<%BANNER%>

Mechanisms of Superoxide Dismutase 1 Aggregate Formation in Familial Amyotrophic Lateral Sclerosis

Permanent Link: http://ufdc.ufl.edu/UFE0024366/00001

Material Information

Title: Mechanisms of Superoxide Dismutase 1 Aggregate Formation in Familial Amyotrophic Lateral Sclerosis
Physical Description: 1 online resource (149 p.)
Language: english
Creator: Karch, Celeste
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: amyotrophic, disease, dismustase, lateral, misfolding, models, mouse, neurodegenerative, protein, sclerosis, superoxide
Neuroscience (IDP) -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Familial amyotrophic lateral sclerosis (FALS) is a late onset neurodegenerative disease that is characterized by selective death of neurons in the upper and lower motor neuron pathways, resulting in progressive paralysis. A subset of FALS cases are linked to dominantly inherited missense mutations in Cu,Zn superoxide dismutase 1 (SOD1). More than 100 mutations in SOD1 are known to cause FALS by a proposed gain of toxic property. One possible mechanism for this toxicity is the formation of aggregates. In the spinal cords of mouse models of ALS overexpressing mutant SOD1, detergent-insoluble, high-molecular-weight, SOD1 aggregates accumulate as mice develop paralysis. The mechanism underlying the formation and function of these aggregates is poorly understood. In this study, we examined the intrinsic factors that mediate the aggregation of SOD1 using a mutagenesis approach and cell culture assay for aggregation. In cell culture, our data suggest that in SOD1 aggregates, cysteines 6 and 111 of SOD1 may play important roles in the early stages of aggregate formation with other structural features of the aggregate providing additional stability. Furthermore, we found that amino acids in beta-strands 6 and 7 in human SOD1 are important for aggregation and that species-specific interactions between these two regions are important for enhancing aggregation. In our animal model of ALS, our findings demonstrate that the accumulation of disulfide cross-linked mutant protein is co-incident with the accumulation of detergent-insoluble aggregates of mutant protein, with both of these events occurring well after the appearance of multiple pathologic abnormalities but concurrent with the onset of symptoms. Together, these studies demonstrate that aggregates are composed of globally misfolded, immature SOD1 protein that contribute to the conversion of disease phenotype. We suggest that alternative forms of misfolded SOD1 proteins impart toxicity early in the disease course.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Celeste Karch.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Borchelt, David Ralph.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024366:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024366/00001

Material Information

Title: Mechanisms of Superoxide Dismutase 1 Aggregate Formation in Familial Amyotrophic Lateral Sclerosis
Physical Description: 1 online resource (149 p.)
Language: english
Creator: Karch, Celeste
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: amyotrophic, disease, dismustase, lateral, misfolding, models, mouse, neurodegenerative, protein, sclerosis, superoxide
Neuroscience (IDP) -- Dissertations, Academic -- UF
Genre: Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Familial amyotrophic lateral sclerosis (FALS) is a late onset neurodegenerative disease that is characterized by selective death of neurons in the upper and lower motor neuron pathways, resulting in progressive paralysis. A subset of FALS cases are linked to dominantly inherited missense mutations in Cu,Zn superoxide dismutase 1 (SOD1). More than 100 mutations in SOD1 are known to cause FALS by a proposed gain of toxic property. One possible mechanism for this toxicity is the formation of aggregates. In the spinal cords of mouse models of ALS overexpressing mutant SOD1, detergent-insoluble, high-molecular-weight, SOD1 aggregates accumulate as mice develop paralysis. The mechanism underlying the formation and function of these aggregates is poorly understood. In this study, we examined the intrinsic factors that mediate the aggregation of SOD1 using a mutagenesis approach and cell culture assay for aggregation. In cell culture, our data suggest that in SOD1 aggregates, cysteines 6 and 111 of SOD1 may play important roles in the early stages of aggregate formation with other structural features of the aggregate providing additional stability. Furthermore, we found that amino acids in beta-strands 6 and 7 in human SOD1 are important for aggregation and that species-specific interactions between these two regions are important for enhancing aggregation. In our animal model of ALS, our findings demonstrate that the accumulation of disulfide cross-linked mutant protein is co-incident with the accumulation of detergent-insoluble aggregates of mutant protein, with both of these events occurring well after the appearance of multiple pathologic abnormalities but concurrent with the onset of symptoms. Together, these studies demonstrate that aggregates are composed of globally misfolded, immature SOD1 protein that contribute to the conversion of disease phenotype. We suggest that alternative forms of misfolded SOD1 proteins impart toxicity early in the disease course.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Celeste Karch.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Borchelt, David Ralph.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024366:00001


This item has the following downloads:


