“Cu, Zn Superoxide Dismutase Misfolding in Amyotrophic ... · Cu/Zn Superoxide Dismutase Misfolding in Amyotrophic Lateral Sclerosis Rishi Rakhit Doctor of Philosophy Graduate Department
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
“Cu, Zn Superoxide Dismutase Misfolding in Amyotrophic Lateral Sclerosis”
By
Rishi Rakhit
A thesis submitted in conformity with the requirements
Graduate Department of Biochemistry University of Toronto
2009
Amyotrophic lateral sclerosis (ALS) is characterized by motor neuron degeneration
resulting in progressive paralysis and death. The only known cause of typical ALS is
mutations in SOD1; these predominantly missense mutations produce a toxic gain-of-
function in the enzyme Cu/Zn superoxide dismutase (SOD1). The prevailing hypotheses
regarding the mechanism of toxicity were a) oxidative damage from aberrant SOD1
redox chemistry, and b) misfolding of the mutant protein. The goal of this thesis was to
investigate the molecular mechanisms of the mutant SOD1 (mSOD1) misfolding and
toxicity.
We proposed that oxidative damage to SOD1 itself could cause its misfolding and
aggregation. To investigate this hypothesis, we subjected purified SOD1 in vitro to metal
catalyzed oxidation. Oxidation of SOD1 produced aggregates reminiscent of those
observed in ALS pathology. Aggregation propensity of zinc-deficient SOD1 and several
mSOD1s known to have lower zinc-binding affinity was proportional to partial
unfolding. Oxidation of SOD1 caused conversion of several His residues to 2-oxo-
histidine.
iii
Because oxidation of SOD1 primarily affected the metal-binding His residues, we
hypothesized that oxidation of wild-type, holo-SOD1 should lead to aggregation.
Increasing the concentration of wild-type SOD1 in oxidation reactions produced
aggregates similar to those observed earlier. Both wild-type and mSOD1 aggregation
kinetics revealed an initial decrease in particle size rather than a monotonic increase
using dynamic light scattering. This was consistent with the conversion of SOD1,
normally an obligate homodimer, into monomers prior to aggregation. This observation
was confirmed using analyatical ultracentrifugation. The common aggregation pathway
for wild-type and mSOD1 suggested a mechanism for sporadic ALS caused by SOD1
misfolding.
To interrogate the in vivo misfolding pathway of SOD1, we used its high-
resolution structure to create an antibody that reacts with monomer/misfolded SOD1 but
not the native dimer. Upon verifying the reactivity of this antibody, we showed that
monomer/misfolded SOD1 is found in a human case of familial ALS and in transgenic
animal models of ALS. Misfolded SOD1 is found primarily in affected cells, motor
neurons. Misfolded SOD1 is also initially absent, but appears prior to symptom onset.
These observations together suggest a causal role for SOD1 misfolding through a
monomeric intermediate in ALS pathogenesis.
Acknowledgements
Everyone who undertakes and completes a PhD knows that the path is not without its
sometimes winding curves. For some, like me, it takes longer to find the way. I thank my
supervisor, Avi Chakrabartty, for taking a chance on me when I needed someone to give
me a break. I thank him for his excellent mentorship, his honesty and for making his lab
an open, exciting place to work. His support, advice and friendship have made the past
years a most enjoyable experience. I also would like to thank my supervisory committee
members, John Glover, Drew Woolley and the late Jim Lepock, for their time and
encouragement over the past five years.
Everyone in the Chakrabartty lab, past and present has enriched my experience
through open debate, thoughtful discussion and friendship. I would particularly like to
thank: Meng Guo, Paul Gorman and Pharhad (Eli) Arselan for treating me like equals
when I was junior; Sandy Go for showing me the ropes and for fun times; my friend,
Sylvia Ho, for helping me in life both inside and outside the lab- I think we grew up
together a bit; Kevin Hadley, with whom I have worked closely in the past few years; and
my summer students- Yi-Ting Chen, Sylvain Helas-Othenin and Alyssa Wong- for
listening to what I had to say.
The work presented in this thesis could not have been completed without the
expertise and effort of all my collaborators over the years. I would like to thank John
Crow for providing reagents that would prove critical in this work; Neil Cashman for
helpful advice, comments and letting me hang out in his lab for some months; Janice
Robertson, for her insight into the mouse models of ALS and her unfailing support; the
iv
team at Caprion Lifesciences for performing mass spectrometry; Don Cleveland and
Christine Vande Velde for their help in establishing the connection of misfolded SOD1
with mitochondria; and the late Patrick Horne for excellent technical expertise in
immunohistochemistry.
My family has been in an invaluable support to me throughout my life. I would
like to thank my mother, Alo, for all her love, my father, Shantanu, for his guidance, and
my sister, Riya, for watching out for me always. Lastly, I would like to thank my partner,
Martha MacDonald, for believing in me, for all her love and encouragement since before
I even began graduate school. Thank you.
v
Table of Contents
STRUCTURE, FOLDING, AND MISFOLDING OF CU,ZN SUPEROXIDE DISMUTASE IN AMYOTROPHIC LATERAL SCLEROSIS ............................................................................................. 2
MECHANISMS OF MUTANT SOD1 TOXICITY AND RECENT DEVELOPMENTS IN ALS RESEARCH................................................................................................................................................ 56
SUMMARY ................................................................................................................................................ 56 MUTANT SOD1 IN NON-NEURONAL CELLS ............................................................................................... 56 SUBCELLULAR DISTRIBUTION OF MUTANT SOD1..................................................................................... 59 TDP-43 IS A NOVEL COMPONENT OF UBIQUITINATED INCLUSION BODIES IN ALS .................................... 66 AUTOPHAGY IN MOTOR NEURON DISEASE ................................................................................................ 68 REFERENCES............................................................................................................................................. 69
OXIDATION-INDUCED MISFOLDING AND AGGREGATION OF SUPEROXIDE DISMUTASE AND ITS IMPLICATIONS FOR AMYOTROPHIC LATERAL SCLEROSIS.................................. 88
INTRODUCTION......................................................................................................................................... 90 MATERIALS AND METHODS...................................................................................................................... 93 RESULTS AND DISCUSSION ....................................................................................................................... 96 METAL CATALYZED OXIDATION OF SOD1 .............................................................................................. 96 FIGURE 1. ................................................................................................................................................. 98 CHARACTERIZATION OF OXIDATIVE MODIFICATION SITES ...................................................................... 99 MORPHOLOGY AND STRUCTURE OF SOD1 AGGREGATES......................................................................... 99 FIGURE 2. ............................................................................................................................................... 101 FIGURE 3. ............................................................................................................................................... 103 STRUCTURAL CHANGES TO SOD1 PRIOR TO AGGREGATION.................................................................. 104 FIGURE 4. ............................................................................................................................................... 105 FIGURE 5. ............................................................................................................................................... 106 CONCLUDING REMARKS......................................................................................................................... 107 ACKNOWLEDGEMENTS ........................................................................................................................... 107 REFERENCES........................................................................................................................................... 108
vi
MONOMERIC CU,ZN-SUPEROXIDE DISMUTASE IS A COMMON MISFOLDING INTERMEDIATE IN THE OXIDATION MODELS OF SPORADIC AND FAMILIAL AMYOTROPHIC LATERAL SCLEROSIS ......................................................................................... 116
SUMMARY .............................................................................................................................................. 117 INTRODUCTION....................................................................................................................................... 118 MATERIALS AND METHODS.................................................................................................................... 121 RESULTS................................................................................................................................................. 123 WILD TYPE SOD1 IS AGGREGATION-PRONE UNDER OXIDATIVE STRESS .............................................. 123 FIGURE. 1. .............................................................................................................................................. 125 FIGURE. 2. .............................................................................................................................................. 127 SOD1 DIMERS DISSOCIATE TO MONOMERS PRIOR TO AGGREGATION ................................................... 128 FIGURE. 3. .............................................................................................................................................. 129 SUPPORTING INFORMATION FIGURE 1 .................................................................................................... 130 MINOR CHANGES TO CONFORMATION OF SOD1 UPON OXIDATION ....................................................... 131 FIGURE. 4. .............................................................................................................................................. 132 SUPPORTING INFORMATION FIGURE 2. ................................................................................................... 133 MAPPING OF OXIDATIVE MODIFICATION SITES...................................................................................... 134 FIGURE. 5. .............................................................................................................................................. 135 FIGURE. 6. .............................................................................................................................................. 136 DISCUSSION............................................................................................................................................ 138 AGGREGATION OF WILD TYPE SOD1 PROVIDES A PLAUSIBLE MECHANISM FOR THE OCCURRENCE OF INCLUSION BODIES IN SPORADIC ALS.................................................................................................... 138 MONOMERIC INTERMEDIATE COMMON TO WILD TYPE AND MUTANT SOD1 AGGREGATION SUGGESTS A COMMON MECHANISM FOR THE PATHOLOGY OF SALS AND FALS........................................................ 140 COMMON THEMES IN SOD1 AGGREGATION IN ALS AND OTHER PROTEIN MISFOLDING DISEASES....... 141 FIGURE. 7. .............................................................................................................................................. 142 ACKNOWLEDGEMENTS ........................................................................................................................... 143 REFERENCES........................................................................................................................................... 143
AN IMMUNOLOGICAL EPITOPE SELECTIVE FOR PATHOLOGICAL MONOMER/MISFOLDED SOD1 IN ALS ........................................................................................... 151
SUMMARY .............................................................................................................................................. 152 INTRODUCTION....................................................................................................................................... 153 MATERIALS AND METHODS.................................................................................................................... 154 METHODS REFERENCES.......................................................................................................................... 161 RESULTS................................................................................................................................................. 162 ANTIBODY DESIGN AND VALIDATION .................................................................................................... 162 FIGURE 1. ............................................................................................................................................... 164 SUPPLEMENTARY FIGURE 1. ................................................................................................................... 166 SUPPLEMENTARY FIGURE 2. ................................................................................................................... 167 MONOMER/MISFOLDED SOD1 IN ALS-MOUSE MODELS ......................................................................... 168 SUPPLEMENTARY FIGURE 3. ................................................................................................................... 169 SELECTIVE DEPOSITION OF MONOMER/MISFOLDED SOD1...................................................................... 170 FIGURE 2. ............................................................................................................................................... 171 SUPPLEMENTARY FIGURE 4. ................................................................................................................... 173 SUBCELLULAR LOCALIZATION OF MONOMER/MISFOLDED SOD1 ........................................................... 174 FIGURE 3. ............................................................................................................................................... 175 SUPPLEMENTARY FIGURE 5. ................................................................................................................... 177 SUPPLEMENTARY FIGURE 6. ................................................................................................................... 178 WILD-TYPE SOD1 CAN MISFOLD IN VIVO ............................................................................................... 179 MONOMER/MISFOLDED SOD1 APPEARS PRIOR TO SYMPTOM ONSET AND CORRELATES WITH MOTOR NEURON LOSS ......................................................................................................................................... 180 MONOMER/MISFOLDED SOD1 IN A HUMAN CASE OF A4V SOD1-ALS.................................................. 180 FIGURE 4. ............................................................................................................................................... 181 DISCUSSION............................................................................................................................................ 182
Fourteen years after the discovery that mutations in Cu, Zn superoxide dismutase (SOD1)
cause a subset of familial amyotrophic lateral sclerosis (fALS), the mechanism by which
mutant SOD1 exerts toxicity remains unknown. The two principle hypotheses are a)
oxidative damage stemming from aberrant SOD1 redox chemistry, and b) misfolding of
the mutant protein. Here we review the structure and function of wild-type SOD1, as well
as the changes to the structure and function in mutant SOD1. The relative merits of the
two hypotheses are compared and a common unifying principle is outlined. Lastly, the
potential for therapies targeting SOD1 misfolding is discussed.
3
Introduction
Cu, Zn superoxide dismutase (SOD1) is a highly conserved enzyme that is the primary
cytoplasm scavenger of superoxide radical (O2-). In 1992, Rosen et al (1) discovered that
the ALS1 gene is SOD1 and that mutations in this gene are associated with amyotrophic
lateral sclerosis (ALS). SOD1 mutations are the most common known cause of ALS,
associated with 2-10% of cases(2); mutations in ALS2 (Alsin, a guanine nucleotide
exchange factor)(3), VAPB (vesicle associated membrane protein B)(4) and ANG
(angiogenin)(5) also account for a small number of cases. Despite more than a decade
since this discovery and more than 1200 publications relating SOD1 and ALS(6), the
causal mechanism behind SOD1 mediated ALS remains elusive. In recent years, a
mechanism involving SOD1 misfolding has gained prominence; this review will focus on
SOD1 structure and function, normal SOD1 folding, and the possible role for SOD1
misfolding in ALS. In each case, the properties of the normal wild-type protein will be
compared and contrasted with those of the mutant protein.
Enzymatic activity of SOD1:
Superoxide (O2-) is generated in a number of cellular processes (for a review,(7)),
including oxidative bursts from immune cells (primarily neutrophils) and as a by-product
of normal respiration. In this case, co-enzyme Q (ubiquinone) may reduce molecular
oxygen, instead of complex I or complex III, producing superoxide; this occurs in ~1% of
electron transport events(7). Superoxide can be produced, in vitro, through reduction of
oxygen in an electrolytic cell(8), by pulse radiolysis(9), or enzymatically with xanthine
and xanthine oxidase(10). The activity of xanthine oxidase on xanthine produces
superoxide as a byproduct of its enzymatic activity, based on its ability to reduce
4
cytochrome c in an oxygen dependent and catalase independent manner. Cu, Zn
superoxide dismutase (known earlier as erythrocuprein) has long been known as the
major copper containing protein in erythrocytes ((11), and refs therein), but its enzymatic
function was not discovered until 1969, when McCord and Fridovich found that
‘erythrocuprein’ dismutes superoxide. That is, it catalyzes the disproportionation reaction
where the first superoxide molecule is oxidized and the second molecule is reduced,
turning two molecules of superoxide into O2 and H2O2. The enzyme catalytic cycle can
be described by the ‘ping-pong’ mechanism(12):
2+ -
Where the co
erythrocuprei
the reduction
in situ(8). Thi
In a related as
system of ribo
colored produ
NBT for the s
activity in a n
competitive in
A microplate
used in attem
Cu
Cu+
O2
O22H+ + O2-
H2O2
pper is at the SOD1 active site. The superoxide dismutase activity of
n (and the major copper containing protein in virtually all tissues) inhibits
of cytochrome c by extrinsically added superoxide or superoxide produced
s reaction forms the basis for the enzyme assay still commonly used today.
say, superoxide is generated by dissolved oxygen and the radical generating
flavin, UV light, and tetramethylene diamine (TEMED), which produces a
ct upon reduction of nitroblue tetrazolium (NBT); SOD1 competes with
uperoxide radical(10). This assay has been adapted to measure SOD1
ative eletrophoretic gel and, by its very nature, is free of interference from
hibition by cytochrome c oxidase present in cell and tissue homogenates.
assay for SOD1 activity is also available(13). SOD1 mimetics have been
pts to protect against the superoxide burst produced upon reperfusion of an
5
ischemic event(14), though these have catalytic rate constants 2-4 orders of magnitude
slower than the authentic Cu, Zn superoxide dismutase(15).
SOD1 structure
Eukaryotic SOD1 is a stable homodimer where each subunit is related to the other by an
approximate C2 axis of rotation in mammals and, strikingly, an exact (non-
crystallographic) C2 axis of rotation in yeast(16). This orients the active site of one
subunit on the opposite side of the molecule relative to the other subunit. The dimer is
held together primarily with hydrophobic contacts, burying approximately 550Å of
hydrophobic surface area in the interface of the two subunits(17). Dimerization of SOD1
reduces the solvent accessible surface area, greatly increasing its stability(18). SOD1 is
also one of the most stable proteins known. The fully metallated protein melts at 85-95°C
(depending on buffer, e.g. (19)) and is enzymatically active in 8M urea or 4M guanidine-
HCl(20). E. coli SOD is highly homologous to the mammalian enzyme, but is
monomeric(21). The active site of the enzyme was more easily denatured than the
mammalian, dimeric SOD1 and the E. coli protein as a whole has a melting temperature
17.7ºC lower than the human protein(21). Several engineered monomeric SOD1s have
been created by mutating residues within the hydrophobic dimer interface to charged
residues(22). These monomeric mutants show disorder in the metal binding loop and the
catalytically important Arg 143(23) is also highly disordered. This disorder results in a
10-fold loss in activity, but this can be abrogated somewhat by compensatory mutations
that stabilize the active site with hydrogen bonding(24). Dimerization may also be related
to cooperative function of the two subunits(25), but this has not been confirmed
experimentally.
6
SOD1 is part of the immunoglobulin-fold family(26), of which each monomer consists
primarily of an eight-stranded beta-barrel with two large loops, the so-called 'electrostatic
loop' and the 'metal binding' loop (Figure 1)(17). Mutations associated with SOD1-ALS
are shown in Figure 1b and comparison of the sequence with mouse and rat SOD1 is
shown in Figure 1c. The electrostatic loop (residues 122-143) has a number of charged
residues, but these are not likely catalytically important because mutating four residues to
uncharged residues actually increases the catalytic rate constant. The ‘metal binding’ loop
(residues 49-84) contains many of the residues necessary for binding of the metals(27).
The topological and hydrogen bonding connectivity of the strands within each SOD1
monomer is reminiscent of the connectivity of lines found in ancient Greek pottery, and
is accordingly called the ‘Greek-key’ fold(28). Because the N-terminus and C-terminus
of SOD1 are adjacent, it was possible to produce circular permutants of SOD1 by
connecting these and ‘cutting’ loops connecting other strands, which creates a new N-
and C- terminus. These permutants have similar expression, stability and rates of
catalysis of wild-type SOD1(28). Strands 1, 2, 3 and 6 are regular, with little twisting,
and are on the opposite face of the barrel from the active site. Strands 4, 5, 7 and 8 are
shorter and more twisted than the other half of the barrel. Several ‘beta bulges’ in these
strands accommodate metal binding(17). It has been hypothesized that the two halves of
the barrel are the product of primordial gene duplication(29). A conserved disulfide bond
between residues 57 and 146 [amino acid numbering is for the human enzyme
throughout] greatly increases SOD1 stability(30). Similar disulfide bonds are present in
other immunoglobulin fold proteins(26).
7
Figure 1. Structure of a SOD1 dimer (pdb code: 1SPD). The left subunit is shown in stick representation for detail. Copper is coloured blue, zinc is coloured lavender. Bridging His 63 is shown in red; the secondary bridge is shown His 46 (yellow)-Asp 124 (magenta)-His 71(yellow). The remaining metal binding histidines are shown in cyan. The right subunit is shown as a cartoon ribbon to illustrate the overall architecture of the SOD1 subunit. The beta barrel is shown in gray, the metal binding loop in green and the electrostatic loop in blue.
8
mouse MAMKAVCVLKGDGPVQGTIHFEQKASGEPVVLSGQITGLTEGQHGFHVHQYGDNTQGCTS rat MAMKAVCVLKGDGPVQGVIHFEQKASGEPVVVSGQITGLTEGEHGFHVHQYGDNTQGCTT human MATKAVCVLKGDGPVQGIINFEQKESNGPVKVWGSIKGLTEGLHGFHVHEFGDNTAGCTS
mouse AGPHFNPHSKKHGGPADEERHVGDLGNVTAGKDGVANVSIEDRVISLSGEHSIIGRTMVV rat AGPHFNPHSKKHGGPADEERHVGDLGNVAAGKDGVANVSIEDRVISLSGEHSIIGRTMVV human AGPHFNPLSRKHGGPKDEERHVGDLGNVTADKDGVADVSIEDSVISLSGDHCIIGRTLVV
Multiple sequence alignment of SOD1 proteins from Mus musculus (mouse, NP_035564), Rattus norvegicus (rat, NP_058746), and Homo sapiens (human, CAG46542) using ClustalW (http://www.ebi.ac.uk/Tools/clustalw2/index.html). Identical sequences are highlighted with an asterisk (*), conserved with a colon (:) and similar with a period (.). Overall, human SOD1 is 83% identical to mouse or rat SOD1, where the rodent SOD1s are 94% identical. SEDI binding sequence is 100% conserved in all species (see Chapter 4). Beta sheet regions (from analysis of structure 1spd in the Protein Database) are marked with lines.
9
0
1
2
3
4
5
6
7
4 14 37 46 59 84 90 100 111 117 132 144 151
Num
ber o
f mut
atio
ns a
t site
Residue Number
Figure 1c. Several SOD1 mutations associated with ALS are shown in relation to the primary sequence (upper). Mutations are spread through all five SOD1 exons (lower).
10
Each SOD1 monomer also contains two metal ions, one copper and one zinc, which play
structural and catalytic roles in the enzyme. The catalytic copper is bound by four
histidines, His46, 48, 63 and 120, in a distorted tetrahedral binding geometry in the
oxidized (Cu(II) ) form and in a distorted trigonal planar geometry, bound by His 46, 48
and 120 in the reduced (Cu(I)) form(17). There is also evidence for a fifth water ligand
for copper in the oxidized state(16). A network of hydrogen bonds may reduce negative
effects of improper binding geometries of the two metals(17). The zinc ion, which is
bound by His 63, 71, 80 and Asp 83 acting as a monodentate ligand, is thought to play a
structural role and act as a positive charge sink. His 63 bridges the copper and zinc ions
where its delta-nitrogen is a zinc ligand and its epsilon-nitrogen is a copper ligand. A
secondary bridge exists where Asp 124 hydrogen bonds to both His 46 (a copper ligand)
and His 71 (a zinc ligand). Mutation of Asp124 to Asn (analagous to the ALS-associated
mutant D124V) precludes formation of this hydrogen bond, leading to a dramatic
decrease in zinc binding affinity(31). This metal binding architecture, where the two
metals are bridged by a histidine side chain as well as by a secondary linkage of hydrogen
bonds, appears to be unique among metallo-enzymes(17).
Mechanism of SOD1 activity
The standard reduction potential for the O2 /O2- couple is –0.33V and for the O2-/ H2O2
couple is +0.89V(12). Thus, any redox-active metal with a reduction potential between
these two values can oxidize and reduce superoxide by catalyzing the spontaneous
transfer of an electron to the metal and to superoxide in a subsequent step. The nominal
reduction potential of the Cu2+/Cu+ couple is +0.16V(32). Verhagen et al measured the
reduction potential of Cu2+-SOD1/Cu+-SOD1 as +0.12V at pH 7.5 in 0.2M salt at 22
11
°C(33). Whether SOD1 actually catalyzes both reactions, or instead catalyzes only the
oxidation or reduction of superoxide and then is restored to the active state with another
redox agent is an issue that was resolved only recently. In fact, SOD1 mimetics often
have only one of these activities. Liochev and Fridovich showed that SOD1 can
participate in a couple with ferrocyanide or ferriccyanide, illustrating that the enzyme
acts as both a superoxide oxidase and a superoxide reductase(34).
The details of the enzymatic mechanism come from a number of high-resolution
X-ray crystal structures and NMR data and lower resolution, but still important EXAFS
and ESR data. The active site is positively charged and makes up approximately 11% of
the total exposed surface(35), but the rest of the surface is negatively charged (Figure 2).
