract
The interaction between proteins and membranes underlies many important biological processes. Proteins influence the structure and fluidity of membranes and likewise, membranes promote structural reorganization in proteins, mutually modifying their functions. In addition, protein-membrane interactions can lead to protein misfolding and the pathogenesis of disease. Despite the importance of these dynamic protein-membrane interactions, less is known about the structure of membrane-bound proteins in comparison to proteins in solution. As an alternative to conventional methods such as X-ray crystallography and NMR, we used electron paramagnetic resonance (EPR) spectroscopy in combination with site-directed spin labeling (SDSL) as the principle technique to study how different environmental conditions can promote the structural transitions that occur in two different membrane-binding proteins, annexin B12 (anxB12, anxXII) and α-synuclein (α-syn). AnxB12 is a member of a family of structurally conserved proteins that are abundantly expressed in most eukaryotic cells. The various proposed functions include vesicle trafficking, membrane fusion, ion channel formation, and cell signaling, indicating that membrane interactions are important for all annexin functions. In order to gain a better understanding of the mechanism of membrane binding and annexin function, the structures of two different membrane-bound states of AnxB12 were studied. In addition, using the data obtained from these studies, we engineered a polarity sensitive annexin-based biosensor applicable to the real-time detection of apoptotic membrane changes in cultured cells and in vivo. The second protein of interest, α-syn, is a natively unfolded cytosolic protein which is observed to bind preferentially to small vesicles (of similar size to synaptic vesicles) and is localized to the presynaptic nerve terminals.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Membrane curvature sensors and inducers studied by site-directed spin labeling
PDF
Structural studies of the IAPP membrane-mediated aggregation pathway
PDF
The structure and function of membrane curving proteins on different membrane shapes and their regulation by post-translational modifications
PDF
Structural features and modifiers of islet amyloid polypeptide: implications for type II diabetes mellitus
PDF
Controlling membrane protein folding using photoresponsive surfactant
PDF
Controlling membrane protein folding with light illumination and catanionic surfactant systems
PDF
Enhancing and inhibiting diabetic amyloid misfolding
PDF
Before they were amyloid: understanding the toxicity of disease-associated monomers and oligomers prior to their aggregation
PDF
Uncovering the protective role of protein glycosylation in Parkinson's disease utilizing protein semi-synthesis
PDF
Protein phosphatase 2A and annexin A5: modulators of cellular functions
PDF
A novel construct to study the pulsatility of insulin secretion in single cells, islets and whole pancreas
PDF
Dual effects of transmembrane proline residues on integrin function
PDF
Oligomer formation of functional amyloid protein - Orb2A
PDF
Molecular and computational analysis of spin-labeled nucleic acids and proteins
PDF
The effect of familial mutants of Parkinson's disease on membrane remodeling
PDF
Physical principles of membrane mechanics, membrane domain formation, and cellular signal transduction
PDF
Site-specific effects of ubiquitin and ubiquitin-like modifier proteins on α-synuclein aggregation
PDF
Novel synthesis of β-glycosides for SPPS of GLCNAC glycoproteins and study of their site-specific biochemical and biophysical consequences
PDF
Molecular basis of CD33 receptor function, and effects of a destabilizing transmembrane motif on αXβ2 integrin
PDF
Amelogenin-ameloblastin protein interaction and function in dental enamel formation
Asset Metadata
Creator
Kim, Yujin E. (author)
Core Title
The folding, misfolding and refolding of membrane proteins and the design of a phosphatidylserine-specific membrane sensor
School
Keck School of Medicine
Degree
Doctor of Philosophy
Degree Program
Biochemistry and Molecular Biology
Degree Conferral Date
2009-12
Publication Date
10/29/2009
Defense Date
09/25/2009
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
EPR,membrane binding,membrane protein structure,OAI-PMH Harvest,oligomers,protein misfolding,α-synuclein
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Langen, Ralf (
committee chair
), Chen, Jeannie (
committee member
), Chow, Robert HP. (
committee member
), Haworth, Ian S. (
committee member
), Markland, Francis (
committee member
)
Creator Email
yujinkim@usc.edu,yujinkim820@hotmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m2700
Unique identifier
UC1208955
Identifier
etd-Kim-3342 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-268665 (legacy record id),usctheses-m2700 (legacy record id)
Legacy Identifier
etd-Kim-3342.pdf
Dmrecord
268665
Document Type
Dissertation
Rights
Kim, Yujin E.
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
EPR
membrane binding
membrane protein structure
oligomers
protein misfolding
α-synuclein