Full Text

PAGE 1

1 MECHANISMS OF SUPEROXIDE DISM UTASE 1 AGGREGATE FORMATION IN FAMILIAL AMYOTROPHIC LATERAL SCLEROSIS By CELESTE MARIE KARCH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Celeste Marie Karch

PAGE 3

3 To my family for their love and support

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank m y advisor, Dr. David R. Borchelt, for his support and guidance. I will always be grateful for the many opportunities that Dr. Borchelt provided. I also thank Dr. Borchelt for challenging me and for teaching me how to think like a scientist. I express my gratitude to my committee memb ers: Dr. Lucia Notterpek, Dr. Wolfgang Streit, and Dr. William Dunn. I would like to thank Dr. Susan Semple-Row land for her contributions to my growth as a scientist. I would also like to thank our collaborators for their advice, support, and assistance: Dr. Joan Valentine and the members of her labor atory in the Department of Chemistry at the University of California at Los Angeles, Dr Julian Whitelegge and the members of his laboratory in the Pasarow Mass Spectrometry Laborat ory at the University of California at Los Angeles, and Dr. P. John Hart and the members of his laboratory at the University of Texas at San Antonio. Finally, I would like to thank past and present memb ers of the Borchelt laboratory for their advice, guidance, and friendship. In particular, I would lik e to thank Mercedes Prudencio for her support.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 ABSTRACT...................................................................................................................................12 CHAP TER 1 INTRODUCTION..................................................................................................................14 Amyotrophic Lateral Sclerosis............................................................................................... 14 Genetics of FALS............................................................................................................... ....14 SOD1-Linked FALS...............................................................................................................16 Genetics of SOD1-Linked FALS.................................................................................... 16 SOD1 Protein...................................................................................................................17 Toxic Property of Mutant SOD1.....................................................................................22 Models of FALS.....................................................................................................................23 Mouse Models of FALS..................................................................................................23 Alternative mouse models of SOD1-linked FALS .................................................. 24 SOD1 toxicity is non-cell autonomous.................................................................... 25 Cell Culture Model of Mutant SOD1 Aggregation......................................................... 26 2 A LIMITED ROLE FOR DISULFIDE CRO SS-L INKING IN THE AGGREGATION OF MUTANT SOD1 LINKED TO FA MILIAL AMYOTROPHIC LATERAL SCLEROSIS...........................................................................................................................27 Introduction................................................................................................................... ..........27 Methods..................................................................................................................................28 Construction of SOD1 Expression Vectors..................................................................... 28 Tissue Culture Transfecti on and Transgenic Mice .......................................................... 28 SOD1 Aggregation Assay by Differential Extraction..................................................... 29 Assay for Disulfide Cross-Linked SOD1........................................................................30 Immunoblotting...............................................................................................................30 Quantitative Analysis of Immunoblots............................................................................31 Results.....................................................................................................................................31 Discussion...............................................................................................................................41 Cysteines 6 and 111 Modulate A ggregation of Mutant SO D1........................................ 45 Structural Features of Cysteine 111................................................................................ 47 Disulfide Bonding in Mutant SOD1 Aggregation........................................................... 48 Conclusions.....................................................................................................................49 3 AGGREGATION MODULATING ELEMENTS IN HUM AN SOD1 PROTEIN............... 51

PAGE 6

6 Introduction................................................................................................................... ..........51 Methods..................................................................................................................................52 Construction of SOD1 Expression Vectors..................................................................... 52 Tissue Culture Transient Transfection............................................................................ 53 SOD1 Aggregation Assay by Differential Extraction..................................................... 53 Immunoblotting...............................................................................................................54 Quantitative Analysis of Immunoblots............................................................................54 Results.....................................................................................................................................55 Discussion...............................................................................................................................62 4 THE ROLE OF MUTANT SOD1 DISU LFIDE OXIDATI ON AND AGGREGATION IN THE PATHOGENESIS OF FAMILIAL ALS.................................................................. 67 Introduction................................................................................................................... ..........67 Methods..................................................................................................................................69 Transgenic Mice.............................................................................................................. 69 SOD1 Aggregation Assay by Differential Extraction..................................................... 69 Assay for Disulfide Cross-Linked SOD1........................................................................70 Immunoblotting...............................................................................................................70 Assay for Detection of Reduced and Oxidized SOD1.................................................... 70 Quantitative Analysis of Immunoblots............................................................................71 Results.....................................................................................................................................71 Discussion...............................................................................................................................86 Redox Chemistry and SOD1 Aggregation in SOD1 Transgenic Mice...........................87 Post-Translational Modification of Mutant SOD1 and Aggregation .............................. 88 Mutant SOD1 Aggregation in Disease Pathogenesis...................................................... 92 SOD1 Aggregation and Toxicity..................................................................................... 96 Conclusions.....................................................................................................................99 5 B CRYSTALLIN IS A MODEST MODIFI ER OF DISEASE PROGRESSION IN MOUSE MODELS OF ALS................................................................................................ 100 Introduction................................................................................................................... ........100 Methods................................................................................................................................102 Tissue Culture Transfecti on and Transgenic Mice ........................................................ 102 SOD1 Aggregation Assay by Differential Extraction................................................... 103 Immunoblotting.............................................................................................................103 Quantitative Analysis of Immunoblots..........................................................................103 Antibodies..................................................................................................................... .104 Histology.......................................................................................................................105 Results...................................................................................................................................106 Discussion.............................................................................................................................116 Myopathy Associated with ALS Disease Course.......................................................... 117 The Role of Heat Shock Protei ns in the ALS Disease Course ...................................... 118 Disease Threshold in SOD1 Transgenic Mice.............................................................. 119 Conclusions...................................................................................................................124