This charge gradient increases the equilibrium concentration of superoxide near the
active site channel, but the electric field gradients show that superoxide would be repelled
from the sides of the channel itself. Based on the rate of diffusion of superoxide and
SOD1, the SOD1 catalytic rate constant of 2 x 109 M-1s-1 is approximately diffusion
controlled once superoxide approaches the SOD1 active site(36). Increasing the positive
charge at the active site while preserving the hydrogen bonding network can increase the
rate of SOD1 catalytic activity by enhancing electrostatic guidance(37). The active site is
also much more evolutionarily conserved (86%) than the rest of the SOD1 sequence
(41%). The catalytic importance of Arg 143 is thought to stem from hydrogen bonding
with the incoming superoxide(35). Arg 143 and Thr 137 also limit the size of anions
incoming to the copper-active site(16). Based on EXAFS and ESR data(38, 39), the
copper becomes tri-coordinated upon reduction to Cu(I) and the bridging histidine no
longer binds to copper. This allows for relaxation of the otherwise unfavourable
12
distortions in the binding geometry of the zinc binding ligands(35) and allows the
histidine to become protonated, which is facilitated by the increased acidity of this ligand
by the zinc ion. This proton is then abstracted from His 63 to the outgoing O22-.
13
Figure 2. Surface charges on SOD1 dimer (pdb code: 1SPD). The dimer is shown with the same orientation as in Figure 1. Positively charged areas are coloured blue, negatively charged areas are coloured red. The active site positive charge enhances electrostatic guidance of superoxide to the copper centre. Negative charges dominate the remainder of the SOD1 dimer surface. Surface and electrostatic calculations were carried out with Swiss PDB viewer (freely available at http://www.expasy.org/spdbv/)(141) using the following parameters: dielectric constant (solvent) =80; use atomic partial charges; map potential to surface; computation method: Coulomb; dielectric constant (protein) =4; colour scale: red = -1.800, white =0.000, blue =1.800.
Crystallization of SOD1 in the Cu(I) state also shows that the histidine bridge is broken
by a rotation of the imidazolate side chain which moves the liganding nitrogen 0.66Å
away from the copper ion(16). Based on a number of crystal structures with copper
oxidized or reduced and with competitive inhibitors bound, Hart et al (16) proposed a
mechanism for SOD1 activity. The first step of the reaction is binding of the superoxide
to the copper active site displacing the axial water ligand. There is an inner sphere
electron transfer (the electron is transferred through a bond) to the copper and the oxygen
diffuses out. This causes a rearrangment to the trigonal bound Cu(I) state, which is then
oxidized by a second, non-covalently bound, superoxide in an outer sphere (the electron
is transferred through space rather than through a bond) mechanism, regenerating the
Cu(II) state (Figure 3).
SOD1 folding and molecular dynamics
Understanding the dynamics of protein motion in addition to the structure of a protein are
important to understanding its function and potential malfunction (40). The dynamics can
be probed using time-resolved fluorescence, NMR or computer simulations. Early
molecular dynamics simulations of SOD1 dynamics (300ps) revealed that there are
instantaneous asymmetric motions of the two SOD1 subunits through essential dynamics
analysis(25). Motion in the beta-barrel of one subunit influences the motion of the other
subunit, including re-organization of the electrostatic channel to enhance the interaction
with the substrate. In an experimental and computational study on SOD1 from
Photobacterium leiognathi, Falconi et al show that enhancement of the intersubunit
communication can increase the cooperativity of the enzyme and increase the catalytic
rate by approximately two fold when introducing a V29G mutation(41). The P. leiognathi
15
Figure 3. Comparison of the SOD1 active site in the Cu(I) and Cu(II) states. Copper is blue, zinc is gray. A) The active site in oxidized SOD1 (Cu(II)) shows copper ligandedwith four histidines, including the bridging histidine. (pdb code: 1SPD) B) SOD1 active site in the reduced state (Cu(I)), where the active site copper is no longer bound by the bridging histidine (pdb code: 1Q0E; note: reduced bovine SOD1 is shown for clarity- the structure of reduced human SOD1 shows double occupancy of the copper in the reduced site and the oxidized site).
16
SOD1 dimer interface, however, is oriented on a different face of the monomer than is
human SOD1, so the role of intersubunit communication in the human protein has not yet
been substantiated experimentally. In a recent five nanosecond molecular dynamics
study, Khare et al found that the intersubunit communication is decreased in mutant
SOD1 when the connectivity of amino acid residues was analyzed using graph theoretic
methods(42). The groups of Banci and Bertini have also studied the dynamics of SOD1
and mutant derivatives with NMR(43-47). Using an engineered monomeric SOD1
derived from the human sequence, they observed NMR line width broadening, indicative
or greater mobility, in loops 4 and 6, which are close to the dimer interface. Perhaps not
surprising based on computational data, they also observe greater mobility in the
electrostatic loop, which is opposite the dimer interface(47). They also observe a decrease
in enzyme activity that is proportional to the width of the active site channel formed from
reorganization of this electrostatic loop; sterically restricting access to the active site
reduced the enzyme activity. There are also large differences in the backbone dynamics
of the monomeric mutant SOD1 compared to the wild-type dimer, which is more
rigid(44). Engineered monomeric SOD1 also shows a loss of hydrogen bonds at the
active site, especially with Asp 124, which serves as a secondary bridge between the
copper and zinc ions and is necessary for reorganization of the zinc binding site(45).
When the dynamic properties of G93A mutant SOD1 were probed with NMR,
similarities to the engineered monomer were found. There was increased disorder in the
loops and an increase in the destabilization of residues necessary for blocking edge-strand
aggregation(48). SOD1 mutations also destabilized the amino acids at the ends of the
beta-barrel in a computer simulation of other fALS associated mutant SOD1s(42).
17
Studies into SOD1 folding have been limited because of the intrinsic difficulties in
studying the folding behaviour of a protein that undergoes extensive post-translational
maturation steps and by the resulting complexity of the equilibrium of various states of
SOD1. The variables in these states include: differing metallation status, whether the
copper is oxidized or reduced, the redox state of the disulfide bond, dimerization, subunit
communication and subunit asymmetry; each of these need to be evaluated properly to
understand the folding behaviour of SOD1. Several groups have, however, attempted to
study the unfolding and refolding of SOD1 under fully metallated or apo-conditions. The
nature of the tryptophan environment has been probed with fluorescence lifetime
measurements, where the lifetime is dependent on both temperature and
unfolding curves where the midpoint is dependent upon proper metallation. However, the
width of the fluorescence lifetime distribution at the transition midpoint, circular
dichroism, analytical ultracentrifugation and fluorescence anisotropy measurements
revealed a partially folded monomeric intermediate in the SOD1 unfolding pathway(50,
51). There is also strong kinetic hysteresis in the guanidine induced denaturation of
SOD1 which reveals a monomeric kinetic intermediate in addition to thermodynamic
intermediates observed earlier(52). Thermal denaturation of holo- and apo-wild-type
SOD1 can be approximated as a two-state unfolding process(53), suggesting that the
folding intermediates are unstable at higher temperatures. However, fitting thermal
denaturation of mutant apo-SOD1 is complicated by competing aggregation
reactions(53). Thus, SOD1 folding is either three-state (native/folded, intermediate, and
18
unfolded) or two-state (native/folded and unfolded), depending on the protein
concentration, temperature and the presence of mutations.
Post-translational processing of SOD1
SOD1 is subject to at least four post-translational maturation steps in addition to N-
terminal acetylation: copper insertion, zinc insertion, dimerization and disulfide bond
formation. Failure or alteration of one or more of these processes could result in a build
up of immature SOD1 which will not be properly folded since the native protein requires
each of these modifications to maintain its structure and stability. Thus, a mutation in a
gene that controls SOD1 post-translational modification may mimic a SOD1 mutation in
fALS by causing improper folding of SOD1 (Figure 4). In familial Alzheimer’s disease,
mutations in APP itself or one of the presenilin genes responsible for post-translational
cleavage of the A-beta peptide from APP can cause disease(54). Since SOD1 mutations
account for only 20-25% of fALS(2), perhaps a gene involved in the post-translational
maturation of SOD1 accounts for some fraction of the remainder.
O’Halloran and co-workers succinctly demonstrated that intracellular copper and zinc
levels are less than one free metal ion per cell(55, 56). Copper is inserted into SOD1 in
yeast cells through the copper chaperone for SOD1 (CCS)(57). CCS consists of three
domains: domains I and III are thought to be necessary for copper insertion into SOD1,
while domain II is homologous to SOD1. SOD1 and CCS may form a heterodimer during
copper insertion using the same interface as that of SOD1 homodimerization(58);
however, another model, where a CCS homodimer and SOD1 homodimer dock in a
heterotetramer avoids the energetically unfavourable dissociation of the SOD1 dimer(59).
It is unknown how CCS obtains copper from outside the cell, but perhaps it acquires it
19
Figure 4. SOD1 undergoes extensive post-translational maturation. Mutations in genes that alter post-translational processing of SOD1 may also affect SOD1 stability and function. The relevant genes are shown under each step.
20
from the copper transporter Ctr(58). Mammalian SOD1 can be activated in a CCS-
dependent or CCS-independent manner(60). CCS-independent activation relies on
reduced glutathione. CCS-independent activation of SOD1 can only take place in the
absence of two proline residues that are present in yeast SOD1. Introduction of these
prolines into human SOD1 renders it completely dependent upon CCS in vivo for proper
copper insertion(61). Insertion of copper into SOD1 from Caenorhabditis elegans occurs
exclusively through a CCS-independent pathway(62).
The presence of an intramolecular (within each subunit) disulfide bond in SOD1 is
somewhat unexpected since it is present at high concentration within the reducing
environment of the cytoplasm. The stability of SOD1 is dependent upon proper disulfide
bond formation(63). The apo-disulfide-reduced form shows a similar circular dichroism
spectrum to the oxidized form, but has an NMR HSQC spectrum similar to the
engineered SOD1 monomer and is also monomeric in size-exclusion chromatography
(63). The monomeric, disulfide reduced, apo-SOD1 also displays two-state folding
behaviour(64). FALS associated mutant SOD1s are also more susceptible to disulfide
reduction than the wild-type protein(65). CCS appears to play an important role in the
formation of the SOD1 disulfide bond. SOD1 activity is disulfide bond dependent and
CCS can activate the enzyme even in the presence of EDTA (a Cu (II) chelator) and
bathocuprein sulfate (a Cu (I) chelator)(66). Moreover, apo-CCS can also catalyze the
formation of the SOD1 disulfide bond, albeit slowly. Whether disulfide bond formation
precedes, is concomitant with copper insertion, or is a product of fast auto-oxidation from
SOD1 active site copper is still unknown to date.
21
The mechanism of zinc insertion is unknown, but because zinc concentrations are
regulated with femtomolar sensitivity(55)(67), we speculate that there is a factor that
inserts zinc specifically into SOD1 or non-specifically into a number of proteins, as
opposed to acquiring it from a freely diffusible pool or only loosely bound zinc. This may
be one of the metallotheineins, or another unknown protein or small molecule. Whether
dimerization occurs spontaneously, or requires a chaperone or some other factor is also
unknown.
SOD1 is ubiquitously expressed, but is present in some tissues at higher
concentration than others(68). Metabolic labeling with 64Cu in mouse fibroblasts shows
that the concentration of apo-SOD1 is inversely proportional to copper levels and that
copper is rapidly incorporated into both pre-formed apo-SOD1 and newly translated
SOD1(69). In addition, SOD1 may be inducible under conditions of oxidative stress. The
CCS-dependent activity of SOD1 is also inducible upon oxidative stress where protein
synthesis is inhibited(70). The mechanism of oxygen induced transcriptional activation of
SOD1 and the control over differential tissue expression of SOD1 are active areas of
research.
Aberrant enzymatic activity of mutant SOD1
Superoxide associated toxicity has been associated with longevity and disease(71). Based
on the tight genetic linkage between some patients with fALS to a locus on chromosome
21q which contains the SOD1 gene and the putative role of oxygen radicals in ageing and
neurological diseases, Rosen and coworkers investigated SOD1 as the causative gene in
fALS(1). Upon conducting PCR on 2 of the 5 SOD1 exons, they found that mutations in
SOD1 were correlated with fALS. Familial ALS is dominantly inherited; how a mutation
22
in one of the copies of a ubiquitously expressed gene causes a dominantly inherited
phenotype was, and remains, the major unanswered question. Rosen suggested that
mutations might actually increase SOD1 activity or have a dominant negative effect
whereby the mutant inactivates the wild-type protein in a heterodimer. However, it was
later shown that several (A4V, G37R, G41D, G93C, I113T) mutant SOD1s have similar
in vitro activity to that of wild-type SOD1. Only G85R, a mutation close to the metal
binding site (see below) lacked activity(72). This and other metal binding region mutant
SOD1s do not rescue the sod1∆ lysine auxotrophy phenotype in yeast(73). Furthermore,
mutant SOD1s do not have a dominant negative effect on wild-type subunit function(74).
In addition to these in vitro experiments, several lines of evidence from mouse genetic
studies point to a SOD1 gain-of-function mutation rather than a loss of function. Mice
expressing greater than normal levels of mutant human SOD1 produced ALS-like motor
neuron pathology, despite the presence of normal endogenous mouse SOD1(75, 76). This
is also true of the G85R SOD1 mutant, even when expressed to lower levels than that
required to produce a phenotype with the G93A and G37R SOD1 mutants(77). Removing
the mouse SOD1 gene did not produce a motor neuron phenotype(78).
In addition to its superoxide dismutase activity, SOD1 also has peroxidase activity,
whereby it generates radicals from hydrogen peroxide(79). A4V and G93A mutant SOD1
increased the rate of hydroxyl radical formation as monitored by the electron spin
resonance change of the spin-trap 5,5'-dimethyl-1-pyrroline N-oxide (DMPO)(80). This
change in reactivity is attributed to a change in the mutant enzyme’s Km for H2O2 (81,
82). Treatment of SOD1 with H2O2 also damages the enzyme by oxidative modification
of histidine to 2-oxo-histidine(83). This modification leads to copper release in the Cu (I)
23
form and subsequent inactivation of SOD1(84). The oxidative modifications of histidine
also leads to damage of the zinc site, which occurs much more rapidly than enzyme
inactivation through copper loss(85). Bicarbonate protects SOD1 from H2O2 induced
inactivation; however, it also catalyzes the formation of a carbonate anion radical,
increasing the overall peroxidase activity of SOD1(86). Hydrogen peroxide first reduces
the Cu (II) to Cu (I), as seen by the loss of the absorbance at 650nm from the Cu d-d*
excitation, and then catalyzes the formation of the carbonate anion radical, which may
then diffuse and oxidize various exogenous substrates(87). Dissolved CO2 may also react
with H2O2 + SOD1, also increasing its peroxidase activity(88). The exact radical species
being generated is highly dependent upon solution conditions and whether the peroxidase
activity also occurs in vivo is unknown.
Superoxide also reacts in a near diffusion-limited reaction with nitric oxide which
produces peroxynitrite(89). Peroxynitrite rapidly degrades under physiological conditions
to produce hydroxyl radical and nitrogen dioxide radical(90). Wild-type SOD1 can also
react with peroxynitrite to produce a highly reactive nitronium species, which can then
nitrate tyrosine residues. Mutant SOD1 could have greater nitrating activity than the
wild-type enzyme because of increased disorder in the residues that mediate anion
selectivity in the active site(91). Mutations produce a partially metallated zinc-deficient
form of SOD1(92) that promotes apoptosis in a nitric oxide dependent manner in cultured
motor neurons replete with growth factors, whereas the holo-protein rescues cultures
deprived of trophic support(93). That is, the holo-protein protects against oxidative
damage caused by trophic factor withdrawal, but the zinc-deficient form produces enough
oxidative stress that it causes cell death in the absence of exogenous oxidative stress
24
(from trophic factor withdrawal). The complex interaction of nitric oxide, superoxide,
peroxynitrite, SOD1 and their targets, however, makes it difficult to predict which
species is the most culpable and what the intracellular targets of modification might
be(94). The ‘SOD1 oxidation/nitration’ hypothesis has been tested in vivo. Cleveland and
co-workers showed that there is an increase in free, but not protein bound nitro-tyrosine
in ALS(95). Free nitrotyrosine is also sufficient to cause motor neuron, but not astrocyte,
apoptosis(96). Aberrant reactivity of mutant SOD1 with nitric oxide has also been linked
to a disruption of nitric oxide signaling. SOD1 catabolizes S-nitrosothiol containing
peptides(97), which depletes the overall S-nitrosothiol content of various cellular
components and interferes with nitric oxide signaling(98).
Despite in vitro and in vivo evidence of aberrant copper chemistry in fALS mutant SOD1
pathogenesis, the hypotheses linking mutant SOD1 toxicity to copper mediated chemistry
has recently come into disfavour because of two key experiments. First, copper is
predominantly inserted into SOD1 by the copper chaperone for SOD1 (CCS) in
mammals(60) and exclusively in yeast(56). Crossing a mouse lacking the endogenous
mouse copper chaperone with a mouse harbouring mutant SOD1 transgenes did not alter
the onset or survival of the double-mutant mouse(60). The decrease in copper loading did
correspond to a decrease in SOD1 dismutase activity, but did not alter the disease course;
from this, the authors concluded that copper chemistry is not critical to mutant SOD1
toxicity. It has been argued that the low levels of residual copper and copper at alternative
binding sites may still contribute to copper mediated mutant SOD1 toxicity(99).
However, an ALS-like phenotype is also observed if a quadruple mutant SOD1, where
each of the copper binding histidines are mutated to alanine, is transgenically expressed
25
in mice(100). Thus, mutant SOD1 maintains its toxicity despite the absence of copper at
the active site in mouse models of SOD1-ALS. This does not mean that oxidative
mechanisms involving copper chemistry are not involved in ALS pathogenesis; rather,
these models may by-pass the necessity for oxidative damage where these mutations
mimic oxidatively modified histidines in their inability to bind metals (see section on
SOD1 misfolding, below).
Structural changes in mutant SOD1
Shortly after Rosen and co-workers’ discovery of mutations in SOD1 exon 2 or 4 can
cause fALS(1), Deng and co-workers found additional mutations in SOD1 exons 1, 2, 4
and 5 and carried out analysis of what effect these mutations might have on SOD1
structure(101)(102)(103). The mutations they found clustered at the dimer interface;
these mutations were thought to disrupt the dimer, disrupt the folding of the beta-barrel,
or both. Mutant SOD1s isolated from these patients showed reduced SOD1 activity in red
blood cells. Tainer (17) point out that the only 5 non-glycine residues in the wild-type
SOD1 structure fall outside the allowed areas in the Ramachandran plot. The three most
common mutant SOD1 expressing mice are G93A, G37R and G85R. Each of these
residues falls outside the allowed area, forcing conformational adjustments in the vicinity
of the changed residues (Figure 5). Destabilization caused by altering these glycine
residues may be a common feature of these mutant SOD1s. Mutation of Gly 93 to various
amino acids destabilized SOD1 by levels corresponding to their preference for this
conformation(53). The A4V SOD1 mutation is the most common found in fALS patients.
Comparison of the A4V and wild-type SOD1 structures shows that there is only a 0.5Å
rmsd when considering both the backbone and side chains(104). This is quite small when
26
Figure 5. SOD1 glycine mutations expressed in commonly used mouse models of ALS fall outside the low energy areas. A) Ramachandran plot showing the backbone torsional angles of SOD1 residues. The commonly used mouse models are circled – G93 (green), G37 (pink), G85 (red). B) Structure of a SOD1 subunit showing the location of the common glycine mutants. G93 (green), G37 (pink), G85 (red).
27
compared to a 1.71Å rmsd between the structures of oxidized (pdb code: 1SPD) and
reduced (pdb code: 1C9V(105)) human SOD1. Local packing disruptions are observed in
the region of the A4V mutation because of the inclusion of two extra methyl groups;
however, it is unclear whether this is sufficient to cause the large difference in stabilities
towards guanidine-HCl denaturation(104). The structure of the H46R SOD1 mutant,
which is a zinc binding ligand, shows that the histidine is replaced by a water
ligand(106). The Asp124 ‘secondary bridge’ which is important for stabilization of the
active site with hydrogen bonds is also disrupted in this mutant. The zinc-deficient
mutant SOD1 shows loss of structure in both the metal binding loop and the electrostatic
loop, but 20% zinc occupancy leads to complete restoration of loop integrity(107). This
loss of structure also allows for edge-strand interactions and leads to crystallization of
H46R, S134N and H43R mutants in arrays resembling fibrous aggregates(106, 108, 109).
The new crystallographic interface between SOD1 subunits observed in these crystal
structures is comparable to that of the native dimer interface(108).
Alterations in biochemical properties of mutant SOD1
Since there are minimal changes in the crystal structure of fully metallated mutant
SOD1 and the role of aberrant copper redox chemistry is unclear, other biochemical
properties of mutant SOD1 must be involved in the pathogenesis of SOD1-ALS. Mutant
SOD1s are less thermostable than the wild-type protein by 1-6°C; interestingly, this was
independent of the metallation of the protein as the metallation seemed to confer equal
stability to wild-type and mutant SOD1s (approximately 14°C for the first metallation
and 22°C relative to the apo- for the second)(110). Mutant SOD1s that cannot bind
metals melt at temperatures 4-12 °C lower than wild-type apo-SOD1 (c.f. 1-6 °C for
28
mutants that can bind metals)(110). When unfolding SOD1 with chaotrope (urea or
guanidine-HCl), mutant SOD1 unfolds more easily than wild-type SOD1, but this effect
is more pronounced in apo-SOD1 than in the holo-protein. This prompted Lindberg et al
to conclude that destabilization of apo-SOD1 is common to fALS associated
mutants(111). However, the apo-forms of some metal binding region mutant SOD1s have
higher melting temperatures than even wild-type SOD1. These mutant apo-SOD1s also
show similar H/D exchange characteristics to that of wild-type apo-SOD1(112).
FALS-associated mutations are also thought to alter the metal binding affinities of
mutant SOD1. Mutations scattered across SOD1 can alter the zinc binding geometry.
This change allows for greater flexibility in the active site, allowing faster reduction of
Cu (II) to Cu (I) by ascorbate. Mutant SOD1s were reduced at rates comparable to zinc-
deficient SOD1(113). The metal binding affinities of SOD1 are so high that they are
difficult to measure directly. Instead, they are calculated by comparing the metal release
rates under mildly denaturing conditions and competition for these metals with chelators
of known binding affinity(92). Using this methodology, Crow et al calculated the
dissociation constants of zinc and copper in wild-type SOD1 to be 4.2 x 10-14 M and 6.0 x
10-18 M, respectively. Mutant SOD1s had weaker zinc and copper binding affinities;
however, the difference in zinc binding was approximately 20- to 30- fold, whereas the
difference in copper binding was only 1-2 fold. Zinc deficient SOD1 also catalyzed the
formation of nitrotyrosine faster than holo-SOD1 (mutant or wild-type)(92).
Recombinant mutant SOD1s isolated from insect cells also showed decreased metallation
compared to wild-type SOD1. This corresponded with a decreased catalytic rate of some
of these mutant SOD1s in pulse radiolysis experiments(114). Zinc-deficient SOD1
29
activity is 10x slower than that of holo-SOD1 at pH 8.0(115). Alterations in the metal
binding geometry also causes a loss of metal binding specificity in fALS-mutant
SOD1s(115).