PAGE 7

7 6 CONCLUSIONS.................................................................................................................. 125 Composition of Mutant SOD1 Aggregates........................................................................... 125 The Role of SOD1 Aggregates in ALS Disease Course....................................................... 127 Modifiers of SOD1 Aggrega tion and FALS Pathogenesis ................................................... 129 Insights into Therapeutics.....................................................................................................129 Future Directions..................................................................................................................131 Conclusions...........................................................................................................................132 LIST OF REFERENCES.............................................................................................................133 BIOGRAPHICAL SKETCH.......................................................................................................149

PAGE 8

8 LIST OF TABLES Table page 1-1 SOD1-linked familial ALS mutations............................................................................... 19 2-1 Aggregation propensity of SOD1 m utants at cysteine residues......................................... 29 3-1 Aggregation propensity of chim eric SOD1 mutants.......................................................... 53

PAGE 9

9 LIST OF FIGURES Figure page 2-1 SOD1 aggregation of FALS cyst eine mutants in transfected cells .................................... 32 2-2 Role of cysteines 6 a nd 111 in SOD1 aggregation ............................................................ 34 2-3 Cysteine residues are not require d for SOD1 aggregate form ation...................................37 2-4 Role of cysteine 111 in mutant SOD1 aggregation ...........................................................39 2-5 Symptomatic G86R mice form detergent in soluble species in sp inal cord tissue. ............ 40 2-6 Mouse SOD1-G86R aggregates in cell culture..................................................................42 2-7 Intermolecular disulfide bonding by SOD1 mutants in expressed cultured cells .............. 43 2-8 Disulfide bonds are not sufficient to m aintain SOD1 aggregate structure........................ 44 2-9 Intermolecular disulfide bonding by SOD1 mutants in expressed cultured cells .............. 47 3-1 Human and mouse SOD1 di ffer at 25 am ino acids............................................................ 54 3-2 Cysteine 111 plays a role in SOD1 aggregation................................................................ 55 3-3 Differences in aggregation propensity between mouse and hum an SOD1 chimeric proteins...............................................................................................................................57 3-4 Amino acid 111 does not predict aggregation alone. ......................................................... 60 3-5 Amino acids 42-50 and 109-123 in human SOD1 are im portant for aggregation............. 63 3-6 Human amino acids 109-123 enhance aggr egation in FALS mu tants throughout the protein ................................................................................................................................64 3-7 Chimeric proteins have diffe rential aggregation propensities ........................................... 65 4-1 The levels of detergent-insoluble SOD1 in creas e dramatically late in the course of disease in SOD1 transgenic mice....................................................................................... 73 4-2 Variation in the levels of detergent-inso luble SO D1 in the spin al cords of paralyzed FALS mice...................................................................................................................... ...76 4-3 Detergent-insoluble SOD1 accumulates to high levels near d isease endstage.................. 77 4-4 Astrogliosis occurs early in disease progression in SOD1-H46R/H48Q mice.................. 78 4-5 Astrogliosis occurs at 150 days in L126Z m ice................................................................. 79