SOD1 misfolding
Before examining the case for SOD1 misfolding in ALS, it is useful to begin with our
definition of ‘misfolding’. We call a protein ‘folded’ if it has regular structure, usually
including elements of secondary structure. Proteins in their native state are often folded,
but not always so(116). A protein is ‘unfolded’ if it is soluble and does not have any
regular structure. We define ‘misfolding’ as a substantial alteration or re-organization of
the native protein structure.
The neuropathological evidence for SOD1 misfolding in ALS is reviewed elsewhere
(e.g.(117)). Briefly, inclusion bodies are observed in almost all forms of ALS, both
sporadic and familial. These fall into three main categories: skein-like, Bunina bodies,
and hyaline. Skein-like inclusions are found in nearly all ALS cases, and stain heavily
with ubiquitin antibodies and are occasionally positive for Dorfin, a ubiquitin E3
ligase(117). Bunina bodies are small eosinophilic inclusions (they are positively charged)
in motor and occasionally other neurons. Hyaline inclusions are ‘glassy’ when stained
and are large, containing neurofilaments. Strong SOD1 staining, suggestive of SOD1
aggregation, of hyaline inclusions is seen in most cases of human ALS carrying the
SOD1 mutation and in mouse models of SOD1-ALS(117-120). In conjunction with
evidence from other neurological disorders where protein aggregation and misfolding is
thought to play a role, these hyaline inclusion/aggregates are taken as the primary
neuropathological evidence for SOD1 misfolding in ALS. With the finding that mutant
30
SOD1 can cause an ALS-phenotype independent of copper loading into the active site
reducing the strength of the pro-oxidation hypothesis of SOD1-ALS(60, 100), a
mechanism involving SOD1 misfolding is coming to the forefront.
We aimed to reconcile these findings by suggesting that the two leading hypotheses
regarding mutant SOD1 toxicity, increased oxidative stress and SOD1 misfolding, are
causally linked(121). We hypothesize that SOD1 misfolding requires and is a result of
oxidation of SOD1 itself. SOD1 is an exceptionally stable protein, the holo-protein
having a melting temperature of ~95ºC(122). Oxidation with hydrogen peroxide causes
metal release and inactivation(84). We have also shown that physiological metal
catalyzed oxidation causes destabilization and aggregation of SOD1(122). Oxidized
SOD1 is also found in a mouse model of ALS(123). A number of factors contribute to the
selective vulnerability of SOD1 within motor neurons to oxidative modification. SOD1 is
present at high concentration in motor neurons(68), and its normal enzymatic function
exposes it to high levels of oxidative stress, predisposing it compared to other proteins to
oxidative damage. It also has a long life-time in motor neurons because of slow transport
in motor axons that potentiates oxidative modification(121). Decreasing SOD1 activity
by limiting or eliminating copper insertion may allow for greater damage to SOD1 itself,
thereby causing irreversible modification to histidine residues that prevent it from
binding to metals and properly folding. If histidines are replaced with other residues, this
mimics oxidation of histidine by also impeding proper metallation.
Mutant SOD1s displayed greater aggregation propensity than wild-type holo-SOD1 when
treated with a relatively mild oxidation system of copper and ascorbic acid. The most
aggregation prone species is zinc-deficient SOD1. Moreover, aggregates produced from
31
oxidation of holo-SOD1 were zinc-deficient (Figure 6). The aggregation propensity was
proportional to the extent of partially unfolding (121). We refer to this species as
‘misfolded’ because SOD1 is normally soluble and because the circular dichroism
spectrum shows a large change upon aggregation. Other, non-physiological
denaturational stresses, such as heating in the presence of 30% TFE(124, 125), also
showed that mutant SOD1s have an increased aggregation propensity over the wild-type.
Aggregation of SOD1 can also be induced to misfold in vitro by heating at denaturing pH
(pH 3.5)(109, 125) or by treating with the relatively strong oxidizing combination of
copper and hydrogen peroxide(126). Molecular dynamics simulations of SOD1 peptides
derived from SOD1 find that N- and C-termini, as well as two beta-strands and two loops
are especially aggregation prone(127).
Whether the final protein aggregates are themselves toxic, or whether some soluble
precursor is the major toxic species is a hotly debated subject. There have been
conflicting reports regarding the toxicity of SOD1 aggregates(128, 129). In order to
investigate whether protein-misfolding intermediates may play a role in ALS
pathogenesis, we looked for intermediates in our in vitro model system(121, 122). Using
dynamic light scattering and analytical ultracentrifugation, we found that monomeric
SOD1 is a misfolding intermediate(122) (Figure 6). The suggestion of monomeric SOD1
formation in mutant SOD1 preparations at 70-80°C was then observed using x-ray
scattering data(130). Monomeric SOD1 has since been identified as the aggregation
prone species in protein folding studies(125, 131) and in a detailed analysis of the SOD1
misfolding pathway at pH 3.5 (125). We are currently examining whether the monomeric
SOD1 intermediate observed in vitro exists in vivo(132).
32
Figure 6. Plausible free energy profile of oxidation induced SOD1 aggregation (adapted from (122)). Mutant SOD1 is less stable than wild-type SOD1(53); upon oxidation this populates a zinc-deficient state and a monomeric intermediate prior to aggregation. Since oxidation causes conversion of histidine to oxo-histidine, this disrupts metal binding capacity and the zinc-deficient intermediate precedes the monomeric intermediate. The activation barrier for conversion from SOD1 monomers to zinc-deficient aggregates is shown as being small because zinc-deficient SOD1 aggregates spontaneously at 37ºC(121) and oxidized SOD1 is more unstable to thermal denaturation than the native form(122). Aggregates produced from in vitro oxidation of holo-SOD1 are also zinc-deficient (unpublished data). Briefly, holo SOD1 is oxidized as in Rakhit et al(122), and the resultant aggregates are pelleted, washed three times with ddH2O repeatedly before the final pellet is dissolved in 6M HCl. The last wash contained very low concentrations of copper and zinc. The protein and metal contents of the pellet were determined by amino acid analysis and graphite-furnace atomic absorbance spectroscopy, respectively.
33
Therapies targeting SOD1 misfolding
If misfolding is the cause of SOD1-ALS pathogenesis, a hypothesis driven approach to
drug design might be to find drugs that stabilize the SOD1 dimer and prevent it from
misfolding. A similar approach has been applied in the case of TTR amyloidoses(133).
Transthyretin (TTR, also called prealbumin) is a tetrameric protein that is implicated in
familial amyloid polyneuropathy (FAP) (134). It also dissociates to monomers prior to
aggregation(135). Treatment with dibenzofurans selectively stabilizes the TTR
tetramer(136). Introduction of a second cysteine residue at the dimer interface was
thought to stabilize the SOD1 dimer by introducing an intermolecular disulfide
bond(137), but structural evidence for this disulfide bond has not been reported. This
mutant appeared to prevent the loss due to aggregation of A4V SOD1 in a gel-filtration
experiment in vitro, but failed to protect against toxicity in a chick embryo(138).
Oxidation of SOD1 Cys 111 to cysteic acid also stabilizes SOD1 to unfolding and
aggregation(139). If a small molecule can mimic this effect, it may have therapeutic
benefit. In a novel approach, Ray et al used computational design to find small molecules
that bind to and stabilize dimeric SOD1, but whether they are effective in cell or animal
models, or ultimately the clinic, remains to be seen(140).
Acknowledgements:
This work was funded by the Neuromuscular Research Partnership, a program of the
Canadian Institutes of Health Research (CIHR), ALS Society (Canada) and Muscular
Dystrophy (Canada).
34
References:
(1) Rosen, D. R., Siddique, T., Patterson, D., Figlewicz, D. A., Sapp, P., Hentati, A.,
Donaldson, D., Goto, J., O'Regan, J. P., Deng, H. X., and et al. (1993) Mutations in
Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral
sclerosis. Nature 362, 59-62.
(2) Majoor-Krakauer, D., Willems, P. J., and Hofman, A. (2003) Genetic
epidemiology of amyotrophic lateral sclerosis. Clin Genet 63, 83-101.
(3) Hadano, S., Hand, C. K., Osuga, H., Yanagisawa, Y., Otomo, A., Devon, R. S.,
Miyamoto, N., Showguchi-Miyata, J., Okada, Y., Singaraja, R., Figlewicz, D. A.,
Kwiatkowski, T., Hosler, B. A., Sagie, T., Skaug, J., Nasir, J., Brown, R. H., Jr., Scherer,
S. W., Rouleau, G. A., Hayden, M. R., and Ikeda, J. E. (2001) A gene encoding a putative
GTPase regulator is mutated in familial amyotrophic lateral sclerosis 2. Nat Genet 29,
166-73.
(4) Nishimura, A. L., Mitne-Neto, M., Silva, H. C., Richieri-Costa, A., Middleton, S.,
Cascio, D., Kok, F., Oliveira, J. R., Gillingwater, T., Webb, J., Skehel, P., and Zatz, M.
(2004) A mutation in the vesicle-trafficking protein VAPB causes late-onset spinal
muscular atrophy and amyotrophic lateral sclerosis. Am J Hum Genet 75, 822-31.
(5) Greenway, M. J., Andersen, P. M., Russ, C., Ennis, S., Cashman, S., Donaghy, C.,
Patterson, V., Swingler, R., Kieran, D., Prehn, J., Morrison, K. E., Green, A., Acharya, K.
R., Brown, R. H., and Hardiman, O. (2006) ANG mutations segregate with familial and
microscope (AFM). Samples deposited and dried onto freshly cleaved mica and dried
under positive pressure. Contact mode images were obtrained using a Si3N4 tip (Digital
Instruments) with nominal spring constant of 0.12 N/m.
Electron Microscopy
94
EM grids (Canemco, Quebec) were floated on 10 µl drops of SOD1 samples, negative
stained with uranyl acetate (MecaLab Inc., Quebec), and examined in an FEI Tecnai 12
transmission electron microscope (80kV accelerating voltage).
Amino Acid Analysis
Amino acid compositions of oxidized and control SOD1 was
determined using the Waters Picotag Amino Acid Analysis system, which utilizes gas
phase acid hydrolysis (6N HCl, 120 ºC), and either precolumn derivitazation with
phenylisothiocyanate or postcolumn derivitazation with Ninhydrin.
Capillary LC-MS/MS
Peptides were analyzed using a Q-TOF Ultima mass spectrometer (Micromass,
Manchester, UK) coupled to a capillary HPLC. Peptides eluted by acetonitrile were
ionized by electrospray and peptide ions were automatically selected and fragmented in a
data dependent acquisition mode. Data base searching was done with Mascot (Matrix
Science).
ANS/Thioflavin T Binding
10 µM SOD1 in 10 mM Tris-acetate (pH 7.0) was incubated for 30 minutes with 20 µM
8-anilino-1-napthalene-sulfonic acid (ANS) or 20 µM thioflavin T before measuring
emission spectrum (excitation at 372 nm and 450 nm respectively).
Congo Red Spectral Shift Assay
SOD1 aggregates were diluted to a final concentration of 3�M (~100�g/ml) and
incubated with 6mM Congo Red for 30 minutes before measuring near UV and visible
absorbance.
Circular Dichroism
95
Zinc deficient SOD1 aggregates were centrifuged for 5 minutes at 13000 g and the
supernatant was removed and replaced with 20mM sodium phosphate buffer, pH 7.0.
Aggregates were then re-suspended by vortex and sonication before circular dichroism
(CD) spectra were recorded on an Aviv circular dichroism spectrometer model 62 DS at
25 °C.
Results and Discussion
Metal Catalyzed Oxidation of SOD1
We employed metal catalyzed oxidation, using CuCl2 and ascorbic acid, to
generate reactive oxygen species because of the physiological relevance of this system.
Metal catalyzed oxidation is the principle source of hydroxyl radicals under normal
physiological conditions(23), and is especially important under conditions of oxidative
stress(24). The concentrations of ascorbic acid used in this study (2 – 4 mM) are well
within the normal concentration range (0.5 – 10 mM) found in neurons and glial
cells(25). We examined the effect of oxidation on three different ALS-associated mutants
of SOD1—A4V, D90A, and G93A, as well as a site-directed mutant (D124N) that has
decreased zinc binding affinity(26) and serves as a model of zinc-deficient SOD1. A4V is
the most common mutation causing FALS, D90A causes a rare autosomal recessive form
of FALS(1), and G93A is the mutant most widely used for the transgenic mouse model of
ALS. We examined the effect of oxidation on the zinc-deficient form of wild-type SOD1
because this species has been implicated in neurotoxicity associated with ALS(11) and
because it can utilize ascorbate to produce superoxide and hydrogen peroxide directly.
We find that at neutral pH, oxidation of each of the three SOD1 mutants and zinc-
deficient wild-type SOD1 induces the formation of large aggregates that scatter light
96
(Fig. 1a). The zinc-deficient protein displayed the most robust aggregation reaction, and
interestingly D90A, the mutation that causes an autosomal recessive form of FALS,
displayed the least amount of aggregate formation. Oxidation of wild-type SOD1 under
identical conditions did not induce the formation of aggregates detectable by right angle
light scattering (i.e. visible aggregates > 350 nm in diameter). With the exception of zinc-
deficient SOD1, aggregates did not form in control samples that lacked oxidants. The
small amount of aggregate observed in the control sample of zinc-deficient protein
suggests that this form of the protein has an intrinsic aggregation tendency. The
aggregation reaction displays distinct pH dependence, with reduced aggregation at pH <
5.5 (Fig. 1b). Similar pH dependence has been observed in the oxidation-induced
aggregation of human relaxin, where oxidation of a single His residue apparently
accounts for the pH-dependence(27). Performing the oxidation reaction under anaerobic
conditions or in the presence of EDTA inhibited aggregation, revealing the absolute
requirement of copper and oxygen for oxidation-induced aggregation (Fig. 1c). On the
other hand, addition of free radical scavengers, mannitol and DMPO, did not inhibit
aggregation (Fig. 1c). Similar results have been obtained with copper-catalyzed
oxidation-induced aggregation of both human relaxin(28) and hamster prion protein(29).
The insensitivity to free radical scavengers and the pH dependence of the oxidation-
induced aggregation are consistent with the site-specific metal-catalyzed oxidation
mechanism, in which there is a requirement for a metal ion binding site that is in close
spatial proximity to the modification sites(23). In this type of oxidation reaction, very few
residues are modified.
97
0
0.5
1.0�
1.5
wild type Zn-Deficient A4V D90A G93A D124N
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5
1.5
1.0
0.5
0
1.0
0.8
0.6
0.4
0.2
0
cont
rol
anae
robi
c
ED
TA
man
nito
l
DM
PO
pH
Rig
ht A
ngle
Lig
ht S
catte
ring
Inte
nsity
(Arb
itrar
y U
nits
)
Figure 1. Zinc Deficient and mutant SODs form visible aggregates upon oxidation, whereas the wild-type protein does not. a) Comparison of right angle scattering signals from various SOD1 solutions upon oxidation with 4 mM ascorbate and 0.2 mM CuCl2 (black) vs. Control (gray) at 37°C, pH 7.0 for 48 hours. Dotted line indicates scattering produced by 10 mM Tris-acetate buffer with 4 mM ascorbate and 0.2 mM CuCl2. b) pH dependence of oxidation induced aggregation of SOD. Zinc-deficient SOD1 forms visible aggregates over a large pH range (5.0-7.5) upon oxidation (triangles). Wild-type SOD1 does not form visible aggregates under similar conditions (circles). Zinc-deficient SOD1 controls also yielded greater than baseline scattering (squares). c) Light scattering signal of zinc-deficient SOD1 treated with copper/ascorbateoxidation under various inhibition conditions (for details see Materials and Methods). Anaerobic conditions and copper-chelated by EDTA prevent aggregation whereas free radical scavengers (mannitol, DMPO) do not. Light scattering measurements were made with a PTI QM-1 fluorometer. Excitation and emission wavelengths were set to 350 nm (bandpass = 2 nm).
98
Characterization of Oxidative Modification Sites
Amino acid analysis was performed on oxidized wild-type protein and on
oxidized and aggregated zinc-deficient SOD1 (Table 1). The most striking feature of the
amino acid analysis of both types of oxidized protein is the loss of histidine residues.
Amino acid analysis suggests that three of the eight histidine residues of the SOD1
subunit were modified. It is known that metal-catalyzed oxidation of proteins leads to
conversion of histidine residues to 2-oxohistidine, 4-hydroxy-glutamate, aspartate, or
asparagine(23). Since the glutamate and aspartate contents do not appear to be altered by
oxidation, it is likely that histidines have been largely converted to 2-oxohistidines.
Further support for the conversion to 2-oxohistidine was obtained by sequencing tryptic
peptides of oxidized wild-type SOD1 by LC-MS/MS (Table 2, Fig. 2). The masses of two
tryptic peptides were increased by 16 mass units, which is consistent with the formation
of 2-oxohistidine. Sequencing of the peptides revealed that both His 80 and His 120
contains an additional 16 mass units; these residues are respectively located at the zinc
and copper binding sites of SOD1.
Morphology and structure of SOD1 Aggregates
The results presented above demonstrate that oxidation of select His residues
induces misfolding and aggregation of SOD1. However, the question remains, are these
in vitro aggregates representative of aggregates seen in ALS? Examination of ALS
inclusion bodies by light, electron, and immunoelectron microscopy have shown them to
be a unique feature of ALS and distinct from the amyloid plaques and neurofibrillary
tangles seen in Alzheimer's disease and the intracellular deposits seen in Parkinson's
disease(30-32). In particular, ALS inclusion bodies are not stained by the amyloid dye,
99
Table 1. Comparison of relative amino acid composition of
oxidized to control SOD1
Ratio oxidized/control
Amino Acid Wild-type SOD1 Zn-deficient SOD1
Aspartic acid 1.00 1.01
Threonine 1.02 1.03
Serine 1.01 1.00
Glutamic acid 1.06 1.00
Glycine 1.00 1.03
Valine 1.00 1.03
Isoleucine 0.95 1.02
Leucine 0.99 0.95
Phenylalanine 0.94 0.94
Histidine 0.63 0.62
Lysine 0.95 0.84
Arginine 0.97 0.87
Proline 0.91 0.93
Alanine 1.07 1.00
100
Figure 2. Oxidative modification sites of SOD1 revealed by tryptic digestion and mass spectrometry. SOD (30 mM) was incubated with 2 mM ascorbate, 25 mM copper, 10 mMsodium acetate, pH 5.0 at 37°C for 24h. The protein was reduced and alkylated with DTT and iodoacetamide in 6 M guanidine hydrochloride, and then digested with trypsin (25:1 substrate to enzyme ratio) at 38°C for 50 h. and analyzed by capillary LC-MS/MS. Ribbon diagram created from the PDB coordinates 1SPD, using the program PYMOL (Delano Scientific). Side chains of modified His residues (80 and 120) are shown in purple, the copper ion is colored blue, and the zinc ion is colored gray.
101
Congo red(30). Instead the inclusion bodies seen in COS7 cells expressing ALS mutants
of SOD1(15), transgenic mouse models of ALS(3, 33), and ALS patients(34-37) are all
composed of a mixture of granular aggregates and some thick fibers, as compared to thin
fibrils seen in amyloid diseases(38).
Our atomic force microscopy (AFM) examination of aggregates formed by oxidation of
zinc-deficient SOD1 revealed large amorphous aggregates (< 10 µm diameter) that were
composed of smaller globular particles (0.2 – 0.5 µm diameter) (Fig. 3a), reminiscent of
in vivo inclusion bodies(34-37). Incubation of oxidized wild-type protein at pH 5
produced a scant number of aggregates that could be detected by negative staining
electron microscopy. These heterogeneous aggregates were composed of amorphous
aggregates along with fibrous aggregates that were 40 nm in diameter and several
micrometers long (Fig. 3b). These fibrous aggregates are thicker than the amyloid fibrils
formed by the Alzheimer amyloid peptide, which are 60 – 90 Å in diameter(38).
Dye-binding experiments using thioflavin T, and Congo red, as well as circular dichroism
(CD) were also used to determine whether the SOD1 aggregates possess amyloid
characteristics. A two-fold enhancement of thioflavin T fluorescence was observed with
the aggregates produced from zinc-deficient SOD1 (Fig. 4a); however, the fluorescence
enhancement seen with amyloid fibrils is usually three orders of magnitude higher(39).
Upon binding to Congo red, there was very little, if any, increase in absorbance or
spectral shift (Fig. 4b), which would have been expected had the aggregates in fact been
amyloid(40), but this is well in keeping with the lack of Congo red binding of SOD
inclusion bodies in vivo(30).
102
Figure 3. a) AFM height image of aggregate formed by zinc-deficient SOD1. Aggregates are large and amorphous. Horizontal scale bar = 10 mm, vertical scale = 2 mm. Inset: close up of protein aggregate revealing aggregate is made up of smaller particles. Horizontal scale bar = 2 mm, vertical scale = 1 mm. b) Transmission electron micrograph of SOD1 incubated in the presence of 25 mM copper and 2 mM ascorbate in 10 mM sodium acetate buffer, pH 5.0 for 48 h at 37°C. Scale bar = 400 nm.
103
The CD spectrum of SOD1 undergoes a large change upon oxidation-induced
aggregation (Fig. 4c). However, the CD spectrum of SOD1 aggregates is indicative of
random coil rather than the characteristic β-sheet spectrum of amyloid. Thus, while it
appears that oxidative damage of SOD1 results in misfolding and aggregation, the
resultant aggregates do not appear to be amyloid. Indeed, the morphology of the
aggregates observed in vitro in this study compare favorably with that of granular SOD1
inclusions observed in ALS models and patients.
Structural Changes to SOD1 Prior to Aggregation
To determine whether susceptibility to oxidation induced aggregation of zinc-
deficient SOD1 and SOD1 mutants results from an altered conformation, ANS dye
binding experiments were performed on untreated, unoxidized protein samples. ANS
binding is a probe of exposed hydrophobic surfaces in proteins. Zinc-deficient SOD1
bound the most ANS, wild-type SOD1 did not show any ANS binding, and the SOD1
mutants displayed varying intermediate degrees of ANS binding (Fig. 5). It is known that
the bound zinc in SOD1 helps maintain the structure of the active site and is not directly
involved in catalysis, and removal of zinc destabilizes the enzyme(41). The ANS binding
experiments indicate that in addition to general destabilization, an alteration in
conformation leading to exposure of hydrophobic surface is also associated with zinc
removal. The intermediate levels of ANS binding observed with the SOD1 mutants may
have resulted from an altered looser conformation of the protein in solution, as has been
suggested by the X-ray crystal structures of mutant SOD1(42). Alternatively, the
intermediate ANS binding may result from heterogeneity in the metallation status of the
104
0
2000
4000
6000
8000
10000
12000
14000
400 450 500 550 600
Wavelength (nm)
A
Figure 4. a) Comparison of Thioflavin T binding of zinc deficient SOD aggregates(thin line) and wild type SOD (thick line). 10 µM SOD1 in 10 mM Tris-acetate(pH 7.0) was incubated with 20 µM Thioflavin T before measuring emission spectrum(excitation at 450 nm). Though there is some increase in observed fluorescence intensity, this is far less than the order of magnitude difference typically seen upon ThioflavinT binding to amyloid fibrils; the increase observed is attributed to sequestering of the fluorophore from quenchers. b) Spectral shift assay of aggregates using Congo Red. 6 mM Congo Red (solid line) had comparable absorbance to 3 mM SOD and 6 mM Congo Red (dotted line). c) CD spectra of SOD aggregates (solid line) and native SOD (dotted line). SOD aggregates do not contain a high proportion of beta sheet structure, and together with the dye binding experiments, indicate that these aggregates are likely not amyloid.
105
0
5
10
15
20
25
30
450 500 550 600
AN
S F
luor
esce
nce
(Arb
itrar
y un
its)
Wavelength (nm)
Zn-
Def
icie
nt
A4V
G93
A
D90
A
D12
4N
Figure 5. Comparison of ANS-binding of wild-type (filled circles) to mutant (dashed lines) and zinc-deficient SOD1 (open circles). 10 mM SOD1 in 10 mM Tris-acetate (pH 7.0) was incubated with 20 mM 8-anilino-1-napthalene-sulfonic acid (ANS) before measuring emission spectrum (excitation at 372 nm). Blue shift and increased intensity of ANS fluorescence in mutants and zinc-deficient SOD1 indicates increase exposure of hydrophobic domains. Inset: Integrated ANS fluorescence signal from mutant and zinc-deficient SODs compared to ANS fluorescence of wild-type SOD1 (dashed line).
106
mutants, where mutant preparations that show the greatest ANS binding contain
significant quantities of incompletely metallated protein, much of which could be zinc
deficient.
Concluding Remarks
We have shown that zinc-deficient SOD1, a site-directed mutant with low zinc
binding affinity and low zinc content (D124N) and three FALS associated SOD1 mutants
are much more susceptible to oxidation induced aggregation than the fully metallated
wild-type protein. These findings, coupled with the long half-life of SOD1 in motor
neurons and the high levels of oxidative damage that are known to occur in neural tissues
of ALS patients(21), provide a possible explanation for the SOD1 aggregates observed in
ALS. While it still remains to be established whether the SOD1 aggregates are
intrinsically toxic, there is mounting evidence that protein aggregates exhibit a general
toxicity that is independent of the function of the protein in its native state(43). Our data
are also consistent with the recent model put forward by Okado-Matsumoto and
Fridovich(18) where anti-apoptotic factors, such as the heat shock proteins, are
sequestered by abundant misfolded/aggregated proteins, such as SOD1 or other
misfolded proteins induced by oxidation/nitration, leading to apoptosis.
Acknowledgements
We thank Dr. Harry Ledebur, Dr. Irene Mazzoni, Dr. Jennifer Griffin, and Eric
Thibaudeau for helpful discussions and Dr. Clarissa Desjardins and Lloyd Segal for
encouragement and support. We thank Dr. Yingxin Zhuang for rigorous and meticulous
efforts to produce consistent, high quality purified SOD1 preparations that proved vital to
our studies. We would also like to thank Dr. I. Fridovich for critical evaluation of this
107
work. NRC holds the Jeno and Ilona Diener Chair of Neurodegenerative Diseases.
Support from Caprion Pharmaceuticals Inc., Temerty Family Foundation (to NRC), and
Canadian Institutes of Health Research (to AC) is gratefully acknowledged.
References:
(1) Cleveland, D. W., and Rothstein, J. D. (2001) From Charcot to Lou Gehrig:
deciphering selective motor neuron death in ALS. Nat Rev Neurosci 2, 806-19.
(2) Borchelt, D. R., Lee, M. K., Slunt, H. S., Guarnieri, M., Xu, Z. S., Wong, P. C.,
Brown, R. H., Jr., Price, D. L., Sisodia, S. S., and Cleveland, D. W. (1994) Superoxide
dismutase 1 with mutations linked to familial amyotrophic lateral sclerosis possesses
significant activity. Proc Natl Acad Sci U S A 91, 8292-6.
(3) Bruijn, L. I., Houseweart, M. K., Kato, S., Anderson, K. L., Anderson, S. D.,
Ohama, E., Reaume, A. G., Scott, R. W., and Cleveland, D. W. (1998) Aggregation and
motor neuron toxicity of an ALS-linked SOD1 mutant independent from wild-type
SOD1. Science 281, 1851-4.
(4) Yim, M. B., Kang, J. H., Yim, H. S., Kwak, H. S., Chock, P. B., and Stadtman, E.
R. (1996) A gain-of-function of an amyotrophic lateral sclerosis-associated Cu,Zn-
superoxide dismutase mutant: An enhancement of free radical formation due to a
decrease in Km for hydrogen peroxide. Proc Natl Acad Sci U S A 93, 5709-14.
(5) Wiedau-Pazos, M., Goto, J. J., Rabizadeh, S., Gralla, E. B., Roe, J. A., Lee, M.
K., Valentine, J. S., and Bredesen, D. E. (1996) Altered reactivity of superoxide
dismutase in familial amyotrophic lateral sclerosis. Science 271, 515-8.
(6) Liochev, S. I., and Fridovich, I. (2002) Nitroxyl (NO-): a substrate for superoxide
dismutase. Archives Of Biochemistry And Biophysics 402, 166-171.
108
(7) Liochev, S. I., and Fridovich, I. (2002) Copper, zinc superoxide dismutase and
H2O2 - Effects of bicarbonate on inactivation and oxidations of NADPH and urate, and
on consumption of H2O2. Journal Of Biological Chemistry 277, 34674-34678.
(8) Lyons, T. J., Liu, H., Goto, J. J., Nersissian, A., Roe, J. A., Graden, J. A., Cafe,
C., Ellerby, L. M., Bredesen, D. E., Gralla, E. B., and Valentine, J. S. (1996) Mutations in
copper-zinc superoxide dismutase that cause amyotrophic lateral sclerosis alter the zinc
binding site and the redox behavior of the protein. Proc Natl Acad Sci U S A 93, 12240-
4.
(9) Crow, J. P., Sampson, J. B., Zhuang, Y., Thompson, J. A., and Beckman, J. S.
(1997) Decreased zinc affinity of amyotrophic lateral sclerosis-associated superoxide
dismutase mutants leads to enhanced catalysis of tyrosine nitration by peroxynitrite. J
Neurochem 69, 1936-44.
(10) Klapper I, H. R., Fine R, Sharp K, Honig B. (1986) Focusing of electric fields in
the active site of Cu-Zn superoxide dismutase:
effects of ionic strength and amino-acid modification. Proteins 1, 47-59.
(11) Estevez, A. G., Crow, J. P., Sampson, J. B., Reiter, C., Zhuang, Y., Richardson,
G. J., Tarpey, M. M., Barbeito, L., and Beckman, J. S. (1999) Induction of nitric oxide-
dependent apoptosis in motor neurons by zinc-deficient superoxide dismutase. Science
286, 2498-500.
(12) Subramaniam, J. R., Lyons, W. E., Liu, J., Bartnikas, T. B., Rothstein, J., Price,
D. L., Cleveland, D. W., Gitlin, J. D., and Wong, P. C. (2002) Mutant SOD1 causes
motor neuron disease independent of copper chaperone-mediated copper loading. Nat
Neurosci 5, 301-7.
109
(13) Culotta, V. C., Klomp, L. W., Strain, J., Casareno, R. L., Krems, B., and Gitlin, J.
D. (1997) The copper chaperone for superoxide dismutase. J Biol Chem 272, 23469-72.
(14) Wong, P. C., Waggoner, D., Subramaniam, J. R., Tessarollo, L., Bartnikas, T. B.,
Culotta, V. C., Price, D. L., Rothstein, J., and Gitlin, J. D. (2000) Copper chaperone for
superoxide dismutase is essential to activate mammalian Cu/Zn superoxide dismutase.
Proc Natl Acad Sci U S A 97, 2886-91.
(15) Koide, T., Igarashi, S., Kikugawa, K., Nakano, R., Inuzuka, T., Yamada, M.,
Takahashi, H., and Tsuji, S. (1998) Formation of granular cytoplasmic aggregates in
COS7 cells expressing mutant Cu/Zn superoxide dismutase associated with familial
amyotrophic lateral sclerosis. Neurosci Lett 257, 29-32.
(16) Johnston, J. A., Dalton, M. J., Gurney, M. E., and Kopito, R. R. (2000) Formation
of high molecular weight complexes of mutant Cu, Zn-superoxide dismutase in a mouse
model for familial amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A 97, 12571-6.
(17) Bruening, W., Roy, J., Giasson, B., Figlewicz, D. A., Mushynski, W. E., and
Durham, H. D. (1999) Up-regulation of protein chaperones preserves viability of cells
expressing toxic Cu/Zn-superoxide dismutase mutants associated with amyotrophic
lateral sclerosis. J Neurochem 72, 693-9.
(18) Okado-Matsumoto, A., and Fridovich, I. (2002) Amyotrophic lateral sclerosis: a
proposed mechanism. Proc Natl Acad Sci U S A 99, 9010-4.
(19) Pardo, C. A., Xu, Z., Borchelt, D. R., Price, D. L., Sisodia, S. S., and Cleveland,
D. W. (1995) Superoxide dismutase is an abundant component in cell bodies, dendrites,
and axons of motor neurons and in a subset of other neurons. Proc Natl Acad Sci U S A
92, 954-8.
110
(20) Borchelt, D. R., Wong, P. C., Becher, M. W., Pardo, C. A., Lee, M. K., Xu, Z. S.,
Thinakaran, G., Jenkins, N. A., Copeland, N. G., Sisodia, S. S., Cleveland, D. W., Price,
D. L., and Hoffman, P. N. (1998) Axonal transport of mutant superoxide dismutase 1 and
focal axonal abnormalities in the proximal axons of transgenic mice. Neurobiol Dis 5, 27-
35.
(21) Bowling, A. C., Schulz, J. B., Brown, R. H., Jr., and Beal, M. F. (1993)
Superoxide dismutase activity, oxidative damage, and mitochondrial energy metabolism
in familial and sporadic amyotrophic lateral sclerosis. J Neurochem 61, 2322-5.
(22) Ferrante, R. J., Browne, S. E., Shinobu, L. A., Bowling, A. C., Baik, M. J.,
MacGarvey, U., Kowall, N. W., Brown, R. H., Jr., and Beal, M. F. (1997) Evidence of
increased oxidative damage in both sporadic and familial amyotrophic lateral sclerosis. J
Neurochem 69, 2064-74.
(23) Stadtman, E. R., and Berlett, B. S. (1997) Reactive oxygen-mediated protein
oxidation in aging and disease. Chemical Research In Toxicology 10, 485-494.
(24) Stadtman, E. R. (1993) Oxidation Of Free Amino-Acids And Amino-Acid-
Residues In Proteins By Radiolysis And By Metal-Catalyzed Reactions. Annual Review
Of Biochemistry 62, 797-821.
(25) Rice, M. E. (2000) Ascorbate regulation and its neuroprotective role in the brain.
Trends In Neurosciences 23, 209-216.
(26) Banci, L., Bertini, I., Cabelli, D. E., Hallewell, R. A., Tung, J. W., and Viezzoli,
M. S. (1991) A Characterization Of Copper-Zinc Superoxide-Dismutase Mutants At
Position-124 - Zinc-Deficient Proteins. European Journal Of Biochemistry 196, 123-128.
111
(27) Khossravi, M., Shire, S. J., and Borchardt, R. T. (2000) Evidence for the
involvement of histidine A(12) in the aggregation and precipitation of human relaxin
induced by metal-catalyzed oxidation. Biochemistry 39, 5876-5885.
(28) Li, S. H., Nguyen, T. H., Schoneich, C., and Borchardt, R. T. (1995) Aggregation
And Precipitation Of Human Relaxin Induced By Metal-Catalyzed Oxidation.
Biochemistry 34, 5762-5772.
(29) Requena, J. R., Groth, D., Legname, G., Stadtman, E. R., Prusiner, S. B., and
Levine, R. L. (2001) Copper-catalyzed oxidation of the recombinant SHa(29-231) prion
protein. Proceedings Of The National Academy Of Sciences Of The United States Of
America 98, 7170-7175.
(30) Okamoto, K., Hirai, S., Yamazaki, T., Sun, X. Y., and Nakazato, Y. (1991) New
Ubiquitin-Positive Intraneuronal Inclusions In The Extra-Motor Cortices In Patients With
Proteinacious intracellular aggregates in motor neurons are a key feature of both sporadic
and familial amyotrophic lateral sclerosis (ALS). These inclusion bodies are often
immuno-reactive for Cu, Zn superoxide dismutase (SOD1) and are implicated in the
pathology of ALS. On the basis of this and similar clinical presentation of symptoms in
the familial (fALS) and sporadic (sALS) forms of ALS, we sought to investigate the
possibility that there exists a common, disease related, aggregation pathway for fALS-
associated mutant SODs and wild type SOD1. We have previously shown that oxidation
of fALS-associated mutant SODs produces aggregates that have the same morphological,
structural and tinctorial features as those found in SOD1 inclusion bodies in ALS. Here,
we show that oxidative damage of wild type SOD at physiological concentrations
(~40µM) results in destabilization and aggregation in vitro. Oxidation of either mutant or
wild type SOD1 causes the enzyme to dissociate to monomers prior to aggregation. Only
small changes in secondary and tertiary structure are associated with monomer formation.
These results indicate a common aggregation prone monomeric intermediate for wild
type and fALS-associated mutant SODs, and provides a link between sporadic and
familial ALS.
117
Introduction
ALS is a fatal neurodegenerative disease that leads to the selective loss of motor neurons.
Although ALS is predominately a sporadic disease, ~10% of cases are inherited in an
autosomal dominant manner, and a subset of these fALS cases are caused by mutations in
the SOD1 gene (1). The gene product of SOD1, cytoplasmic Cu/Zn superoxide dismutase
(SOD1), is a ubiquitously expressed enzyme that catalyzes the disproportionation
reaction of superoxide radicals (1). There are several lines of evidence that SOD1
mutations result in a gain, rather than loss of function that causes ALS. For instance,
some fALS-associated mutant SOD1s retain full enzymatic activity (2). In addition,
SOD1 knock-out mice lack ALS symptoms, while transgenic mice expressing the fALS
associated mutant G93A SOD1 develop ALS-like symptoms, despite expression of
endogenous mouse SOD1 (3). Lastly, over-expression of human wild-type SOD1 fails to
alleviate symptoms in this transgenic mouse model for ALS (3). One hypothesis of the
gain of function of SOD1 is that misfolding of the mutant alters the catalytic mechanism
to allow production of oxidants, such as peroxynitrite (4) and possibly hydrogen peroxide
(5). These reactive nitrogen and oxygen species cause toxicity by accumulated damage to
proteins, nucleic acids and lipids. Another major hypothesis is toxicity caused by
intracellular aggregation of SOD1. SOD1 inclusion bodies, which also react with anti-
ubiquitin antibodies, are a common pathological finding in motor neurons and
neighboring astrocytes of ALS patients (6). These two hypotheses, however, are not
mutually exclusive when considering that oxidative modification of proteins may
contribute to aggregation and protease resistance. Protein aggregates are a common
pathological feature of many neurological disorders (7), including Huntington’s,
118
Alzheimer’s and Parkinson’s diseases. In each case, misfolding and aggregation appears
to be linked to cytotoxicity (7). Similarly, the aggregation propensity of mutant SOD1s
may be the mechanism by which over 100 disparate mutations cause a common ALS
phenotype. SOD1 aggregates may be inherently toxic or cause motor neuron toxicity by
sequestering chaperones and blocking proper functioning of the proteosome (8). Indeed,
over-expression of either the chaperone Hsp70 (9), or the ubiquitin E3 ligase Dorfin (10),
preserves the viability of cells expressing fALS mutant SOD1s by decreasing the number
of intracellular SOD1 aggregates.
Understanding the details of the biophysical pathway by which SOD1 aggregation occurs
might then yield avenues for novel therapeutics for ALS. In order to carry out such an
investigation, we previously developed a physiologically relevant cell-free model of
SOD1 aggregation (11). As a natural free radical scavenger and metalloprotein, SOD1 in
motor neurons is exposed to greater oxidative stress than other proteins. Based on this
occupational hazard, as well as the long half-life(12) and high concentration of SOD1 in
motor neurons(13), we proposed that SOD1 is especially susceptible to oxidation-induced
aggregation in motor neurons. Indeed, it has been shown that ALS patients display higher
levels of oxidative stress than age-matched controls(14) and transgenic mouse models for
fALS also show higher levels of oxidized SOD1 (15). In our cell-free model, we used
metal catalyzed oxidation (MCO), a physiologically relevant process, to induce oxidative
stress. MCO may represent the major cause of SOD1 aggregation in vivo, because it is
the major source of hydroxyl radicals in vivo, it has been implicated as a major cause of
protein modification in aging, and it marks proteins for degradation (16). Our results
showed that fALS-associated mutants of SOD1 were extremely prone to form visible
119
aggregates upon oxidation, compared with wild-type protein (11). Examination of the
morphology, structure and tinctorial properties of these aggregates demonstrated their
similarity to those found in intracellular inclusion bodies seen in ALS patients and in cell
lines . ALS inclusion bodies are not stained with the amyloid dye Congo Red(17), and are
all composed of a mixture of granular aggregates and some thick fibers(18-21), as
compared with thin fibrils seen in amyloid diseases(7). Examination of SOD1 aggregates
produced in our model system by electron microscopy, atomic force microscopy and dye
binding demonstrated that they had identical properties with in vivo aggregates. Finally,
the in vitro aggregates had circular dichroism spectra consistent with non-specific
aggregation seen in vivo(11). In the present study, we use the same model to dissect the
kinetics of the SOD1 aggregation pathway.
SALS and fALS have similar clinical features, and inclusion bodies are seen in both
forms of the disease (22). Hyaline-like inclusion bodies that are SOD1 immuno-reactive
are typical of fALS (23), while skein-like inclusion bodies, which can be SOD1 immuno-
reactive (24-26), are found in motor neurons of sALS patients (23). We use our model
with demonstrated in vivo relevance (11) to compare and contrast the aggregation
pathways of wild type and mutant SOD1 to understand the similarities and differences in
the sporadic and familial forms of the disease. We find that SOD1, normally a dimeric
enzyme, dissociates to monomers prior to aggregation for both wild-type and mutant
proteins; this dissociation is accompanied by minimal changes in the secondary structure
of SOD1. This common intermediate suggests a common pathway for the aggregation of
mutant and wild type SOD1 and provides a mechanistic link between sporadic and
familial ALS.
120
Materials and Methods
In Vitro Aggregation of SOD1
Wild-type SOD1 from human erythrocytes and all other reagents were from Sigma-
Aldrich. Mutant human SOD1s were prepared as previously reported (11). MCO
reactions consisted of 10-100µM SOD1, 4mM ascorbic acid and 0.2mM CuCl2 in 10mM
Tris-acetate buffer, pH 7.0, whereas control reactions were 10-100µM SOD1 in buffer;
reactions were incubated at 37°C, 24-48 hrs.
Right Angle Light Scattering
Scattering measurements were made with a Photon Technology International QM-1
fluorescence spectrophotometer. Excitation and emission wavelengths were set to 350
nm. All samples were vortexed prior to measurement to dislodge aggregates.
ANS Binding
Oxidized and control SOD1 were incubated for 30 minutes with 20µM 8-anilino-1-
napthalene-sulfonic acid (ANS) before measuring emission spectrum with excitation at
372 nm. The signal was integrated from 400 to 600nm.
Dynamic Light Scattering (DLS)
Data was collected with a DynaPro Protein Solutions DLS module at 37°C, 10s
averaging time. MCO reactions for either 40µM wild type SOD1 or 10µM mutant SOD1
(A4V) were initiated by the addition of CuCl2. Consecutive measurements (>10) were
used for the regularization analysis by DYNAMICS software. Autocorrelation
coefficients for each of the minimum 10 measurements for each time point were averaged
and normalized to a maximum of 2 and a minimum of 1 for comparability.
Analytical Ultracentrifugation
121
Sedimentation was performed at 20°C on a Beckman XLI analytical ultracentrifuge using
an AN50-Ti rotor. Sedimentation equilibrium runs using six-channel charcoal-Epon cells
were performed for 24hrs prior to data acquisition. Molecular weight determinations
involved global analysis of data acquired at 5 different speeds using Beckman XL-I
software, where absorbance versus radial position data were fitted to the sedimentation
equilibrium equation using non-linear least squares techniques. The partial specific
volume and density of the sample were calculated using the program SEDENTERP from
the amino acid sequence and buffer composition, respectively.
Differential Scanning Calorimetry
Differential scanning calorimetric studies were performed using a Nano DSC
(Calorimetry Sciences, Provo, UT, USA). Samples of 10µM oxidized and control SOD1
were concentrated to 30µM (~1mg/ml) using Amicon membrane concentrators (15kDa
molecular weight cut off). DSC traces of 1 mg/ml fresh, oxidized and control SOD1s in
10mM Tris-acetate buffer, pH 7.0 were obtained sequentially. The identical solution
without protein was used as a reference. Both sample and reference solutions were
degassed at 4°C before scanning from 20 to 120°C at 1°C/min.
Circular Dichroism (CD)
The CD spectra of oxidized and unoxidized SOD1 were recorded on an Aviv Circular
Dichroism Spectrometer Model 62 DS at 25 °C in a 1mm path length cell, using 1nm
bandwidth.
Tryptophan Fluorescence
122
Fluorescence measurements were taken on a Photon Technology International QM-1
fluorescence spectrophotometer with excitation wavelength set to 280nm, and a band
pass of 4nm.
Capillary LC-MS/MS
SOD1 (30µM) was incubated with 2 mM ascorbate, 25 �M copper, 10 mM sodium
acetate, pH 5.0 at 37°C for 24h. The protein was reduced and alkylated with DTT and
iodoacetamide in 6 M guanidine hydrochloride, diluted and then digested with trypsin
(25:1 substrate to enzyme ratio) at 38°C for 50 h. and analyzed by capillary LC-MS/MS.
Peptides were analyzed using a Q-TOF Ultima mass spectrometer (Micromass,
Manchester, UK) coupled to a capillary HPLC. The fraction of the sample that did not
bind to the C18 column (mostly guanidine HCl) was redirected to the waste. Peptides
eluted by acetonitrile were ionized by electrospray and peptide ions were automatically
selected and fragmented in a data dependent acquisition mode. Database searching was
done with Mascot (Matrix Science).
Results
Wild Type SOD1 is Aggregation-Prone Under Oxidative Stress
Protein aggregation is inherently a highly concentration-dependent process. In order to
ensure the physiological relevance of our in vitro aggregation system, we used
neurophysiological concentrations of SOD1. Kurobe et al (27) found that the SOD1
content of erythrocytes is 0.95 ± 0.07 µg SOD1/mg hemoglobin, which translates to a
molar concentration of 10µM (based on normal values for mean corpuscular hemoglobin
and volume (28)). The concentration of SOD1 is approximately 4.5 times greater in brain
than in erythrocytes (27), corresponding to a concentration of 45µM. Intracellular SOD1
123
concentration in motor neurons may be even higher (13). Therefore we used the 0 to
100µM concentration range.