PAGE 10

10 4-6 B crystallin is upregulated in astrocytes at diseas e endstage in L126Z mice ..................80 4-7 The appearance of disulfide cross-linked SOD1 is coincid ent with the accumulation of detergent-insoluble mutant protein................................................................................82 4-8 SOD1 aggregates resist di ssociation in high concentrati ons of reducing agents. ............ 84 4-9 SOD1 aggregates are largely composed of disulfide-reduced form s of SOD1 in fresh spinal cord tissue............................................................................................................. ...85 4-10 Reduced forms of mutant SOD1 are com ponents of NP40-insoluble aggregates. ......... 90 4-11 Reduced hSOD1 protein is preferentially incorpo rated into detergent-insoluble aggregates..........................................................................................................................91 5-1 B crystallin reduces m utant SOD1 aggregation in cell culture..................................... 105 5-2 Overexpressed B crystallin is upregulated in response to mutant SOD1 in cell culture. .............................................................................................................................107 5-3 Hsp40 and Hsp70 are constitutivel y induced in HEK293-FT cells................................. 108 5-4 Reduction or elimination of B crystallin in m utant SOD1 transgenic mice does not substantially a lter survival...............................................................................................110 5-5 SOD1 aggregation is restricted to the brainstem and spinal co rd in Gn.L126Z mice..... 111 5-6 SOD1 aggregation is absent in muscle tissue.................................................................. 113 5-7 Reduction or elimination of B crystallin in SO D1 transgenic mice does not alter aggregation in spinal cord tissue...................................................................................... 114 5-8 SOD1 aggregation propensity in mi ce expressing varying levels of B crystallin. ......115 5-9 SOD1 aggregation propensity in PrP. G37R m ice expressing varying levels of B crystallin...........................................................................................................................116 5-10 B crystallin is upregulated in SOD1 transgenic m ice.................................................... 117 5-11 SOD1 antibodies recogni ze denatured for ms for SOD1.................................................. 118 5-12 In Gn.G37R mice, B crystallin does not alter loca lization of SOD1 accumulation.. .... 120 5-13 In Gn.L126Z mice, B crystallin does not alter local ization of SOD1 accum ulation.... 121 5-14 In PrP.G37R mice, B crystallin do es not alter local ization of SOD1 accumulation. .. 122 5-15 B crystallin is upregulated in astrocytes of Gn.L126Z m ice.........................................123

PAGE 11

11 6-1 Mutant SOD1 folding pathways throughout the ALS disease course .............................127 6-2 Disease progression in SOD1 transgenic mice................................................................ 128

PAGE 12

12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MECHANISMS OF SUPEROXIDE DISM UTASE 1 AGGREGATE FORMATION IN FAMILIAL AMYOTROPHIC LATERAL SCLEROSIS By Celeste Marie Karch May 2009 Chair: David R. Borchelt Major: Medical Sciences Neuroscience Familial amyotrophic lateral sc lerosis (FALS) is a late ons et neurodegenerative disease that is characterized by selective death of neurons in the upper and lower motor neuron pathways, resulting in progressive paralysis. A subset of FALS cases ar e linked to dominantly inherited missense mutations in Cu,Zn superoxide dismutase 1 (SOD1). More than 100 mutations in SOD1 are known to cause FALS by a proposed gain of toxic property. One possible mechanism for this toxicity is the forma tion of aggregates. In the spinal cords of mouse models of ALS overexpressing mutant SOD1, de tergent-insoluble, hi gh-molecular-weight, SOD1 aggregates accumulate as mice devel op paralysis. The mechanism underlying the formation and function of these aggregates is po orly understood. In this study, we examined the intrinsic factors that mediate the aggregation of SOD1 using a mutagenesis approach and cell culture assay for aggregation. In cell culture, our data suggest that in SOD1 aggregates, cysteines 6 and 111 of SOD1 may play important ro les in the early stages of aggregate formation with other structural features of the aggregate providing additi onal stability. Furthermore, we found that amino acids in -strands 6 and 7 in human SOD1 are important for aggregation and that species-specific interactions between th ese two regions are important for enhancing aggregation. In our animal model of ALS, our findings demonstrate that the accumulation of

PAGE 13

13 disulfide cross-linked mutant protein is co-incident with the accumulation of detergent-insoluble aggregates of mutant protein, with both of these events occurr ing well after the appearance of multiple pathologic abnormalities but concurrent with the onset of symptoms. Together, these studies demonstrate that aggreg ates are composed of globally misfolded, immature SOD1 protein that contribute to the conversion of disease phe notype. We suggest that alternative forms of misfolded SOD1 proteins impart toxi city early in the disease course.