MCO of SOD1 at neutral pH was used to induce aggregation. The concentration
dependence of the aggregation reaction measured by right angle light scattering is shown
in Fig. 1A. At low concentrations of SOD1, light scattering from both oxidized and
control samples were comparable to buffer, indicating that protein aggregates were
absent. Increasing SOD1 concentrations caused the appearance of large visible
aggregates (> 350nm diameter) in oxidized but not in control samples, where scattering
remains at buffer baseline levels. Light scattering levels were proportional to SOD1
concentrations; however, the absolute scattering intensity from 100µM oxidized SOD1 is
significantly lower (~30%) than the scattering intensity from only 10µM zinc-deficient
SOD1 (11). The amount of scattered light is proportional to both the size and number of
aggregates so there must be either fewer or smaller aggregates in the case of wild type
holo-SOD1 than zinc deficient SOD1. These data suggest that the aggregation propensity
of wild-type SOD1 is lower than that of zinc deficient SOD1 or fALS-associated SOD1
mutants.
ANS fluorescence is a selective probe of the molten globule state of a protein and is
proportional to hydrophobic surface area available for binding fluorophores (29). The
fluorescence enhancement due to ANS-aggregate interactions is shown in Fig. 1B. ANS
shows a slight increase in fluorescence in the MCO buffer over the control buffer in the
absence of SOD1 (0µM SOD1) and ANS fluorescence increases with concentration of
unoxidized SOD1, possibly due to minimal binding to native SOD1. These effects,
however, are small compared to the large increase in fluorescence upon binding SOD1
124
Figure. 1. Wild type SOD1 aggregation. Top: Right Angle Light Scattering measurements of different concentrations of oxidized (filled) and control (open) SOD1. Bottom: ANS binding of SOD1 aggregation under oxidative (filled) and control (open) conditions.
0
50
100
150
200
250
300
350
0 20 40 60 80 100
R
ight
Ang
le
Ligh
t Sca
tterin
g
0
5
10
15
20
25
30
SOD concentration (µM)
AN
S F
luor
esce
nce
0 20 40 60 80 100
A
B
125
aggregates. The increase observed (Fig. 1B) is a sensitive indicator of the amount of
aggregated protein and parallels the increase in right angle light scattering (Fig. 1A).
Fig. 2 shows the differential scanning calorimetry (DSC) thermogram obtained for the
thermal denaturation of both oxidized and unoxidized SOD1. The DSC trace of freshly
prepared SOD1 (solid line) shows two endothermic transitions, with Tm=102°C and
85°C. The higher temperature transition is dominant and corresponds to the heat
denaturation of the fully metallated SOD1, while the lower transition is representative of
overnight in Tris/acetate buffer, pH 7.0, produced a DSC trace that is qualitatively similar
to that of the freshly prepared sample. The proportion of metal-deficient SOD1 increases,
and the beginning of a high temperature transition is visible at 100°C. However, the high
temperature transition is largely masked by a very large exotherm centered at 110°C,
which is probably due to aggregation. Thus, there are some changes to the SOD1
structure upon incubation at 37°C. Oxidized SOD1 shows a large exothermic transition in
the DSC trace from 55°C to 78°C, which is indicative of an in situ aggregation event
commencing at a temperature 50°C lower than for unoxidized SOD1. An endothermic
transition corresponding to the thermal denaturation of partially metallated SOD1 at 83°C
is also observed, but the magnitude of this transition is difficult to determine accurately
due to baseline instability above 80°C. No evidence of a transition at Tm=102°C
corresponding to native SOD1 was observed for the oxidized SOD1. Taken as a whole,
these results indicate that metal catalyzed oxidation of wild type SOD1 yields a species
that is less stable and more aggregation prone, but only aggregates at high concentrations
(>40µM).
126
0 20 40 60 80 100 120 140-40000
-20000
0
20000
Temperature 0C
Cex
p(c
al/K
/mole
dim
er)
Figure. 2. DSC thermograms of SOD1. Unoxidized SOD1 (solid line); unoxidized SOD1 incubated at 37°C for 18hrs (dashed line). Oxidized SOD1 (dotted line) displayed an exothermic peak at 75°C.
127
SOD1 Dimers Dissociate to Monomers Prior to Aggregation
Cu, Zn SOD1 is present in eukaryotes ubiquitously as a stable dimer, held together with
hydrophobic contacts (31). To investigate the kinetics of SOD1 aggregation, we used
dynamic light scattering (DLS) and regularization analysis to deconvolute the
autocorrelation decay curves and obtain multiple species fits. The molecular weight is
then extracted by volume shape hydration. We found that unoxidized SOD1 has a
hydrodynamic radius of 2.54nm, corresponding to a molecular weight of 35.4kDa, or
approximately the sequence molecular weight of the dimer (data not shown). When
studying the kinetics of aggregate formation using DLS, the regularization results
indicate that SOD1 is initially dimeric, as expected (Fig. 3, 2 minutes). Upon oxidation
for 30 minutes (Fig. 3, 30 minutes) however, the mass weighted regularization results
show the major species to be 1.94nm particles, corresponding to a molecular weight of
17kDa, or approximately the monomer molecular weight. After 140 minutes (Fig. 3, 140
minutes), the appearance of large particles (>100nm) dominates the regularization
analysis. Normalized autocorrelation coefficient curves representative of each time point
are shown in Fig. 3, bottom panel. Longer correlation times are indicative of smaller
diffusion coefficients (larger particles). An overlay of the autocorrelation coefficients for
different time points clearly shows an initial decrease in correlation time (smaller
particles) followed by an increase. Equivalent results were obtained using the fALS
mutant A4V SOD1 (Supporting Information Fig. 1). Analysis of kinetic data for DLS
thus indicates the formation of SOD1 monomers as an intermediate prior to aggregation.
Equilibrium analytical ultracentrifugation (AUC) was used to determine the
molecular weight of oxidized and unoxidized SOD1. Sedimentation equilibrium data at
128
Figure. 3. DLS analysis of SOD1 during oxidation. Top three panels: Mass-weighted distributions of particles present after 2, 30 and 140 minutes of MCO. Dimeric SOD1 (radius~2.5nm) dissociates to monomers (radius~1.94nm) prior to aggregation (radius >100nm). Bottom Panel: DLS Correlation decay curves after 2 minutes [dimer] (black), 30 minutes [monomer] (open) and 140 minutes [aggregated] (gray) SOD1.
129
Supporting Information Figure 1. DLS analysis of G93A SOD1 during oxidation. Top Panel:Unoxidized G93A SOD1 ~ dimeric (radius~2.5nm), Middle Panel: After oxidation for 10 minutes ~ monomers (radius~1.94nm), Bottom Panel: After oxidation for 30 minutes, G93A SOD1 exists primarily as aggregates.
130
rotor speeds of 20-45k rpm in 5k increments was fit globally using the Beckman XL-I
software package. Representative data for oxidized and unoxidized wild type SOD1 at
35k rpm are shown in Fig. 4. The weight-averaged molecular weight of unoxidized
SOD1 was 30,649 Da (95% confidence interval 30,200- 31,098 Da), which is close to the
fully metallated molecular weight (31950 Da). Oxidized SOD1, however, has a weight
averaged molecular weight of 27,299 Da (95% confidence interval 26,880- 27,716 Da),
only 85.3% of the fully metallated dimer molecular weight. The data for oxidized SOD1
can also be described by a monomer-dimer equilibrium with Kd = 1.8 x 10-5 M. Analysis
of AUC data for the FALS mutant G93A SOD1 yielded a weight averaged molecular
weight of 30,517 Da (95% confidence interval 30,037- 30,998 Da) for the control and
27,683 Da (95% confidence interval 27,184- 28,181 Da) for the oxidized G93A SOD1
sample (Supporting Information Fig. 2). Again, the data for oxidized G93A SOD1 could
be described by a monomer-dimer equilibrium with Kd = 1.2 x 10-5 M. Soluble
unoxidized SOD1 samples are stable dimers for both wild type and G93A mutant SOD1
(Kd is too small to measure by AUC), whereas both wild type and G93A mutant SOD1
likely exist as a combination of monomers and dimers upon oxidation. Based on the
micromolar dissociation constant of the oxidized SOD1s and the normal cellular
concentrations in motor neurons (>40µM), a significant fraction of the oxidized proteins
should be monomeric in vivo.
Minor Changes to Conformation of SOD1 upon Oxidation
We have observed that SOD1 becomes aggregation prone and becomes at least partially
monomeric upon MCO. High-resolution structure determination (x-ray crystallography,
NMR) is usually reliant upon high protein concentration, much higher than those at
131
Radius (cm)5.9 6.0 6.1
0.0
0.5
1.0
1.5
Absorb
ance
-6
0
5
Resid
uals
Absorb
ance
Resid
uals
A
B
6.9 7.0 7.1
0.0
0.5
1.0
1.5
-6
0
4
Figure. 4. Analytical ultracentrifugation of oxidized and control SOD1. a) Unoxidized SOD1 has a weight average molecular weight of 30657 Da. b) Oxidized SOD1 gives a weight-averaged molecular weight of 27266 Da. The data was also adequately described by a monomer-dimer equilibrium with Kd = 1.8 x 10-5(M).
132
6.9 7.0 7.1
0.0
0.5
1.0
1.5
Abso
rbanc
e
RadiusM = 30517 ( 30037, 30998 )
-10-8-6-4-20246
Resi
dual
s
M = 27682 ( 27184, 28181 )
5.9 6.0 6.1
0.0
0.5
1.0
1.5Abso
rbanc
e
Radius
-8-6-4-2024
Resi
dual
s
A B
Supporting Information Figure 2. Analytical Ultracentrifugation of oxidized and control G93A-SOD1 a) a) Unoxidized G93A-SOD1 has a weight average molecular weight of 30517 Da. b) Oxidized SOD1 gives a weight-averaged molecular weight of 27266 Da. The data was also adequately described by a monomer-dimer equilibrium with Kd = 1.2 x 10-5(M).
133
which oxidized SOD1 will precipitate. In order to elucidate the changes to SOD1
structure upon oxidation, we utilized CD to evaluate changes in secondary structure and
tryptophan fluorescence to study changes in tertiary structure (Fig. 5a and 5b,
respectively). The CD spectra of oxidized and freshly prepared wild type SOD1 are
almost identical (Fig. 5a). There is thus little change in secondary structure in SOD1,
which is composed of a β-barrel and two metal ion binding loop regions that may
contribute to a CD signal. Tryptophan fluorescence, a highly environmentally sensitive
fluorophore, is comparable in oxidized and control SOD1 (Fig. 5b). There is no change in
the emission maximum and little change in intensity, indicating that the tryptophan
residue is in a very similar chemical environment in the oxidized and unoxidized forms of
SOD1.
Mapping of Oxidative Modification Sites
We have previously shown that histidine residues of SOD1 are selectively modified by
MCO (11). Here, we present a comprehensive analysis of the sites of oxidative
modification. LC-MS/MS results of oxidized and unoxidized SOD1 are summarized in
Fig. 6 and in Supporting Information Table 1. Three active site histidine modifications
are observed (His48, 80, 120), well in keeping with the idea that MCO produces
modifications local to metal binding sites (16). Interestingly, two non-active site
modifications (His110, Phe20) are also observed, suggesting that there may be additional
transient metal binding sites in SOD1. This site may be part of a weak non-active site
copper-binding motif comprising residues 109-111 (32). Phe20, normally buried in the
hydrophobic core of the SOD1 β-barrel, is also modified, indicating that even though the
134
Wavelength (nm)
Try
pto
phan F
luore
sce
nce
0
10
20
300 350 400 450
-0.03
0
0.03
0.06
190 210 230 250 270
Mean R
esid
ue E
lipticity A
B
Figure. 5. a) CD spectra of oxidized (but not aggregated) SOD1 vs. unoxidized SOD1. CD spectrum of 10mM oxidized wtSOD (filled; aggregation does not occur at this concentration) and unoxidized SOD1 (open). b) Tryptophan fluorescence of SOD1. Oxidized (filled) and unoxidized(open) fluorescence spectra.
135
Figure. 6. Oxidative modification sites of SOD1. Ribbon diagram created from the PDB coordinates 1SPD, using the program PYMOL (Delano Scientific). Side chains of modified His residues (48, 80, 110 and 120) are shown in purple, the modified Phe residue (20) is shown in pink, the copper ion is colored blue, and the zinc ion is colored gray.
136
Supplementary Information Table 1. Summary of Mass Spectroscopic Analysis of Tryptic Fragments of Oxidized SOD1
Experimental Mass Tryptic Peptide Sequence Position
Gredal,O., Wong,P.C., Williams,D.S., and Cleveland,D.W. (2004). Toxicity of Familial
ALS-Linked SOD1 Mutants from Selective Recruitment to Spinal Mitochondria. Neuron
43, 5-17.
3. Okado-Matsumoto,A. and Fridovich,I. (2001). Subcellular distribution of
superoxide dismutases (SOD) in rat liver: Cu,Zn-SOD in mitochondria. J.Biol.Chem.
276, 38388-38393.
Results
Antibody Design and Validation
Investigating protein conformation in vivo is a challenging problem. One possible
strategy is to design an antibody that will recognize specific misfolded conformations but
not the native protein. This immunological approach has been previously applied to other
neurodegenerative disorders involving protein aggregation, but these designs have relied
on low resolution biophysical information on the structure of the misfolded protein(14,
15). Our strategy employs the use of high-resolution X-ray crystal structure data to design
an antibody against misfolded SOD1. Examination of the X-ray structure of the native
SOD1 dimer (pdb code: 1SPD)(16) reveals that residues 145-151 are sequestered in the
SOD1 dimer interface and are inaccessible in native SOD1. An antibody raised against
this epitope is predicted to recognize misfolded forms of SOD1 where the native dimer
interface is disrupted and exposed, such as in monomers and non-native oligomers.
Accordingly, we named this the SOD1 Exposed Dimer Interface (SEDI) antibody. We
162
synthesized a multiple antigenic peptide where each branch of the dendrimer had the
sequence ggRLACGVIGIggkg; the capitalized sequence is part of the SOD1 sequence
(residues 143-151; Fig. 1a-b). SOD1 residues 143 and 144 were added to the antigenic
peptide to increase its solubility; the N-terminal and C-terminal Gly/Lys linkers were
added to contextualize the epitope to an internal sequence, increase solubility, and
increase molecular weight for enhanced immunogenicity. Rabbit anti-serum produced
from immunization with this antigen was affinity purified using an immobilized linear
peptide with identical sequence to the antigen. To demonstrate that the SEDI antibody
reacts selectively with mis/unfolded SOD1, its reactivity with native folded SOD1, urea
unfolded SOD1 (Supplementary Fig. 1, online), and oxidation-induced aggregates of
SOD1 was evaluated in ELISA. The SEDI antibody reacts with urea mis/unfolded SOD1
and with oxidation-induced aggregates that arise from a transient monomeric SOD1
intermediate(12), but not with natively folded SOD1 (Fig. 1d). The specificity of the
SEDI antibody for dimer-disrupted/dissociated forms of SOD1 was further demonstrated
in immunoprecipitation reactions where it reacts with urea unfolded SOD1 (Fig. 1c, lane
5) but not native dimeric SOD1 (Fig. 1c, lane 4). In contrast to commercially available
SOD1 antibodies that detect both native and misfolded SOD1, the SEDI antibody reacts
selectively with SOD1 conformers in which the dimer interface epitope is exposed, but
not with native SOD1 (Supplementary Fig. 2). The SEDI antibody thus satisfies the
design criteria and provides a unique tool for studying the in vivo misfolding pathway of
SOD1.
163
Figure 1. (next page) Design and validation of SEDI antibody, which selectively recognizes monomer/misfolded SOD1, but not native dimeric SOD1. a) Surface representation of native dimeric SOD1 (green) with buried epitope shown in red. b) Surface representation of monomeric SOD1 with now exposed epitope shown in red. Figures were prepared using PyMol(Delano Scientific). c) Examination of SEDI specificity in immunoprecipitation reactions, as detailed in Supplementary Methods. SEDI reacted only with unfolded SOD1 (U, lane 5) and not with the native dimer (F, lane 4), where the amount of SOD1 available for immunoprecipitation was similar (lanes 1 and 2). Lane 3 is left blank for clarity. d) SEDI specificity in ELISA. Unfolded SOD1 (as in 1g), oxidized SOD1 aggregates (as per Rakhit et al(12)), folded native SOD1 or lysozyme were deposited on hydrophobic ELISA plates. SEDI recognizes unfolded or aggregated SOD1, but not native SOD1, whereas commercially available anti-SOD1 (StressGen) recognizes all three forms. This falls within the linear range of the SEDI antibody (data not shown). e) Anti-SOD1 Western blot of SEDI immunoprecipitation from wild-type overexpressing (wt O/E) or G93A-SOD1 mice. Lane 1: SOD1 positive control, purified from human erythrocytes (directly loaded onto gel, no IP); Lanes 2-9, IP: SEDI, Western: Stressgenanti-SOD1. Lane 2: IP of spinal cord homogenate from nontransgenic littermate of G93A SOD1 mouse; lanes 3-5: SEDI IP of wild-type SOD1 overexpressing mouse tissues (lane 3: wt O/E brain pellet, lane 4: wt O/E brain supernatant, lane 5: wt O/E spinal cord supernatant); lanes 6-8: SEDI IP of G93A-SOD1 mouse tissues (lane 6: G93A brain pellet, lane 7: G93A brain supernatant, lane 8: G93A spinal cord supernatant). f) Anti-SOD1 Western blot alongside SEDI immunoprecipitations from 750mg of G37R(top), G85R(middle) or G93A(bottom) SOD1 ALS-mouse spinal cord homogenates. Lanes 1-4 are 1ml, 2ml, 4ml and 10ml of 0.5% immunoprecipitation supernatant; the lower bands in the G37R and G93A spinal cord correspond to endogenous mouse SOD1. G85R SOD1 runs at the same molecular weight as mouse SOD1. Lane 5 is the SEDI immunoprecipitation with 10mg of antibody. Bands were quantitated using Image J (NIH).
164
Figure 1.
165
Supplementary Figure 1. SOD1 unfolding was followed by changes to tryptophan fluorescence (ex. = 280nm, em. = 350nm). SOD1 unfolds over the course of a few hours in the presence of 8M urea, 2mM DTT and 1mM EDTA (red curve), and does not refold quickly (blue curve). Tryptophan fluorescence was normalized to the highest value for the respective experiment to eliminate dilution effects from adding unfolded SOD1 to the refolding buffer (PBS). Note, SOD1 was unfolded overnight (12hrs) prior to in vitro SEDI binding experiments (Fig. 1).
166
Supplementary Figure 2. Western blots were performed to further demonstrate the specificity of the SEDI antibody. a) The SEDI antibody reacts only with monomeric SOD1 and not with dimeric SOD1 even when present at comparable concentrations. SOD1 is sufficiently stable that under denaturing conditions, it runs predominantly as the monomer, with some dimer still detectable. Since the samples were boiled in SDS, the SOD1 running at the dimer position may have partially unfolded but the SEDI epitope remains sequestered in the dimer interface. Several different concentrations of SOD1 were examined to obtain comparable levels of dimeric and monomeric SOD1 in these blots (10mg, 5mg, 1mg, 500ng, and 100ng). The intensity of the upper dimer band (as probed with anti-SOD1, right) in lane 1 (10mg of SOD1) is between the intensity of the monomer band of lane 4 and lane 5 (500-100ng SOD1). Identical blots were probed with either SEDI (left) or a commercial anti-SOD1 antibody (right, SOD100, StressGen). Note: the SEDI Western blot was deliberately overexposed to reveal any possible reactivity with the dimer. We performed denaturing Western blots because a direct comparison of SEDI reactivity between native dimeric SOD1 and mis/unfolded SOD1 is complicated by the fact that the native protein does not transfer to hydrophobic membranes well(20). b) Antibody specificity was confirmed by competition with the antigenic peptide. SEDI was pre-incubated with 500x molar excess of linear peptide with the same sequence as the antigenic peptide. Western blotting of 1mg of SDS-denatured SOD1 was then performed with either SEDI (left) or SEDI pre-incubated with peptide (right). SEDI reactivity is almost completely ablated by the addition of competing peptide.
167
Monomer/misfolded SOD1 in ALS-mouse models
Immunoprecipitation experiments with the SEDI antibody were conducted on
spinal cord tissue from G93A and G37R-SOD1 (enzymatically active mutants(9)) and
G85R-SOD1 (enzymatically inactive mutant(9)) mouse models of SOD1-ALS to test for
the presence of monomer/misfolded SOD1. These mice develop symptoms and pathology
resembling human ALS and are useful disease models(17-19). Since the SEDI antibody
is directed against a sequence identical in both mouse and human SOD1, it has the
potential to recognize endogenous mouse SOD1 as well as the transgenic human SOD1.
However, under non-denaturing conditions, the SEDI antibody detected
monomer/misfolded SOD1 from the soluble fraction of spinal cord homogenates of
mutant SOD1 mice (Fig. 1e). Non-specific binding was minimal, since only trace levels
of SOD1 were detected with pre-immune IgG (Supplementary Fig. 3)(20). We compared
the total SOD1 content of spinal cord homogenates with the amount immunoprecipitated
with SEDI to estimate the proportion of monomer/misfolded SOD1 (Fig. 1f). Less than
0.5% of total SOD1 was immunoprecipitated from the spinal cord homogenates from
each late-presymptomatic mouse (80 day old G93A SOD1, 8.5 month old G37R SOD1,
or 11-month old G85R SOD1 ALS-mice). This is the lower limit of total misfolded
SOD1 since only soluble misfolded species are detected in this experiment. We suspected
that large misfolded SOD1 aggregates might be present in the spinal cord insoluble pellet
fraction obtained after homogenization. While we did examine resuspended pellet
fractions for aggregated SOD1, control experiments using pre-immune IgG indicated that
SOD1 aggregates in the pellet fraction exhibit significant non-specific binding
(Supplementary Fig. 3). It is evident, however, that only a subset of the mutant protein is
168
Supplementary Figure 3. Comparison of immunoprecipitation reactions using SEDI antibody or pre-immune IgG. a) Immunoprecipitation reactions on 750�g of G93A SOD1 mouse spinal cord was carried out with either SEDI or pre-immune IgG. Binding to monomer/misfolded SOD1 in the soluble fraction following homogenization of tissue was almost completely specific (only trace SOD1 in the pre-immune IgG fraction, lane 5); however, significant non-specific binding was observed in the immunoprecipitation reactions of resuspended pellet fractions (lane 4). b) Immunoprecipitation of solubilized mitochondrial fraction from G85R SOD1 spinal cords. The spinal cords from two 5 month old G85R SOD1 mice (and one non-transgenic littermate) were pooled and fractionated (‘5 month’). Three 10 month old G85R SOD1 mice spinal cords were similarly pooled and fractionated (‘10 month’). The mitochondrial fractions were solubilized in RIPA buffer and immunoprecipitated with SEDI and pre-immune IgG. The loading controls, 0.5% of the immunoprecipitation supernatant (‘IP sup’), are shown in the first four lanes, followed by the immunoprecipitations (‘IP’). SEDI immunoprecipitates monomer/misfolded SOD1 from the 10 month sample (lane 8), but not the 5 month sample (lane 7). Pre-immune IgG pulls out a trace amount of SOD1 from the 10 month sample (lane 6). The greater amount of SOD1 in the SEDI IP relative to the pre-immune IgGfurther demonstrates the specificity of this antibody for monomer/misfolded SOD1. c, d)Bargraphs are densitometry results of this experiment using either maximum intensity (peak height, c) or integrated band intensity (area, d) measurements using Image J (NIH).
169
misfolded in these mice. The G127X-SOD1 ALS model expresses a natively unfolded
truncated protein that also accumulates to very low levels, but is sufficient to cause
disease(21).
Selective deposition of monomer/misfolded SOD1
To determine whether monomer/misfolded SOD1 is distributed uniformly
throughout the spinal cord, where only motor neurons are vulnerable to this stress, or
localized within specific cell types, we examined paraffin-embedded spinal cord sections
of transgenic SOD1 animals via immunohistochemistry. The SEDI antibody selectively
labels vacuolated structures localized within motor neurons of the ventral horn and
highlights the ventral root path in G37R-SOD1 (8.5 months old) mice (Fig. 2a). At higher
magnification, the localization of monomer/misfolded SOD1 specifically to motor axons
in the G37R-SOD1 mice is clearly evident (Fig. 2b), as well as in axonal processes of the
G85R-SOD1 mouse (Supplementary Fig. 4). Furthermore, inclusion bodies in cells that
morphologically appear to be motor neurons or astrocytes are detected with the SEDI
antibody in late stage G85R- (Fig. 2c, d) and G93A-SOD1 spinal cord (Fig. 2e, inset). No
labeling was observed in the dorsal horn, in axonal processes of the dorsal root (Fig. 2a),
nor in spinal cord sections of non-transgenic littermates (Fig. 2f). It should be reiterated
that SOD1 is expressed in all cells of the spinal cord, as demonstrated by the global
distribution of SOD1 staining using a commercial antibody that recognizes both folded
and misfolded SOD1 (Fig. 2h). This contrasts with the remarkable specificity of SEDI
antibody labeling, which reveals monomer/misfolded SOD1 predominantly in motor
neurons (Fig. 2a-b). The majority of monomer/misfolded SOD1 is concentrated along the
170
Figure 2. Misfolded SOD1 deposits predominantly on the periphery of vacuoles and aggregates in motor neurons of ALS-mice. In sections a-g) cross sections of mouse spinal cord are shown labeled with SEDI antibody, brown, and counter-stained with hematoxylin in all sections(blue); a) Labelling in G37R SOD1 mouse spinal cord ventral horn follows neuritic tracts (arrows). b) Monomer/misfolded SOD1 present in vacuoles in the axons of motor neurons of G37R-SOD1 (ventral root). c) and d) SEDI labeling in the G85R mouse appears primarily as aggregates in astrocytes (arrows, in d)), motor neurons (arrow, in c), but also as diffuse staining of motor neurons (double arrow, in d)). e) Primarily vacuolar labeling in G93A-SOD1 mouse spinal cord, (arrow); inset Inclusion bodies, similar to those reported earlier(1), also contain misfolded SOD1 (G93A-SOD1 mouse) f) SEDI antibody staining of spinal cord ventral horn from non-transgenic mouse littermate of the G93A-SOD1 mouse, note: absence of staining. g) SEDI antibody staining of spinal cord ventral horn from transgenic mouse overexpressing human wild-type SOD1, note: limited staining of ventral horn. h) SOD1 is ubiquitous and found in every cell type (brown); labeled with non-discriminating commercial SOD1 antibody (Biodesign; G93A-SOD1 mouse). Scale bars: a) 200 mm, b-d) 10 mm e) 20 mm f-h) 50 mm
171
Rishi
Note
Accepted set by Rishi
Supplementary Figure 4. SEDI labeling of a mutant SOD1-transgenic mouse spinal cords. a) G85R SOD1 mouse spinal cord can appear as diffuse staining following an axonal process (arrows, brown). b) The major site of SEDI antibody staining is around vacuoles in ventral horn of G93A mouse spinal cord. Arrows: motor neurons containing large numbers of vacuoles with abundant SEDI antibody staining. c) While the dominant structures labeled with SEDI antibody are vacuoles, labeling can also appear as diffuse staining of motor neuron perikaryon (arrow). Scale bar = 20µm
173
periphery of intracellular vacuoles in the spinal cords of 8.5 month old G37R- (Fig. 2b,
4e) and 100 day old G93A-SOD1 transgenic mice (Fig. 2e, Supplementary Fig. 4).
Misfolded SOD1 also appears as diffuse deposits within motor neuron perikarya
(Supplementary Fig. 4), illustrating that the SEDI antibody reveals monomer/misfolded
SOD1 that is distinct from inclusion bodies and vacuolar deposits. These new features are
unobservable with traditional antibodies, and perhaps these species may participate in
misfolded SOD1 toxicity.
Subcellular localization of monomer/misfolded SOD1
We investigated the intracellular distribution of SEDI-reactive SOD1 using
immunoprecipitations from subcellular fractions(5). SOD1 is classically defined as a
cytoplasmic protein and is mitochondrially localized in certain instances. Disease
affected (spinal cord, brain) and non-affected (liver) tissues from G93A- or (human)
wild-type SOD1 overexpressing rats or G85R-SOD1 mice were collected. Since large
molecular weight protein complexes containing SOD1 are present in terminal, but not
pre-symptomatic SOD1-G85R mice(22), our studies were performed using pre-
symptomatic animals. Subcellular fractionation of the tissue produced mitochondrial,
microsomal (P100) and cytosolic fractions, each of which was solubilized with mild
detergent and its protein concentration normalized prior to immunoprecipitations. In
animals overexpressing enzymatically active G93A-SOD1 and wild-type SOD1,
monomer/misfolded SOD1 is present in both the mitochondrial and cytosolic spinal cord
fractions, with only very small amounts detectable in the microsomal fraction (P100)
(Fig. 3). In contrast, only minor amounts were immunoprecipitated from corresponding
fractions isolated from liver and brain collected from the same G93A SOD1 rat and
174
Figure 3. Subcellular distribution of monomer/misfolded SOD1. Affected (spinal cord, brain) tissues and unaffected tissues (liver) were dissected from G93A SOD1 rats, wild-type SOD1 overexpressing rats or from G85R SOD1 mice. Tissues were fractionated as per Liu et al(5) and anti-SOD1 Western blotting was performed as indicated in Materials and Methods. Loading controls (‘load’) were 0.25% of the immunoprecipitation reaction supernatant. 100mg of protein from each sample was SEDI immunoprecipitated as indicated (‘mito’= gradient purified mitochondria, ‘P100’= 100,000 xg pellet, containing microsomes, ‘cyto’ = cytoplasm), except the G85R-SOD1 spinal cord mitochondrial fraction, 30mg of which was used.
175
nearly undetectable from these tissues of the wild-type SOD1 overexpressing rat.
Conversely, monomer/misfolded G85R-SOD1 was enriched in mitochondrial and
microsomal fractions from spinal cord and brain, while lesser amounts were recovered
from the cytosol. The G85R-SOD1 mice express the protein at much lower levels than
the G37R- and G93A-SOD1 mice, but it has considerably more monomer/misfolded
SOD1 in the density fractions containing mitochondria from the spinal cord and brain.
Immunoprecipitation reactions on the G85R SOD1 spinal cord mitochondrially enriched
fraction pulled out substantial amounts with the SEDI antibody, but only a trace amount
with pre-immune IgG, demonstrating that the SEDI antibody surpasses any non-specific
binding effect (Supplementary Fig. 3). Because our analyses are with tissues from pre-
symptomatic G85R-SOD1 animals (collected prior to the appearance of large protein
aggregates) and because any such protein aggregates would have been removed by
sedimentation to higher density than that of mitochondria, it is unlikely that the
monomer/misfolded SOD1-G85R we identify results from incomplete separation of
putative, non-mitochondrial aggregates. However, since variable amounts of
monomer/misfolded SOD1 is immunoprecipitated from mitochondrial fractions of the
G85R-SOD1 spinal cord of different ages (Fig. 3, Supplementary Fig 5), and the
possibility of co-precipitation can not be completely excluded, a more reliable estimate of
the amount of monomer/misfolded SOD1 association with G85R-SOD1 spinal cord
mitochondria needs to be measured in the future. Monomer/misfolded SOD1 is also
detected in these same fractions isolated from G93A spinal cords that do not have
significant levels of such protein aggregates(22). Association with membrane-containing
fractions (Fig. 3 and Supplementary Fig. 6) is consistent with recent findings of ER-stress
176
Supplementary Figure 5. SEDI antibody immunohistological labeling is specific and can be saturated by competition with the antigenic peptide. a) G37R-SOD1 mouse spinal cord ventral root labeled with SEDI antibody. b) Same as a), but antibody reaction competed with excess antigenic peptide.
177
Supplementary Figure 6. Our immunohistochemistry experiments show that monomer/misfolded SOD1 accumulates around vacuoles (Fig. 3b). These vacuoles are histopathological features observed in the G93A- and G37R-SOD1 mice, but not other mutant SOD1 mice(1). It has been proposed that these vacuoles have mitochondrial(18) and peroxisomal origins(30); in fact, we also observe localization of mitochondrial markers, such as Tom20, superimposed with, or adjacent to, SEDI labeling around these vacuoles (a-c). While mitochondrial association of G93A or G37R-SOD1 is implicated in vacuolization(30), the significant enrichment of monomer/misfolded SOD1 in the mitochondrial fraction of G85R-SOD1 mice, which do not possess vacuoles, suggests the apparent association of monomer/misfolded forms of SOD1 to mitochondria is not sufficient to cause vacuolization. a) Monomer/misfolded SOD1 primarily localizes around vacuoles (brown); counterstained with hematoxylin (blue). b) Vacuoles labeled with marker for mitochondrial outer membrane, TOM-20 (blue). Note: not counterstained with hematoxylin. c) Co-localization of monomer/misfolded SOD1 (brown) and TOM-20 (blue) (arrows).
178
and activation of the unfolded protein response in mutant SOD1-mice(23, 24). Native
dimeric SOD1 is primarily cytosolic, as expected and no monomer/misfolded SOD1 was
detected in any fraction of liver tissue. Since mitochondria are the principle source of
oxidative stress in vivo, the observation that monomer/misfolded SOD1 deposits are
concentrated in fractions enriched in mitochondria, especially relative to the distribution
of total SOD1, is consistent with the hypothesis that oxidative stress triggers SOD1
misfolding(12).
Wild-type SOD1 can misfold in vivo
It has been suggested from in vitro studies that a pool of monomeric immature SOD1
exists in cells(25); the absence of staining in the non-transgenic mouse (Fig. 1e, 2f) NTD:
check this suggests that either monomeric immature SOD1 is present at levels below the
detection limit of this antibody or that this antibody does not recognize immature SOD1
in this animal. Interestingly, small amounts of misfolded SOD1 were also observed in
atypical vacuoles in the mouse expressing high levels of human wild-type SOD1 (Fig.
2g). This is corroborated by our observation that small amounts of SOD1 are
immunoprecipitated using the SEDI antibody from this mouse (Fig. 1e) and that wild-
type SOD1 can misfold in vitro(13). It has been reported that transgenic mice
overexpressing wild-type human SOD1 do develop pathological features of ALS, but
when much older than mice expressing the mutant protein(26). Our finding of misfolded
SOD1 localized to vacuoles in the wild-type SOD1 mouse is consistent with these
previous findings. Indeed, the presence of some misfolded SOD1 in the wild-type SOD1
mouse may explain a long-standing mystery of how high-level overexpression of human
wild-type SOD1 can actually exacerbate disease by causing earlier onset in mice(26, 27).
179
Monomer/misfolded SOD1 appears prior to symptom onset and correlates with motor
neuron loss
Understanding the production and accumulation of misfolded SOD1 in the
context of disease progression is critical to the timing of potential therapeutics and may
yield clues about the source of SOD1 denaturational stress. In our colony, the G93A-
SOD1 disease phenotype is highly synchronous with progressive hindlimb weakness
developing around 100 days of age. The mice are no longer viable at 120-130 days of
age. We examined mice at 20, 63, 100 and 120 days of age for the presence of misfolded
SOD1. Monomer/misfolded SOD1 was initially absent (20 days; Fig. 4a), but could be
detected in presymptomatic mice by 63 days of age (Fig. 4b). Misfolded SOD1 was also
present at the onset of hindlimb weakness (100 days; Fig. 4c), but declined with a
concomitant loss of motor neurons at end stage (120 days; Fig. 4d). There are also
reduced levels of labeling with the SEDI antibody at end stage in the G37R-SOD1
mouse, where the majority of motor neurons have degenerated (Figure. 4e-f). The
apparent decrease in SEDI-labeling at disease end stage is attributed to neuron loss, the
primary site where misfolded SOD1 is seen. These observations demonstrate a temporal
linkage of SOD1 misfolding and motor neuron degeneration in these mice:
monomer/misfolded SOD1 accumulation is correlated to disease onset and motor neuron
degeneration, and disappears concomitant with motor neuron loss.
Monomer/misfolded SOD1 in a human case of A4V SOD1-ALS
Spinal cord sections obtained at autopsy from a SOD1-ALS patient carrying the
A4V SOD1 mutation were also examined with the SEDI antibody to test for the presence
180
Figure 4. Age-dependent accumulation and decrease of misfolded SOD1 with concomitant loss of motor neurons in mouse models of ALS. (a-d, G93A SOD1 mouse; e-f, G37R SOD1 mouse) a) Monomer/misfolded SOD1 is initially absent from the mouse spinal cord (age: 20 days). b) Monomer/misfolded SOD1 appears in pre-symptomatic G93A-SOD1 mouse spinal cord (63 days). c) Monomer/misfolded SOD1 staining peaks at onset of rear-leg weakness (100 days). d) Decline in levels of monomer/misfolded SOD1 at end-stage. Some monomer/misfolded SOD1 is still present as round deposits (arrows), but obvious vacuolar deposition is minimal (120 days). e) Monomer/misfolded SOD1 in G37R-SOD1 model of ALS at 8.5 months of age (pre-symptomatic) and f) 12 months of age (disease end-stage). Scale bar in a) 50 mm, b-f) 25 mm. SEDI antibody staining is in brown, hematoxylin counterstain is in blue.
181
of monomer/misfolded SOD1. SEDI antibody labeling in the human case paralleled our
observations of labeling in the end-stage ALS-mouse spinal cords. Both human and
mouse spinal cords had reduced SEDI antibody staining and very few healthy motor
neurons at disease end-stage; however, clear motor neuron labeling was seen in the
human spinal cord with the SEDI antibody (Fig. 5a). We also found numerous deposits of
monomer/misfolded SOD1 (Fig. 5b) that are morphologically similar to those found at
the disease end stage in mouse models of ALS (Fig. 4d, f). This is in contrast with
ubiquitous labeling of an adjacent section with a commercial SOD1 antibody (Fig. 5c).
This confirms the presence of monomer/misfolded SOD1 detected with the SEDI
antibody in a human case of ALS.
Discussion
The SEDI antibody, a direct in vivo probe of SOD1 conformation, has established the
presence of monomer/misfolded SOD1 in a human A4V SOD1-ALS case, a rat model of
ALS and mice with highly expressed, enzymatically active G93A- and G37R-SOD1 as
well as lower expressed, enzymatically inactive and physiologically unstable G85R-
SOD1. We found that monomer/misfolded SOD1 accumulates in motor neurons prior to
neurodegeneration, whereas natively folded wild-type and mutant SOD1 is ubiquitous
and found in all cell types. This observation, coupled with demonstration that expression
of mutant SOD1 in motor neurons drives initiation of disease and early progression in
transgenic mice(28), makes a case for the neuronal toxicity of misfolded SOD1.
Furthermore, while monomer/misfolded SOD1 comprises only a small fraction of the
total mutant SOD1 population, it is enriched in membrane-containing subcellular
fractions, especially those containing mitochondria, thus implying some significance for
182
Figure 5. Monomer/misfolded SOD1 in a case of human SOD1-ALS. a) A motor neuron labeled with SEDI antibody in ventral horn of human spinal cord from ALS case carrying the A4V-SOD1 mutation.a) Misfolded SOD1 present as small round deposits in human A4V-SOD1 ALS parallels the round deposits observed in end-stage ALS-mouse models (Fig. 4d, f). SEDI antibody staining is in brown, hematoxylin counterstain is in blue.c) Labelling of an adjacent section of human case of A4V SOD1-ALS with a with commercial SOD1 antibody (Biodesign). In contrast with limited, specific labelling with the SEDI antibody, the commerical antibody reveals the ubiquitous presence of SOD1 (brown). Scale bar=10mm (Figure 5a, b) 20mm (Figure 5c)
183
misfolded protein deposition and possible disruption of normal mitochondrial function in
SOD1-ALS toxicity. The SEDI antibody recognizes a large stretch of hydrophobic
residues normally buried in the native dimer interface. Consequently the accumulation of
monomer/misfolded SOD1 in membrane fractions may be mediated by the exposed
hydrophobic dimer interface. Furthermore, exposure of the dimer interface epitope
reveals sites of potential de novo protein-protein interactions, including the exposure of
two putative Hsp70 binding sites(29) in the SOD1 sequence (Supplementary Fig. 7).
Motor neurons may be selectively vulnerable because of their inability to efficiently
upregulate protein chaperones(3).
While further testing is needed to precisely correlate phenotypic disease progression and
monomer/misfolded SOD1 levels, the antibody may allow a method to follow disease
course using CSF or other samples during therapies. As such, it also has potential use in
diagnosis if monomer/misfolded SOD1 can be detected in CSF. The SEDI antibody may
have utility in drug discovery efforts aimed at identifying molecules that prevent SOD1
misfolding by stabilizing native SOD1. In addition, the SEDI antibody may have direct
therapeutic benefit whereby passive immunization blocks aberrant interactions with
misfolded SOD1. This new research tool should expand experimental possibilities within
ALS research and thus will be made suitably available.
184
Supplementary Figure 7 . Putative Hsp70 binding sites in SOD1 – 4-5 hydrophobic residues flanked by basic residues. a) SOD1 dimer with N-terminal binding site shown in red (residues 4-8), b) SOD1 dimer with C-terminal binding site shown in gray (residues 144-151). c) SOD1 monomer with N-terminal binding site shown in red. d) SOD1 monomer with C-terminal binding site shown in gray.
185
Acknowledgements
We thank M. Strong (University of Western Ontario) for providing human spinal cord
sections and P.H. Chan (Stanford University) for providing human wild-type SOD1 rats.
We also thank N. Ng and members of the Chakrabartty lab and CRND for helpful
comments and review, V. Mulligan, J. Kim and W. Zhou for technical assistance and C.
Accardi for assistance with animal care. Funding was from the Neuromuscular Research
Parternship- the Canadian Institutes of Health Research (CIHR), ALS Society (Canada)
and Muscular Dystrophy Association (Canada) (J.R., and A.C.), ALS Association (U.S.)
and MND Association (U.K.) (J.R.), CIHR, and Temerety Family Trust (N.R.C). C.V. is
funded by the Muscular Dystrophy Association and D.W.C. receives salary support from
the Ludwig Institute. R.R. receives a Doctoral Research Award from CIHR Institute of
Neuroscience, Mental Health and Addication (CIHR-INMHA) and ALS Society
(Canada). J.R. holds a Canada Research Chair in Molecular Mechanisms of ALS and
N.R.C. holds a Canada Research Chair in Neurodegeneration and Protein Misfolding
Diseases. Requests for the SEDI antibody should be directed to
(27) Deng, H. X., Shi, Y., Furukawa, Y., Zhai, H., Fu, R., Liu, E., Gorrie, G. H., Khan,
M. S., Hung, W. Y., Bigio, E. H., Lukas, T., Dal Canto, M. C., O'Halloran, T. V., and
Siddique, T. (2006) Conversion to the amyotrophic lateral sclerosis phenotype is
associated with intermolecular linked insoluble aggregates of SOD1 in mitochondria.
Proc Natl Acad Sci U S A 103, 7142-7.
(28) Boillee, S., Yamanaka, K., Lobsiger, C. S., Copeland, N. G., Jenkins, N. A.,
Kassiotis, G., Kollias, G., and Cleveland, D. W. (2006) Onset and Progression in
Inherited ALS Determined by Motor Neurons and Microglia. Science 312, 1389-92.
(29) Rudiger, S., Germeroth, L., Schneider-Mergener, J., and Bukau, B. (1997)
Substrate specificity of the DnaK chaperone determined by screening cellulose-bound
peptide libraries. Embo J 16, 1501-7.
(30) Higgins, C. M., Jung, C., and Xu, Z. (2003) ALS-associated mutant SOD1G93A
causes mitochondrial vacuolation by expansion of the intermembrane space and by
involvement of SOD1 aggregation and peroxisomes. BMC Neurosci 4, 16.
191
Discussion and Future Directions
Summary
Here I have developed an in vitro model system that faithfully recapitulates the ALS-
associated SOD1 misfolding pathway. I have shown that the SOD1 misfolding pathway
populates a monomeric intermediate in vitro and in vivo. In vivo, SOD1 misfolding was
initially absent, but was found prior to symptom onset. Additionally, misfolded SOD1
was localized to motor neurons, the primary pathology in ALS. These spatio-temporal
correlations make a strong case for a causative role of SOD1 misfolding in ALS
pathogenesis.
Discussion
Understanding the underlying causes of ALS is crucial to developing a therapy that treats
more than the downstream effects of motor neuron degeneration. In fact, because ALS is
diagnosed on neurological criteria alone(1, 2), it is better described as a syndrome, a
condition characterized by common pathology or symptoms, rather than a single disease
– there could be multiple causes of upper and lower motor neuron degeneration. A
number of different etiologies for ALS have been proposed. We investigated the
molecular mechanisms by which SOD1 misfolding might cause ALS. Other potential
causes include chemical/drug dependent motor neuron degeneration in atypical Guanian
ALS/Parkinsonism(3), which may be related to the consumption of cycad. A very small
number of familial cases of ALS are attributed to a proline to serine mutation at amino
acid 56 (P56S) in the vesicle associated membrane protein B (VAMP-B), but this is
associated with atypical ALS(4). The tar-DNA binding protein (TDP-43) was found to be
a component of inclusion bodies found in ALS pathological inclusion bodies(5, 6).
192
Interestingly, it has also recently been shown that TDP-43 mutations are found in a small
(<1-4%) proportion of ALS cases(7, 8). However, there was insufficient strength in the
genetic studies to show a causal relationship between TDP-43 mutations and ALS and it
was not found in genome-wide association studies(9); additionally, TDP-43 is found on
chromosome 1(10), whereas the remaining (unknown gene) ALS-associated loci are
found on different chromosomes(11). It remains that mutations in SOD1 are the only
known cause of typical adult-onset ALS.