PAGE 14

14 CHAPTER 1 INTRODUCTION Amyotrophic Lateral Sclerosis Am yotrophic lateral sclerosis (ALS) is a pr ogressive neurodegener ative disease that selectively targets motor neurons (1). Motor neur on death occurs in the motor cortex, brainstem, and spinal cord, primarily affecting the corticobulb ar and corticospinal tracts. Atrophy in these motor tracts results in the development of mu scle weakness, fasiculations, atrophy, dysarthria, dysphagia, and respiratory failure (2). Respiratory failure and de ath occurs three to five years after onset of ALS symptoms (1). The current therapy for all forms of ALS is riluzole (3), which blocks glutamate receptors. However, riluzole does little to extend the surv ival of ALS patients. The majority of the cases of ALS have no appare nt cause, termed sporadic ALS, while 10% of the cases are genetically linked, termed familial ALS (FALS). Genetics of FALS Eight FALS loci and six ALS-related genes have been identifie d and linked to ALS. Four for ms of the disease are inherited by autoso mal dominance and have clinical features indistinguishable from sporadic ALS: ALS1, AL S3, ALS6, and ALS7. In ALS1, mutations were identified at chromosome 21q22.21 in the gene encoding superoxide dismutase 1 (SOD1) protein, which is a superoxide scavenging enzyme (4). ALS3 involves mutations occurring in chromosome 18q21 at an unknown gene (5). ALS6 is associated with mutations in the fused in sarcoma (FUS) gene located at chromosome 16q12, which is involved in the regulation of transcription and RNA splicing (6). ALS 7, located at chromosome 20ptelp13, has an unidentified gene associated with the illness (7, 8). Recently, muta tions have been identified in TAR DNA binding protein (TARDBP) at chromosome 1p36; howev er, to date, only a single family has been described (9). Four forms of FALS are inherited by autosomal dominance and

PAGE 15

15 have clinical features that di ffer from sporadic ALS: ALS-FT D, ALS with Parkinsonism and dementia, Progressive lower motor neuron dise ase, and ALS8. ALS-FTD (frontotemporal dementia) results from mutations that occu r in chromosome 9q21q22 (10). ALS with Parkinsonism and dementia is associated with mutations at chromosome 17q21.1 (11) in a gene that encodes for microtubule associated pr otein tau (MAPT), which functions to stabilize microtubules and promote their assembly (12). Progressive lower motor neuron (LMN) disease is associated with chromosome 2q13 in the p150 subunit of the dynactin 1 (DCTN1) gene (13), which binds to microtubules and cytoplasmic dynein during vesicle transport (14). ALS8 results from mutations at chromosome 20q13.3 (15, 16) in the vesicle associated membrane protein B (VAPB) gene, which is a membrane protein that associates with mi crotubules (17). The clinical features of this form of ALS are marked by lower motor neuron symptoms, postural tremor, cramps, and fasiculations. ALS4 is a juvenile onset, autosomal dominant form of ALS that is associated with mutations at chromosome 9q34 in the SETX gene (18, 19). SETX is homologous to genes that function in RNA processing. Two forms of the disease are juvenile onset and autosomal recessive: ALS2 and ALS5. ALS2 results from mutations in the ALS2 gene at chromosome 2q33, which encodes for the Alsin protein and functions as a guanine exchange factor for Ran, Rho, and Rab GTPases (20, 21). ALS2 has a slow di sease progression and includes primary lateral sclerosis. ALS5 is associated with mutations at chromosome 15q15.1q21.1 (22) and has no pseudobulbar signs with slow disease progression. Together, the variability in gene loci and re sulting disease phenotypes have l ead some in the field to propose that ALS is a spectrum disease.

PAGE 16

16 SOD1-Linked FALS Genetics of SOD1-Linked FALS Approxim ately 20% of the FALS cases are caused by dominantly inherited mutations in Cu-Zn superoxide dismutase (SOD1) (Table 1-1) (4 ). SOD1-linked FALS is characterized as an autosomal dominant disorder, meaning that a sing le copy of the mutant SOD1 gene is sufficient to cause the ALS phenotype, and one copy of the wild-type gene is present in every cell. However, as more SOD1-linked FALS families are identified, it appears that mutations in SOD1 can be grouped into three types: autosomal do minant with complete penetrance, autosomal dominant with incomplete penetrance, and rece ssive. SOD1 mutations that produce complete penetrance are those in which all mutant gene carriers develop disease in an age dependent manner: A4V, G37R, L38V, G41S, H43R, D76V L84F, L84V, N86K, E100G, D101H, I104F, G108V, C111Y, I112M, G114A, L126X, G127X, G141E, L144F, V148G, V148I (Table 1-1). SOD1 mutations that produce incomplete penetran ce are those in which not every carrier of the mutant gene develops the disease: A4T, L8Q, V14G, G16S, N19S, E21G, N65S, G72S, D76Y, N86S, A89V, D90A, G93S, A 95T, D101N, S105T, I113T, V118L, V118KTGPX, L126S, N139H (Table 1-1). A small subset of mutations are recessively inherited and require two copies of the mutant gene for disease to manifest: D 90A and D96N (Table 1-1) (P. Andersen, personal communication). Disease pathology and progression of SOD1-linke d FALS is similar to sporadic ALS. Mutations in SOD1 produce a spec trum of disease variants in AL S, ranging in the age of disease onset, the location of first symptoms, the rate of progression, and the length of disease duration. A subset of mutations result in a more aggressive form of the disease, in which patients die within 3 years of onset: A4T, A4V, C6F, C6 G, V7E, L8Q, G10V, G4 1S, H43R, H48Q, D90V, G93A, D101G, D101H, D101Y, L106V, I112T, R115G, D125H, S134N, V148I, V148G (Table

PAGE 17

17 1-1). A second set of mutations result in longer than average survival afte r disease onset: K12R, E21G, G41D, H46R, N65S, D76V, A89V, D90A, G93C, G93D, G93S, G93V, E100K, I104F, I113F, L126S, L144F, I149T (Table 1-1). Nevert heless, there are many mutations that produce variable survival times in patients: E21G G37R, L38V, D76Y, L84F, L84V, N86S, D90A, G93R, I113T, L144S (Table 1-1) (P Andersen, personal communication). More than 100 mutations in SOD1 have been identified in FALS cases (Table 1-1). The vast majority are point mutations located within -strands, a prominent structural element of the protein (23). An additional subset of SOD1 mutations includes frameshift mutations and truncation mutations that occur near the carboxy l terminus of the protein and cause early termination of the protein. The effects of th e mutations on normal enzyme activity, protein turnover, and folding vary considerab ly (24-26). In cell culture and in vitro models, enzyme activity ranges from undetectable to near normal (24, 27-30); most mutati ons accelerate the rate of protein turnover (24, 29); and many mutations increase the susceptibility of SOD1 to disulfide reduction (31). Because some SOD1 mutants retain high activity (29) and because the targeted deletion of SOD1 in mice does not induce ALS-like symptoms (32), mutant SOD1 is proposed to cause disease by the acquisiti on of toxic properties. This work focuses on SOD1-linked FALS as th is genetically linked form of the disease most closely mimics disease progr ession and symptoms of classic, sporadic ALS. Thus, insights into disease progression and disease mechanisms may be translatable to the sporadic form of the disease. SOD1 Protein SOD1 is a m etalloenzyme responsible for me tabolizing oxygen radicals that are produced during normal cellular respiration (33). SOD1 converts oxygen radicals into hydrogen peroxide

PAGE 18

18 and oxygen. SOD1 is ubiquitously expressed in all ce ll types. Within the cell, SOD1 is primarily located in the cytosol but ha s been found at lower levels in nuclei, peroxisomes, and mitochondria (34-36). The active enzyme is a homodimer of two 153 amino acid subunits. Each subunit binds one atom of copper in the active site (H46, H48, H63, and H120) and one atom of zinc in the zinc loop (H63, H71, H80, and D83), wh ich provides structural stability (37). Each normally folded SOD1 subunit is characterized by a -barrel containing eight anti-parallel strands, an electrostatic loop that directs the substrate into the active site, and an intramolecular disulfide bond between cysteine 57 and cysteine 146 (38). Togeth er, these structural features produce an extremely stable protein, retaining its structure in 1% SDS and 8M urea (39).

PAGE 19

19 Table 1-1. SOD1-linked familial ALS mutations SOD1 mutant Penetrance*Disease course* Activity Aggregation Reference A4S N/A N/A N/A N/A (40) A4T Incomplete Aggressive N/A ++++ (41) # A4V Complete Aggressive Normal +++ (42-44) V5L N/A N/A N/A N/A C6F N/A Aggressive N/A +++ (45, 46) C6G N/A Aggressive N/A +++ (46, 47) V7E N/A Aggressive Normal N/A (48, 49) L8Q Incomplete Aggressive Normal N/A (49, 50) L8V N/A N/A N/A N/A (51) G10V N/A Aggressive N/A N/A (52) G12R N/A Slow N/A N/A (53) V14G Incomplete N/A N/A +++ (54) # V14M N/A N/A N/A N/A (55) G16A N/A N/A N/A N/A (51) G16S Incomplete N/A N/A N/A (56) N19S Incomplete N/A N/A N/A (51) F20C N/A N/A N/A N/A (51) E21G Incomplete Variable N/A ++ (50) # E21K N/A N/A N/A ++ (57) Q22L N/A N/A N/A N/A (51) V29A N/A N/A N/A N/A V29insA N/A N/A N/A N/A (58) G37R Complete Variable Normal ++ (4, 43, 49) L38R N/A N/A N/A N/A (59) L38V Complete Variable Normal N/A (4, 44, 49) G41D N/A Slow Normal ++ (4, 43, 49) G41S Complete Aggressive Normal +++ (4, 44) # H43R Complete Aggressive Normal +++ (4, 44) # F45C N/A N/A N/A N/A (60) H46R Complete Slow Reduced ++ (43, 44, 61) V47A N/A N/A N/A N/A (51) V47F N/A N/A N/A N/A (51) H48Q N/A Aggressive Reduced ++ (43, 49, 62) H48R N/A N/A N/A N/A (51) E49K N/A N/A N/A N/A (59) T54R N/A N/A N/A N/A (51) S59I N/A N/A N/A N/A (51) N65S Incomplete Slow N/A N/A (63) P66A N/A N/A N/A N/A L67R N/A N/A N/A N/A (59) G72S Incomplete N/A Normal N/A (44, 64) G72C N/A N/A N/A N/A (65)