SOD1 mutations were first shown to cause ALS in 1993(12). Prior genetic studies
had narrowed the linkage with ALS to chromosome 21q(13); SOD1 was sequenced as a
candidate gene within this chromosome because oxidative stress was thought to play a
critical factor in motor neuron degeneration and SOD1 is the principle scavenger of
cellular superoxide. A simple enzyme loss-of-function mutation was thought to underlie
these cases of ALS(12); however, a number of observations thereafter made it clear that
SOD1 mutations caused ALS by a new toxic gain-of-function. First, SOD1 mutations
cause ALS in an autosomally dominant inherited manner (reviewed in (11)); while this
could be attributed to haploinsufficiency, it is more likely that for a highly expressed
enzyme like SOD1, a single copy mutation would cause disease by a gain of function. In
vitro, mutant SOD1s associated with ALS can have similar enzymatic activity to that of
the wild-type protein(14). Lastly, in mouse genetic studies, expression of human mutant
SOD1 produced an ALS-like phenotype (motor neuron wasting and paralysis) despite the
presence of endogenous mouse wild-type SOD1(15, 16), whereas genetically removing
the mouse SOD1 did not produce any motor neuron phenotype(17). Over-expression of
193
human wild-type SOD1 with human mutant SOD1 failed to rescue the motor neuron
phenotype in mouse models(18).
Two divergent observations attempted to explain the SOD1 toxic gain-of-
function. First, Crow et al observed that a common property of the many missense SOD1
mutants is their relatively weak zinc binding affinity(19). They and others showed that
mutant SOD1 can behave as a net pro-oxidant under certain conditions(20). Kopito and
coworkers showed that mutant SOD1 can form structures called aggresomes when
transiently expressed under conditions of protein folding stress(21). These are
reminiscent of inclusion bodies found in human ALS and mouse models of ALS(21).
Aggresomes were then showed to inhibit proteosomes(22, 23). Each of these studies has
their relative strengths and weaknesses. First, the in vivo zinc binding capacity of mutant
SOD1s relative to wild-type has been difficult to quantify(24). Large increases, relative to
age-matched controls, in oxidative markers are not observed in ALS(25, 26). Also,
aggresomes formed by SOD1 in vitro have some differences than those found in the
disease(27). Furthermore, proteasome inhibition is a downstream effect of protein
misfolding, the molecular details of which are still not understood.
As expounded in previous chapters, we sought to uncover (one of) the
mechanism(s) by which mutant SOD1s cause motor neuron degeneration in ALS.
Because SOD1 is extremely stable – a Tm of ~100°C(28) and active in 8M urea(29) – and
is only marginally destabilized by mutations associated with ALS(30), we proposed that
there exists a physiological stress that causes it to denature/misfold(31). This is in
contrast with the Aβ peptide and tau in Alzheimer’s disease(32) and alpha-synuclein in
Parkinson’s disease(33), which each form aggregates, but each is natively unfolded and
194
so does not have to first unfold prior to forming aggregates. While SOD1 might not
substantially modify other proteins/intracellular targets, it might very well become
oxidatively modified itself as a consequence of its normal enzymatic activity. This could
be local reactive oxygen species (ROS) generated by SOD1, mitochondrial or cellular
ROS. Even if SOD1 does not locally generate ROS, it might still be modified because of
its high concentration in the central nervous system (CNS) – even if there is no selectivity
for SOD1 oxidation, it may occur predominantly because of its mere presence. In
addition, the life-time of SOD1 molecules in motor neurons may span several years
because of its transport in long axons by slow-component B increases the probability of
SOD1 oxidation(31). We used metal-catalyzed oxidation, similar to ROS potentially
generated by SOD1 itself and other intracellular sources, to show that this physiologically
relevant stress can cause SOD1 to misfold and aggregate. This provided strong evidence
that physiological stresses like oxidation can cause extremely stable proteins like SOD1
to misfold, something that otherwise only occurs under strongly denaturing conditions in
vitro and a plausible explanation for how such a stable protein aggregates in vivo.
Oxidative stress could thus be an important factor in ALS. A protein-lifetime’s
accumulations of oxidative modifications could also explain why ALS is a late-onset
disease.
Oxidation induced SOD1 misfolding was related to covalent modification of
histidine residues, including those normally involved in binding metals. Because binding
zinc is thought to stabilize SOD1, oxidation of zinc-binding residues could result in a loss
of metal binding affinity and cause the protein to misfold. In our experiments, zinc-
deficient wild-type SOD1 was the most aggregation prone of all the SOD1 species tested
195
and the aggregation propensity of a particular SOD1 species was proportional to how
much ANS it bound prior to oxidation. That is, partial unfolding (population of an ANS
binding state) predisposed SOD1 to aggregation. Oxidation of holo-SOD1 also produces
misfolded SOD1 aggregates, albeit at higher concentrations than mutant SOD1s, which
are deficient in zinc, providing further evidence that zinc release is associated with SOD1
misfolding. Because even wild-type SOD1 can misfold and aggregate, there is a finite
probability of anyone developing ALS, which is simply increased by mutations in SOD1.
This provides a plausible mechanism by which aggregation of wild-type SOD1 might
cause sporadic ALS and a model where misfolding of SOD1 underlies both mutant SOD1
and non-mutant SOD1 ALS cases.
Analysis of in vitro SOD1 aggregation kinetics revealed a monomeric misfolding
intermediate from an otherwise obligate homodimer(28). Residual monomeric SOD1 was
also observed in the soluble fraction following oxidation induced aggregation of both
wild-type and mutant SOD1. The native SOD1 homodimer is held together by burying a
large hydrophobic surface(34); monomerization immediately suggests a mechanism for
both aggregation and cellular toxicity. If specific monomer-monomer contacts are lost,
non-specific aggregation might occur to minimize exposed hydrophobic surface area.
Toxicity might arise from binding to other proteins and altering signaling, inactivating
essential cellular components or overwhelming the protein-folding homeostasis
machinery.
Our in vitro SOD1 aggregation system recapitulated several features also
observed in vivo. Most other protein misfolding diseases (e.g. transthyretein in senile
systemic amyloidosis, Aβ in Alzheimer’s disease) are characterized by the deposition of
196
amyloid: conversion of the soluble protein into an insoluble form that is high in β-sheet
content, contains unbranched fibrils ~7-10nm in diameter and specifically binds to the
dyes Thioflavin T and Congo red(35, 36). Oxidation induced SOD1 aggregates have none
of these properties; additionally, the ultrastructure of SOD1 aggregates resembles that of
SOD1 aggregates found in motor neurons of pathological SOD1-ALS cases. Our in vitro
SOD1 aggregation system has two major predictions about misfolding of SOD1 in vivo.
First, the monomeric aggregation intermediate that we observed in vitro should also be
present in vivo. Aggregation is a disease-associated process and the monomeric
intermediate should only be found in pathological specimens, and if the gain-of-function
mutation causes toxicity, the monomeric species should be initially absent, but found
prior to symptom onset. Folding of SOD1 may also proceed through a monomeric
intermediate. If a population of immature monomeric SOD1 exists in cells, it may have a
different structure than the monomeric aggregation intermediate, or, alternatively,
monomeric SOD1 may have a higher population in pathological cases than in non-
pathological ones. Secondly, we purport oxidative stress is the denaturational stress
underlying SOD1 misfolding and aggregation. If our model holds true, misfolded SOD1
should also be oxidized in vivo. Because oxidative modifications to SOD1 are partially a
result of its long lifetime in vivo, a corollary of the oxidative hypothesis is also that
misfolded SOD1 should be initially absent.
Careful analysis of the high-resolution structure of SOD1 revealed that a β-strand
(amino residues 146-151) is sequestered in the dimer interface of the SOD1 homodimer
and is solvent inaccessible. Upon monomerization, however, these residues should be
exposed to solvent. We created a peptide antibody (SEDI) to these residues reasoning that
197
it should not react with the native dimer (because these residues are solvent inaccessible),
but that it should react with monomeric intermediate, and any other forms where the
native dimer is disrupted (monomer/misfolded SOD1)(37). Because native SOD1 is an
obligate homodimer, any form where these residues are solvent exposed correspond to
misfolded SOD1- where any residual immature monomeric SOD1 can be considered to
be misfolded because it has failed to fold to the native dimer. We confirmed that the
SEDI antibody has the predicted SOD1 monomer selectivity. SEDI reacted with unfolded
SOD1 in immunoprecipitation reactions, Western blots and ELISA, but not with folded,
dimeric SOD1. This was, to the best of my understanding, the first example of an
antibody designed from the high-resolution structure of a protein to achieve selectivity
for certain folding states of a protein and not other folds.
We first tested whether monomer/misfolded SOD1 was found in mouse models of
SOD1. Because the SEDI epitope is identical in mouse SOD1 and human SOD1, SEDI
would recognize monomer/misfolded mouse SOD1 or immature endogenous monomeric
SOD1 if it were present. However, no reactivity was found in non-transgenic mice.
Monomer/misfolded SOD1 was found in several different mouse lines, each expressing a
different mutant SOD1. The full human mutant gene, inclusive of introns, is expressed at
high copy number, producing variable levels of mutant SOD1 mRNA and protein(15,
38). Lines expressing high levels of human protein develop ALS-like symptoms. Only a
small fraction (0.1%) of total mutant SOD1 was immunoprecipitated using SEDI;
endogenous mouse SOD1, which has slightly different mobility than human SOD1 in
SDS-PAGE, was not immunoprecipitated. Most human SOD1 is properly folded when
expressed in mouse; this suggests a reason why high expression is required to produce the
198
motor phenotype in these mice, where the misfolded species must reach some threshold
concentration. We found that mice expressing human wild-type SOD1 also had some
monomer/misfolded SOD1, but less than in the G93A SOD1 mouse; this mouse does
exhibit a very slight phenotype(18). Consistently, when individually expressed human
wild-type or A4V mutant SOD1 are non-symptomatic, but expressed an ALS-phenotype
is produced in double transgenic mice(39). This may be caused by reaching a threshold of
misfolded SOD1 in these animals.
If monomer/misfolded SOD1 is a causative agent in ALS, it should be present
predominantly in motor neurons, or, motor neurons could be selectively vulnerable to this
protein misfolding stress. In immunohistochemical experiments, SEDI reactivity was
found only in the anterior horn of the spinal cord – the location of degenerating motor
neurons – and in the motor axons of the ventral root itself. It is unclear from our
experiments whether monomer/misfolded SOD1 is generated and degraded in all cells
except motor neurons or generated selectively in motor neurons. The accumulation of
misfolded SOD1 in motor neurons may signal an insufficiency in cellular clearance
mechanisms, the requirement of some chemical modification over time of SOD1 that
triggers misfolding, or possibly a combination thereof. The observed selectivity of
monomer/misfolded SOD1 deposition implies that there may be convergence of these
risk factors. Also, our observation that monomer/misfolded SOD1 is found primarily in
motor neurons does not eliminate the possibility that motor neurons are also selectively
vulnerable to this stress. Not all motor neurons/axons exhibited SEDI reactivity;
however, not all motor neurons express mutant SOD1 to high levels in these mice(15,
38). It is unclear from our study whether motor neurons lacking monomer/misfolded
199
SOD1 expressed mutant SOD1 and were spared/pre-disease or simply did not express the
mutant protein. A more comprehensive examination correlating loss of function in a
particular motor unit (e.g. strength) to the load of misfolded SOD1 could be carried out to
dissect this relationship. For example, it is well known from physiological and anatomical
studies as to which motor neurons innervate which muscles (e.g. (40). Transgenic mice
would be sacrificed at different post-symptomatic stages to examine possible correlation
between SEDI load in the innervating motor neurons and muscle function.
Subcellular localization of monomer/misfolded SOD1 might provide some clues
on the source of denaturational stress and/or mode of misfolded protein toxicity.
Although SOD1 is classically a cytosolic protein, a small fraction has previously been
found in the mitochondrial intermembrane space and in the extracellular milieu. We
performed subcellular fractionation of several tissues from several different transgenic
SOD1 animals. Solubilized mitochondria, endosomes and cytoplasmic fractions were
immunoprecipiated using the SEDI antibody. Monomer/misfolded SOD1 was
concentrated in the cytoplasmic fractions of the spinal cord and brain, as expected,
however, significant amounts were also found in the mitochondrial fractions of these
tissues. Recent work suggests that mitochondria may be the proximal site of mutant
SOD1 toxicity; we have shown that a subset of the normal SOD1 associated with the
mitochondria becomes misfolded in transgenic mice and rats. Because mitochondria are
the principle source of reactive oxygen species intracellularly(41), our observation that
monomer/misfolded SOD1 accumulates in the mitochondria is consistent with our
original hypothesis that oxidation causes SOD1 misfolding(31).
Implications/Predictions from Thesis Work
200
During the course of my thesis work, we discovered a monomeric SOD1 misfolding
intermediate and associated it with pathology in transgenic animals expressing human
mutant SOD1 and a human case of ALS caused by a SOD1 mutation. Because mutations
in SOD1 are known to cause the disease and monomer/misfolded SOD1 is found only in
pathologically relevant tissues, a strong case is made for role of this misfolding
intermediate in SOD1 mediated toxicity. Direct demonstration of misfolded SOD1
toxicity has not been made. Experiments where misfolded SOD1 is added directly to
cells, or perhaps directly into the brain or spinal cord of mice, might address this
problem. In cell culture, cells treated with misfolded SOD1 (for example, oxidized
SOD1) should have lower viability than cells treated with native SOD1 or buffer. Non-
transgenic mice treated with misfolded SOD1 might develop local motor neuron
pathology, whereas treatment with transgenic mice expressing mutant SOD1 might
trigger the disease and result in earlier onset in addition to local pathology. Technical
limitations of getting sufficient misfolded SOD1 protein into the relevant cells, however,
are significant. While it has long been known that severity of motor phenotype in mutant
SOD1 animals correlates strongly with gene dosage(42), correlation with misfolded
SOD1 load has not been determined. This experiment, however, may be complicated by
timing of the assay for misfolded SOD1, since the amount of misfolded SOD1 varies
widely with the disease course. I predict that a relatively constant amount of misfolded
SOD1 may be required to trigger the disease. This hypothesis could be tested by
examination of the misfolded SOD1 load in presymptomatic transgenic mice expressing
mutant SOD1. Because these mice are very well characterized and the timing of the
phenotype is well known, spinal cords could be dissected from mice immediately prior to
201
symptom onset from mice expressing varying levels of mutant SOD1. Reducing the load
of misfolded SOD1 should also delay symptom onset and lengthen disease duration. Liu
and coworkers recently demonstrated that immunization of several different mouse
models of ALS with the SEDI antigen or oxidized SOD1 aggregates caused a reduction
in misfolded SOD1 load and a delay in symptom onset and increase in disease duration
(43). These observations, coupled with our association of misfolded SOD1 with motor
neuron pathology, make a strong case for monomer/misfolded SOD1 as a toxic moiety in
ALS.
I have shown that our in vitro model system of oxidation mediated SOD1
aggregation successfully recapitulates several features of SOD1 mediated ALS in vivo.
However, several hypotheses stemming from our in vitro system and our in vivo findings
have yet to be examined in detail. First, oxidation of key metal binding residues in SOD1
causes aggregation in vitro, but we have not shown that oxidation of SOD1 causes
misfolding and aggregation in vivo. It is difficult to test this hypothesis directly because
oxidative stress can not be targeted to SOD1 in an animal. Several corollaries of the
oxidation hypothesis can be tested. First, misfolded SOD1 isolated from animal models
and/or human pathological specimens should be oxidatively modified. Oxidative
modifications to SOD1 have been observed, but it is unclear whether this segregates with
misfolded SOD1. Misfolded SOD1 could simply accumulate oxidative modifications
after misfolding, but observation that it is oxidized would support the oxidative
hypothesis. A challenge here is to isolate adequate amounts of misfolded SOD1 to
examine its oxidation status using conventional tryptic digestion and mass spectrometry
sequencing. Protein identification requires the identification of only one to several
202
peptides in the LC-MS/MS experiment; however, to identify post-translational
modifications, peptides that cover the entire protein sequence must be obtained.
Additionally, the modified peptides must make up a significant fraction of the total, or
modified peptides will not be detected(44). Typically, a hundred to a thousand fold more
protein is required to examine post-translational modifications exhaustively. Because
monomer/misfolded SOD1 makes up only <0.1% of total SOD1, and is found primarily
in the spinal cord- which is small in mice- it is difficult to isolate adequate amounts of
misfolded SOD1 for this analysis. Experiments to this end, however, where we use spinal
cords from a transgenic rat model expressing mutant SOD1(45), that are large enough to
yield significant amounts of misfolded SOD1, are on going. In a more direct approach,
efforts are underway to test whether oxidative insults, such as hydrogen peroxide or
paraquat, can increase the amount of oxidized/misfolded SOD1 in cells expressing wild-
type or mutant SOD1s. Oxidation of non-SOD1 substrates may occur, but should not
result in an increase in oxidized/misfolded SOD1 unless there is an increase in non-
specific aggregation which can be controlled by quantitative measurement of proteasome
function or ANS binding.
Our in vitro model suggested that, in addition to mutant SOD1, wild-type SOD1
can also aggregate through a monomeric misfolding intermediate. From these results, we
predicted that everyone has a finite probability of developing ALS based on the
oxidation-induced aggregation of wild-type SOD1. Small amounts of
monomer/misfolded SOD1 was also present in transgenic mice and rats overexpressing
human wild-type SOD1. Misfolding of SOD1 may thus also underlie a subset of ALS
cases where there are no mutations in SOD1. In our examination of two other cases of
203
ALS where there is no SOD1 mutation, no monomer/misfolded SOD1 was found.
Histological or biochemical immunologic testing with the SEDI antibody of many more
ALS cases is needed to ascertain whether SOD1 misfolding can cause ALS independent
of mutation in the gene.
Because SOD1 is transported in motor axons by a protein complex called slow
component B(46), we calculated that the time it would take SOD1 translated in the
perikaryon to reach a distal motor end plate to be up to 500 days(31). This calculated
long life-time of SOD1 could potentiate oxidative modification of SOD1. We, however,
did not directly measure the life-time of SOD1 in humans or in mice. Confirmation of
this life-time, by, for example, 35S methionine pulse-chase experiments, and correlation
with oxidation modifications would help strengthen the oxidation induced SOD1
misfolding hypothesis. Alternatively, ligand controlled expression of epitope-tagged
SOD1 in mice could be used to measure the life-time of SOD1 in various tissues,
including motor neurons. In each case, labelled SOD1 would be detected in tissues
dissected from animals sacrificed at various times post-innoculation/translation.
Dectection might be via mass spectrometry, radioactive decay/emission or
immunohistochemistry. Simultaneous measurement of SOD1 oxidative modification by
mass spectrometry and/or antibodies selective for the oxidized form could then correlate
the propensity of SOD1 to become oxidized with its life-time. If the long life-time
potentiated oxidative misfolding of SOD1, misfolded SOD1 should be initially absent
and accumulate over time. Monomer/misfolded SOD1 was initially absent from G93A
SOD1 mice and was found prior to the onset of symptoms, in keeping with a causative
role of SOD1 misfolding in ALS etiology. Our observation is consistent with our
204
oxidation-induced SOD1 misfolding hypothesis, but could also be explained by the
accumulation of small amounts of misfolded SOD1 from protein that failed to fold
properly, unfolded stochastically, or misfolded from a non-oxidative denaturational
stress. Correlating protein life-time, oxidation of side chains and SOD1 misfolding in
animal models would greatly strengthen our assertion of a causative link between these
factors.
Mutant SOD1 was recently shown to associate with mitochondria in several
disease models of ALS. Expression of mutant SOD1 causes mitochondrial dysfunction in
cell culture(47) and radical alteration in the ultrastructure of a subset of brain and spinal
cord mitochondria in wild-type-, G93A-, or G37R- SOD1 mice. Because a subset of
SOD1 is normally localized to the mitochondrial intermembrane space, the potential
pathological role of SOD1 in mitochondrial dysfunction was unclear until we
demonstrated an accumulation of misfolded SOD1 in the mitochondria of several
different animal models of ALS. A number of issues regarding the role of misfolded
SOD1 in mitochondrial dysfunction in ALS are unresolved. It is currently unknown
whether misfolded SOD1 causes mitochondrial toxicity or what the mechanism of
toxicity might be. Suggestions that misfolded SOD1 bind to essential mitochondrial
substituents need to be rigorously tested. Additionally, the selective degeneration of
motor neurons might arise from the unique properties of spinal cord/brain mitochondria
or from the load of misfolded SOD1 in/on these mitochondria. Lastly, it is unclear
whether misfolded SOD1 in the mitochondria are generated in situ or whether they are
generated in the cytoplasm, or some other subcellular compartment, and then bound to
mitochondrial proteins or membranes. Elliott and coworkers recently demonstrated that
205
overexpression of the copper chaperone for SOD1 (CCS) caused an increased
mitochondrial load of mutant SOD1, mitochondrial pathology and greatly reduced the
lifespan of G93A SOD1/CCS double transgenic mice(48). Mitochondrial pathology has
also been described in human cases of ALS(49). If misfolded SOD1 is toxic to
mitochondria, this activity should be reproducible in an in vitro system analogous to
Koch’s postulates for proving the toxicity of a particular agent. Mitochondria isolated
from non-transgenic animals or cells treated with misfolded SOD1 should produce the
same functional and morphological changes observed in animals and mitochondria
isolated from animal models of ALS should be worsened by this treatment. In particular,
energy production/oxygen utilization should be decreased, and expansion of the
intermembrane space to resemble vacuoles should be observed. Mitochondria obtained
from various tissues may have differing vulnerability to misfolded SOD1 stress;
mitochondria may have to be isolated directly from spinal cord/brain of animals. If
misfolded-SOD1 mitochondrial toxicity arises from a biological process rather than a
simple physical association, misfolded SOD1 could be added to cells directly (e.g.
through liposomes) or the expression of a constituatively unfolded mutant SOD1 could be
induced. This would be followed by functional and morphological characterization of
mitochondria from treated cells.
Uses of SEDI in basic research (mechanisms of ALS)
A number of future experiments utilizing the SEDI antibody have been outlined above.
The selectivity of the SEDI antibody allows it to be used in further research into the
mechanisms of SOD1-mediated ALS. The basic uses of the SEDI antibody are those of
any antibody: localization and quantification. The SEDI antibody could be used to show
206
localization of monomer/misfolded SOD1 in mice or cells using conventional
immunohistochemistry, immunofluorescence or immuno-gold electron microscopy.
Localization within cells as a function of disease would yield clues to the mechanism of
misfolded SOD1 toxicity. For example, monomer/misfolded SOD1 might be present
initially in the cytoplasm and mitochondria, but not cause motor neuron disruption/death
until it is present in the endosomal pathway. Using electron microscopy, localization
within the mitochondria and/or co-localization with additional mitochondrial markers
could yield potential sites of misfolded SOD1 binding and clues to its mode of putative
mitochondrial toxicity. Western blot densitometry following SEDI immunoprecipitation
or ELISA could be used to measure the load of monomer/misfolded SOD1 in cells or
tissues and correlate this with factors including: protein life-time, intrinsic and extrinsic
oxidative stress, tissues of interest, localization within spinal cord, disease time-course,
etc. Co-immunoprecipitation experiments with SEDI might reveal novel disease-specific
SOD1-protein interactions.