PAGE 20

20 Table 1-1. Continued SOD1 mutant Penetrance*Disease course* Activity Aggregation Reference D76V Complete Slow N/A N/A (66) D76Y Incomplete Variable Normal N/A (44, 54) H80R N/A N/A N/A + (67) L84F Complete Variable N/A N/A (64) L84V Complete Variable Normal +++ (49, 55) # G85R N/A Variable Reduced +++ (4, 44) N86I N/A N/A N/A N/A N86D N/A N/A N/A N/A (68) N86K Complete N/A N/A N/A (69) N86S Incomplete N/A Normal N/A (49, 70) V87A N/A N/A N/A N/A (51) V87del N/A N/A N/A N/A (51) T88delACTGCTGAC N/A N/A N/A N/A (51) A89V Incomplete Slow N/A N/A (71) A89T N/A N/A N/A N/A (51) D90A Incomplete Variable Normal +++ (44, 72) # D90A Recessive Slow Normal +++ (44, 72) # D90V N/A Aggressive N/A N/A (73) G93A N/A Aggressive Normal +++ (4, 44) # G93C N/A Slow Normal +++ (4, 43, 49) G93D N/A Slow N/A ++ (74)(49) # G93R N/A Variable Normal ++++ (75)(49) # G93S Incomplete Slow N/A +++ # G93V N/A Slow Normal +++ (76) # A95T Incomplete N/A N/A N/A (60) A95V N/A N/A N/A N/A (51) D96N Recessive N/A N/A N/A (77) D96V N/A N/A N/A N/A V97M N/A N/A N/A N/A (51) I99V N/A N/A N/A N/A E100G Complete Variable Normal +++ (4, 49) # E100K N/A Slow Normal ++++ (44) # D101G N/A Aggressive Normal ++++ (78) (49) # D101H Complete Aggre ssive N/A N/A (79) D101N Incomplete Aggressive Normal + (44, 80) # D101Y N/A Aggressive N/A N/A (81) I104F Complete Slow N/A N/A (82) I104del N/A N/A N/A N/A (51) S105L N/A N/A N/A N/A (51) S105delTCACTC Incomplete N/A N/A N/A (51) L106V N/A Aggressive N/A N/A (42) G108V Complete N/A N/A N/A (83) C111Y Complete N/A N/A ++ (46, 84)

PAGE 21

21 Table 1-1. Continued SOD1 mutant Penetrance*Disease course* Activity Aggregation Reference I112M Complete N/A N/A N/A (63) I112T N/A Aggressi ve N/A N/A (74) I113F N/A Slow N/A N/A (51) I113T Incomplete Variable Normal ++ (4, 43, 49) G114A Complete N/A N/A N/A (51) R115G N/A Aggressive Normal N/A (49, 85) V118L Incomplete N/A N/A N/A (51) V118insAAAAC N/A N/ A N/A N/A (86) V118KTGPX Incomplete N/A N/A N/A D124G N/A N/A N/A N/A (51) D124V N/A N/A Reduced +++ (76) D125H N/A Aggressive Reduced + (44, 62) # D125delTT N/A N/A N/A N/A (87) L126S Incomplete Slow Reduced (88) L126stop N/A N/A N/A +++ (50) L126insTT Complete N/A Reduced ++++ (89) G127insTGGG N/A N/ A N/A N/A (54) Q132insTT N/A N/A N/A N/A (83) Q132del N/A N/A N/A N/A (76) E133V N/A N/A N/A N/A E133del N/A N/A N/A + (76) E133insTT N/A N/A N/A N/A (83) S134N N/A Aggressive Reduced + (44, 90) # N139H Incomplete N/A N/A N/A (91) N139K N/A N/A Reduced +++ (89) A140G N/A N/A N/A N/A (92) G141E Complete N/A N/A N/A (79) G141stop N/A N/A N/A N/A (51) L144F Complete Slow Normal ++ (49, 93) # L144S N/A Variable N/A ++ (94) # A145G N/A N/A N/A N/A A145T N/A N/A N/A N/A (94) C146R N/A N/A N/A N/A (46, 50) G147R N/A N/A N/A N/A (51) V148G Complete Aggressive Normal +++ (49, 93) # V148I Complete Aggressive N/A + (82) I149T N/A Slow Normal N/A (49, 89) I151S N/A N/A N/A N/A (51) I151T N/A N/A N/A N/A (95) + low (no aggregation in 24 hours), ++ modera te (<0.5), +++ high (0.7-1.7), ++++ extreme (>1.8) # Prudencio M, Hart PJ, Borchelt DR, Andersen PM. Submitted. PM Andersen, personal communication http://alsod.iop.kcl.ac.uk/Als/index.aspx