Another use for the SEDI antibody might be in reverse genetic screening for
genes that alter SOD1 maturation, stability or structure. In Alzheimer’s disease,
mutations in APP, which is proteolytically cleaved in two places to produce the plaque-
forming Aβ peptide, can cause an alteration in Aβ species or aggregation(50). Mutations
in genes that are involved in APP proteolytic processing have also been implicated in
Alzheimer’s disease by altering Aβ concentration and ratio in the presence of wild-type
APP(50). SOD1 undergoes extensive post-translational modification prior to maturation:
N-terminal acetylation, dimerization of the nascent polypeptides, insertion of zinc,
insertion of copper and formation of the intrasubunit disulfide bond. Failure or alteration
207
in efficiency of one of these might cause SOD1 to fail to fold or misfold. Copper
insertion into SOD1 can occur through interactions with the copper chaperone for SOD1
(CCS)(51) or through an unknown CCS-independent pathway for copper insertion in
mammals and C. elegans(52, 53). A role for CCS-dependent SOD1 disulfide bond
formation has also been postulated, but has not been confirmed(54). Other genes that
might alter SOD1 processing or stability are unknown. A gene-deletion or gene-silenced
library could be screened for genes that increase the amount of monomer/misfolded
SOD1. Yeast SOD1 is 55.6% identical to human SOD1, and nearly identical in the SEDI
epitope (RPACGIVIL vs. RLACGVIGI); monomer/misfolded SOD1 from yeast may
bind to SEDI or a specific yeast SEDI antibody could be made. Yeast from the gene-
deletion array covering the entire yeast genome(55) would then be probed with SEDI and
fluorescently tagged. SEDI-labelled yeast would then be separated using fluorescence
activated cell-sorting (FACS) and identified using deletion associated ‘barcode’
sequencing tags or arrays where the position of each deletion mutant is known could be
scanned for SEDI-related fluorescence. CCS-/- cells should be a positive control: lack of
CCS should cause SOD1 to not fold properly. The entire array could then be subjected to
oxidative stress via treatment with hydrogen peroxide or paraquat to increase the amount
of misfolded SOD1. Because zinc is thought to play a primarily structural role in SOD1,
loss of whatever gene is responsible for zinc insertion into SOD1 should also increase the
amount of monomer/misfolded SOD1. Genes responsible for disulfide bond formation or
trafficking SOD1 to various subcellular compartment might be uncovered in a similar
manner. Analogous reverse genetic screens using mammalian cells could be carried out
208
using the library of siRNAs that covers the entire human genome(56-58), though this
array is not as well validated as the yeast gene-deletion array.
Uses of SEDI antibody in translational research
ALS has been understood as a disease for over one hundred years, but its diagnosis is
based on neurological criteria and elimination of similar diagnoses(1). There is only one
FDA approved medication for ALS, riluzole, which extends life on average by only 1-2
months(59). Riluzole inhibits glutamate transporters, which are thought to play a role in
excitotoxicity in ALS. Because mutations in SOD1 remain the only known cause of
typical ALS, treatments that target the mechanism of SOD1 toxicity could dramatically
affect the lives of individuals with SOD1 mutations. If misfolding of SOD1 is found to
underlie a subset of sporadic ALS cases, these individuals may also benefit from
treatments targeting SOD1. Because the SEDI antibody selectively targets misfolded
SOD1, it could have many uses in translational research, including diagnosis, drug
discovery, and in therapy.
SOD1 mutations can be screened in familial cases of ALS by PCR sequencing
from exons of genomic DNA. A significant number (~5%) of sporadic ALS cases also
have SOD1 mutations. Routine genetic testing is avoided because of ethical and
economic reasons; in the USA, families of patients with SOD1 mutations might be denied
health care coverage(60). SOD1 mutants are typically highly penetrant by the seventh
decade of life in familial ALS, but the penetrance and modifying factors in sporadic ALS
is unknown(61). Because ALS patients with mutations in SOD1 have variable ages of
onset, it is currently unknown when to start physical therapy or riluzole and relies on the
same neurological criteria for diagnosis as sporadic (non-SOD1) ALS. We have shown
209
that monomer/misfolded SOD1 detected with the SEDI antibody correlates well with
disease; because SOD1 can become extracellular, we may be able to detect misfolded
SOD1 in CSF or blood of ALS patients to serve as a biochemical test for ALS where
currently none exists. Additionally, clinical trials for ALS are currently conducted where
survival and functional rating score (FRS) are the primary and secondary outcome
measures(62). Quantification of monomer/misfolded SOD1 using SEDI in easily
extracted fluids could be used as an objective marker for disease progression. Correlation
of SEDI reactive protein from CSF with other proteins/small molecules may reveal novel
biomarkers for ALS that can be generalized to non-misfolded SOD1 ALS. Use in a
preclinical setting is more straight-forward because animals under trial can be sacrificed
to examine tissues of interest directly. The SEDI antibody can be used as an auxiliary
biochemical marker, in addition to survival and other phenotypic measures, to measure
the efficacy of potential therapeutics by directly monitoring disease associated proteins.
If monomer/misfolded SOD1 or aggregated SOD1 that is produced through a
similar monomeric intermediate mechanism is a toxic species in ALS, reducing the
amount of this misfolded protein should have therapeutic benefit. Two not mutually
exclusive strategies might involve: a) stabilizing the native dimer to prevent
monomerization and subsequent aggregation, or b) removing/degrading the misfolded
species or otherwise neutralizing its toxic effect. So called ‘chemical chaperones’ have
been used to stabilize proteins and prevent them from mis/unfolding in transthyretin
(TTR) amyloidosis(63) and acid beta-glucosidase (GCase)(64) in Gaucher’s disease.
Lansbury and coworkers used computational screening to find molecules that stabilize
the native SOD1 dimer in vitro(65); however, in the single trial with these molecules, it
210
did not affect viability in a chick embryo expressing mutant SOD1(66). Other molecules
with improved pharmacological properties (Lipinski’s ‘Rule of Five’) and penetration of
the blood-brain barrier may have to be found. We recently found that minocycline, which
had been shown to increase life-span in ALS mice and penetrate the blood brain barrier,
actually binds to mutant SOD1 and increases its melting temperature by approximately
7°C. Unfortunately, binding of minocycline to SOD1 is weak and perhaps non-specific,
requiring large amounts of the drug to stabilize physiological amounts of protein(67). An
ALS clinical trial using minocycline, even at lower doses than required to stabilize SOD1
meaningfully, showed an acceleration of disease progression(68, 69). The SEDI antibody
could be used in a cell-based or cell-free screen for compounds that stabilize SOD1. In a
cell-based assay, cells would be subjected to some protein denaturational stress, such as
oxidative stress through treatment with hydrogen peroxide or paraquat. The amount of
misfolded SOD1 could be quantified using SEDI to screen for compounds that reduce the
amount of misfolding. In a cell-free assay, purified SOD1 protein would be unfolded
using chaotropes in the presence of library compounds and the amount of
monomer/misfolded SOD1 would again be quantified using the SEDI antibody.
Selective removal of misfolded SOD1 could be accomplished by increasing the
efficiency by which it is targeted for degredation and actually broken down. Autophagy
and the ubiquitin-proteasome system are complimentary and overlapping means of
degrading misfolded proteins, coupled through the activity of histone deacetylase 6
(HDAC6)(70), which is also a critical component of aggresomes and stress granules(71).
SOD1 is degraded through both the ubiquitin-proteasome system and the lysosome
through autophagy(72). The ubiquitin E3 ligase Dorfin ubiquitinates mutant SOD1 and
211
targets it for degredation(73); if the activity of Dorfin or the concentration of Dorfin
could be increased by pharmacological intervention, this might form the basis for a
therapeutic strategy for SOD1-ALS. Lysosomes degrade organelles such as the
mitochondria at the end of their lifetime through autophagy. It has recently been shown
that mice deficient in neuronal autophagy by knocking out either of Atg5 or Atg7
experience motor neuron degeneration accompanied by the development of ubiquitin
positive inclusions in these neurons(74, 75). Increasing the efficiency of autophagy
through pharmacological means might then protect motor neurons by disposing of both
misfolded proteins and dysfunctional mitochondria. Autophagy is increased in G93A
SOD1 mutant mice(76). Autophagy can be induced in mammalian cells by treatment with
rapamycin, which inhibits mTOR, itself a negative regulator or autophagy, or lithium,
which activates autophagy through inhibition of inositol monophosphatase(77). In a
small single-blind clinical trial, lithium slowed disease progression in ALS as measured
by ALS-FRS and survival(78). These drugs are promising candidates as ALS
therapeutics, but have pleiotropic effects, a narrow therapeutic concentration range and
potential side effects. Additionally, enhancing the ubiquitin-proteasome system or
autophagy helps to clear intracellular proteins, but perhaps not extracellular proteins that
may be responsible for spread of ALS pathology. Because SEDI binds specifically to
monomer/misfolded SOD1, it could potentially neutralize extracellular misfolded SOD1
and perhaps slow disease progression. Endogenous antibodies could be raised to the
SEDI antigen (active immunization) or SEDI antibodies could be injected intrathecally
(passive immunization). Reducing the total amount of total SOD1 by immunization with
mutant SOD1 led to a moderate increase in lifespan in ALS mice. Liu et al recently
212
showed that immunization of G37R SOD1 mice with either the SEDI antigen or
oxidation induced SOD1 aggregates increased survival and disease duration
significantly(43, 79). Phenotypic effect was proportional to immune response and
corresponded to lower amounts of misfolded SOD1. These results support our hypothesis
that monomer/misfolded SOD1 is a toxic species in ALS, that reducing the concentration
of misfolded SOD1 ameliorated disease and that immunization with the SEDI antigen is
an effective method to reduce the toxic load of misfolded SOD1. Because this strategy
targets only pathological, misfolded SOD1, this immunization strategy may avoid
problematic auto-immunity/inflammation problems associated with recent clinical trials
using a non-structure specific immunization of Aβ in Alzheimer’s disease(80, 81).
Generalizability of SEDI strategy
To the best of my knowledge, we are the first to use the high-resolution structure of a
protein to predict an epitope on a protein to generate antibodies selective for certain
conformational forms of that protein. This approach was successful for SOD1, but
whether this strategy is generalizable and can be applied to other proteins is unknown. If
successful, this strategy may see the generation of a suite of antibodies targeting proteins
in misfolding diseases, or otherwise, that may have important applications analogous to
those discussed earlier in this chapter for the SEDI antibody. Because monoclonal
antibodies can be used as therapeutics, structure guided design of conformation-
specific/selective antibodies may be a straight-forward method to implement structure-
guided drug-design.
The following algorithm describes a generalization of the strategy we used to
generate the SEDI antibody:
213
1. Get the high resolution structure of the protein.
2. Have/acquire information on the protein’s misfolding pathway or other
pathological activity.
3. Find residues buried in the native structure, but exposed in misfolded/pathological
structure.
4. Make peptide, antibody, and validate.
The first requirement to generate conformation-selective antibodies is to know the
structure of the protein of interest; in many cases, this is already known and publicly
accessible through the Protein DataBase (PDB). Proteins associated with diseases are
high-value targets for structure elucidation by x-ray crystallography, NMR, computer-
modeling or a combination of these techniques. As such, the high-resolution structures of
many disease-associated proteins have been solved. Structural genomics initiatives have
also yielded significant numbers of new protein structures, of which some proteins may
be linked to disease. In addition to gaining insight into the disease process through
analysis of the high-resolution structure of disease-associated proteins, these protein
structures could also be used to generate conformation-selective antibodies.
Detailed biochemical and biophysical analyses of the protein misfolding pathway
or pathological processes have also been undertaken for many disease proteins. For
protein misfolding diseases, this might include alterations in mRNA splicing(82),
proteolytic cleavage(50), dynamics of certain parts of the protein structure(83, 84),
changes to its oligomerization(28, 85), changes to binding interfaces with other
proteins(86), etc. For foreign pathological proteins, this might include the identification
of surfaces important for binding to host proteins(87), invariant regions of proteins,
214
essentiality of proteins, cell surface expression, etc. This data allows us to distinguish
pathological proteins from normal ones and to identify regions of foreign proteins that are
less likely to avoid immune recognition by mutation. In each case, a region of the whole
protein is identified that is critical for its toxicity/virulence. These data can be integrated
and mapped onto the protein’s structure. Oligomerization interfaces and protein-protein
interaction surfaces in general are good targets if these interactions either occur only in
the disease or in the absence of disease. Hydrogen/deuterium exchange experiments also
yield useful information on solvent exposure at a residue-specific level when designing a
selective antibody. The targeted region must be solvent, or at least antibody, inaccessible
in native form and accessible in toxic form, rather than any buried sequence. Solvent
exposure in various forms of the protein or protein complex may be predicted using
software tools such as Swiss PDB viewer from Expassy(88).
The selected region/surface may be composed of either a contiguous segment of
the polypeptide or from distal amino acids close in three dimensions. Producing a three
dimensional surface that is a subset of the protein to make antibodies selective for that
region is technically challenging; I will limit the discussion to cases where a linear
peptide composed of at least five amino acid residues. Multiple antigenic peptides
(MAPs) are chemically synthesized using standard Fmoc-based chemistry where each
strand of the dendrimer is a linear peptide within the protein’s region of interest. Rabbit
polyclonal or other antibodies are then made to these MAPs. These antibodies should be
selective for misfolded conformations or pathologically critical regions of the protein
under investigation, but this must be verified. Because antibodies can potentially react
differently in various applications, these putatively selective antibodies should be tested
215
in a variety of immunological assays for specificity. We compared the reactivity of SEDI
with native, folded SOD1 and unfolded SOD1 in Western blots, ELISA and
immunoprecipitation reactions(37). Assays to compare the reactivity of novel selective
antibodies may be analogous. Alternatively, if monoclonal antibodies with desired
specificity are known, competitive binding of the monoclonal antibody and
uncharacterized antibody to a substrate in ELISA or other assay may be used to verify the
specificity of the novel antibody.
Several protein misfolding diseases are attractive candidates to test the
generalizability of the structure-guided antibody design approach. In two of these,
transthyretin (TTR) and β2-microglobulin (β2m), the structure of the protein implicated
is very closely related to the SOD1 structure: each polypeptide forms a Greek-key β-
barrel fold in the native state(89, 90). TTR is tetrameric in the native state where
dissociation into monomers precedes amyloid formation(63); dissociation of β2m, which
is part of the major histocompatibility complex and forms a dimer with either MHC class
I or MHC class II polypeptides, is also required for amyloid formation(91). Each also has
a β-strand sequestered in a native oligomerization interface that should become exposed
in the monomer and perhaps the final aggregated form. Because of the striking
similarities between the TTR and SOD1 structure and misfolding pathways, it has
greatest chance of replicating our success with the SEDI antibody.
TTR misfolding and amyloid deposition is implicated in several disorders
including senile systemic amyloidosis, familial amyloid cardiomyopathy and familial
amyloid polyneuropathy(92). Both wild-type TTR and mutant TTR can form amyloid
deposists, where 4% of African Americans carry TTR mutations predisposing them to
216
amyloidosis(93). TTR is a serum protein that is produced in the liver(94); currently,
diagnosis is by biopsy and the only treatment is liver transplant, which is of limited
efficacy. If we could produce antibodies selective for misfolded TTR, this might have
potential diagnostic and therapeutic uses analogous to those for the SEDI antibody. I
chose an epitope in the β-strand in the TTR oligomer interface that is buried in the native
TTR homotetramer and exposed in the monomer/misfolding intermediate (Figure 1);
antibodies raised to a MAP derived from this sequence might be specific for
monomer/misfolded TTR. We raised rabbit polyclonal antibodies to this segment of TTR
(misTTR antibody) and investigated its reactivity in vitro. We used competition ELISA
to examine the reactivity of the misTTR and a whole-molecule TTR antibody from
Sigma with native TTR, unfolded TTR, the misTTR MAP antigen and TTR amyloid.
Competition ELISA has two principle advantages over other immunological assays: both
the target antigen and antibody are in solution and the assay is quantitative. By examining
a solution interaction, potentially protein-structure altering effects by binding to the
ELISA surface can be avoided. The apparent binding constant of the antibody in question
can also be quantified easily and using this we can directly calculate the amount of
antigen in solution. As predicted, the misTTR antibody reacted with the misTTR MAP
antigen, but not folded TTR (Figure 2a). Commercial TTR polyclonal antibodies react
with folded or unfolded TTR, but do not react with the TTR MAP antigen, indicating that
it is not selective for TTR structure and that it has no overlapping activity with the
misTTR antibody (Figure 2b). Importantly, the misTTR antibody also binds to TTR
amyloid without showing competing binding to folded TTR at similar concentrations
(Figure 2c). Thus, the misTTR antibody satisfies the requirements for an antibody
217
Figure 1. Spacing filling models of transthyretin (TTR) from pdb 1BMZ, prepared with PyMol (Delano Scientific). Top: Native TTR tetramer with misTTR epitope buried. Bottom:Monomeric TTR with misTTR epitope exposed.
218
misTTR competition ELISA
0
10000
20000
30000
40000
50000
60000
0.001 0.01 0.1 1 10
competitor (uM)
RFU TTR
MAP
anti-TTR Competition ELISA
0
10000
20000
30000
40000
50000
0.001 0.01 0.1 1 10
competitor (uM)
RFU TTR
MAP
a) b)
c) misTTR Ab Competition ELISA
0
10000
20000
30000
40000
50000
60000
70000
80000
0.01 0.1 1 10
[competitor] / uM
Fluo
resc
enc
folded TTRTTR fibrils
Figure 2. Competition ELISA illustrates specificity of TTR antibodies (concentration in monomer units)
a) Antibodies raised to buried epitope do no react with native TTR (misTTR), even at high concentrations whereas they do react with the cognate multiple antigenic peptide (MAP)
b) Commercial TTR antibodies react with folded TTR, but not with the misTTR MAP, indicating that the MAP epitope is not exposed during normal antibody production
c) misTTR antibodies react with TTR amyloid fibrils, but not with native TTR at similar concentrations
219
specific for misfolded TTR in vitro. Reactivity of the misTTR antibody with
biopsy/autopsy and serum samples from amyloidosis patients and controls is currently
under investigation.
Members of our laboratory have pursued a structure-guided antibody design
strategy for other proteins, including PrP and p53. These proteins have a different fold
than SOD1 and from one another. Using the structure-guided approach, antibodies
selective for misfolded PrP have been produced. These antibodies are directed at an
epitope that is buried in the native PrP monomer but exposed in the misfolded monomer.
Oligomerization of p53 occurs through a small C-terminal tetramerization domain
consisting of an α-helix and a β-strand. I hypothesized that mutations in the distal DNA
binding domain of p53 might alter its oligomerization; this could be tested in vivo using
an antibody selective for the p53 monomer (Figure 3). This and other putative p53
structure-specific antibodies are currently under investigation. The structure-guided
antibody design approach thus appears to yield antibodies with the desired selectivity in
at least several cases; whether it will be generally applicable remains to be seen.
Conclusion
Because SOD1 is an extremely stable protein, we hypothesized that a physiological
denaturational stress is necessary to cause it to aggregate. SOD1 is stabilized by
homodimerization and binding to Cu2+ and Zn2+; oxidation of key metal-binding histidine
residues destabilizes SOD1 and causes it to aggregate in vitro through a monomeric
misfolding intermediate. To investigate whether SOD1 undergoes a similar misfolding
pathway in vivo, we designed an antibody selective for monomer/misfolded SOD1. Using
this antibody, we found that SOD1 misfolding in vivo behaved very similarly to how we
220
predicted: monomer/misfolded SOD1 was localized primarily within motor neurons in
the spinal cord, wild-type SOD1 also misfolded in vivo, misfolded SOD1 was initially
absent but appeared prior to symptom onset and misfolded SOD1 accumulates in the site
of principle source of oxidative stress in vivo, the mitochondria within the brain and
spinal cord. The SEDI antibody and other structure-guided antibodies might prove further
useful in basic and translational research; these uses have been summarized below
(Figure 3).
221
SEDI
Generalizability• test concept of structure based antibody design on other proteins
Diagnostic• CSF/blood test for ALS• correlation with other markers• Follow disease course in animals treated with lead compounds
Therapeutic• active/passive immunization strategies to selectively inactivate misfolded SOD1
Research
Drug Discovery• structural probe in small molecule screens that increase SOD1 stability
Basic Science• toxicity of misfolded SOD1• SOD1 oxidation-misfolding connection• screen for genes that modify SOD1 structure/stability• species that interact with misfolded SOD1
Figure 3. Summary of uses for SEDI antibody.
222
References:
(1) (1996) The Amyotrophic Lateral Sclerosis Functional Rating Scale. Assessment
of activities of daily living in patients with amyotrophic lateral sclerosis. The ALS
CNTF treatment study (ACTS) phase I-II Study Group. Arch Neurol 53, 141-7.
(2) Cudkowicz, M. E., Shefner, J. M., Schoenfeld, D. A., Brown, R. H., Jr., Johnson,
H., Qureshi, M., Jacobs, M., Rothstein, J. D., Appel, S. H., Pascuzzi, R. M.,
Heiman-Patterson, T. D., Donofrio, P. D., David, W. S., Russell, J. A., Tandan,
R., Pioro, E. P., Felice, K. J., Rosenfeld, J., Mandler, R. N., Sachs, G. M.,
Bradley, W. G., Raynor, E. M., Baquis, G. D., Belsh, J. M., Novella, S.,
Goldstein, J., and Hulihan, J. (2003) A randomized, placebo-controlled trial of
topiramate in amyotrophic lateral sclerosis. Neurology 61, 456-64.
(3) Steele, J. C., and McGeer, P. L. (2008) The ALS/PDC syndrome of Guam and the
cycad hypothesis. Neurology 70, 1984-90.
(4) Gkogkas, C., Middleton, S., Kremer, A. M., Wardrope, C., Hannah, M.,
Gillingwater, T. H., and Skehel, P. (2008) VAPB interacts with and modulates the
activity of ATF6. Hum Mol Genet 17, 1517-26.
(5) Neumann, M., Sampathu, D. M., Kwong, L. K., Truax, A. C., Micsenyi, M. C.,
Chou, T. T., Bruce, J., Schuck, T., Grossman, M., Clark, C. M., McCluskey, L. F.,
Miller, B. L., Masliah, E., Mackenzie, I. R., Feldman, H., Feiden, W.,
Kretzschmar, H. A., Trojanowski, J. Q., and Lee, V. M. (2006) Ubiquitinated
223
TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis.
Science 314, 130-3.
(6) Arai, T., Hasegawa, M., Akiyama, H., Ikeda, K., Nonaka, T., Mori, H., Mann, D.,
Tsuchiya, K., Yoshida, M., Hashizume, Y., and Oda, T. (2006) TDP-43 is a
component of ubiquitin-positive tau-negative inclusions in frontotemporal lobar
degeneration and amyotrophic lateral sclerosis. Biochem Biophys Res Commun
351, 602-11.
(7) Sreedharan, J., Blair, I. P., Tripathi, V. B., Hu, X., Vance, C., Rogelj, B.,
Ackerley, S., Durnall, J. C., Williams, K. L., Buratti, E., Baralle, F., de
Belleroche, J., Mitchell, J. D., Leigh, P. N., Al-Chalabi, A., Miller, C. C.,
Nicholson, G., and Shaw, C. E. (2008) TDP-43 mutations in familial and sporadic