PAGE 22

22 Toxic Property of Mutant SOD1 Two hypotheses have been proposed to explai n the toxic property that m utant SOD1 acquires: the oxidative damage hypothesis and the oligomeriz ation hypothesis. The oxidative damage hypothesis suggests that mutant SOD1 aberrantly produces hydrogen peroxide and peroxynitrite, resulting in damage to the cell (96). Under normal conditions, wild-type SOD1 has a very low propensity to produce these t oxic side products: wild -type SOD1 may undergo peroxidation through interaction with hydrogen peroxide, produc ing toxic hydroxyl radicals (30, 97, 98), or wild-type SOD1 may react with peroxyn itrates, which result in nitronium ions that can covalently modify tyrosines (99-101). Unde r the oxidative stress hypot hesis, mutations in SOD1 may increase the propensity to form these t oxic side products. The caveat to the oxidative damage hypothesis is that it re quires mutant SOD1 to bind c opper, an essential element for SOD1 activity. A subset of FALS-linked SOD1 mu tants are classified as metal binding mutants due to defects at or near me tal binding sites (26, 27, 102, 103). Metal binding mutants have no SOD1 activity and do not produce hydrogen peroxide or peroxynitrite (26). In transgenic mice modeling FALS, mice overexpressing SOD1 wi th two (H46R/H48Q) or four (H46R/H48Q/H H63G/H120G) of the copper binding sites mutated to FALS residues develop paralysis (43, 104). Thus, mouse studies have demonstr ated that SOD1 activity and the correct bindi ng of copper is not required for FALS phenotype. Alternatively, the oligomerization hypothesi s proposes that misfol ded SOD1 proteins interact to form increasingly high-molecular-w eight oligomers (105). Mutations in SOD1 destabilize the native state protein and possibl y promote aggregation by diminishing metal binding and altering the secondary, tertiary, or quaternary struct ures (106-111). Tissues from FALS patients (112) and from SOD1 transgenic mice (113) contain SOD1 positive inclusions in the brainstem and spinal cord. Aggregated forms of mutant SOD1 can also be detected by

PAGE 23

23 detergent-insolubility or size exclusion filtration (23, 43, 114, 115). Some suggest that SOD1 aggregate-mediated toxicity occurs by sequester ing cellular proteins a nd heat shock proteins (116); the clogging of proteasomes by indigestible protein fragments (117); or the localization of SOD1 aggregates to vi tal organelles (118). Because SOD1 mutants have a diverse range of biophysical characteristics and produce a similar ALS phenotype, we are searching for a co mmon mechanism of toxicity. We propose that mutations in SOD1 predispose the protein to sel f-oligomerize and that SOD1 aggregates play a significant role in toxicity and disease course. Models of FALS Mouse Models of FALS Mouse m odels of FALS provide insight into the role of mutant SOD1 in motor neuron death, the progression of the disease, and the eff ectiveness of possible drug targets. All mouse models of FALS that overexpress mutant human SOD1 share a similar phenotype of motor neuron loss, muscle wasting, and hindlimb paralysi s. Mouse models of FALS differ from the human ALS phenotype in that th ere is no upper motor neuron pathology. In these mouse models, human genomic SOD1 is expressed unde r the control of the human SOD1 promoter, which produces ubiquitous expression of the hum an protein. FALS-linke d SOD1 mutants that have been expressed in mice include: G93A (119), G37R (120), G85R (113), L126Z (115), G86R (121), D90A (122), and Gins127TGGG (123). Mouse models have also been developed to express forms of mutant SOD1 that co mbine disease-linked mutations and include experimental mutations to study the mechanis m of SOD1 toxicity: H46R/H48Q (104) and H46R/H48Q/H63G/H120G (43). There is some variability in survival of different mutant mice, which may be due to expression levels of the transgene.

PAGE 24

24 FALS-linked SOD1 mutant mice are predom inantly characterized by loss of motor neurons, inclusions, and neuropils in the brainstem and spinal cord (119, 124, 125). Prominent neuronal loss occurs in the ventral horn of the spinal cord (125). Surviving, atrophic motor neurons exhibit hyaline filamentous, neurofilament-r ich, spheroidal inclusions (125). Within the ventral horn, SOD1 positive inclusions are predom inantly detected in neurons (125). Protein inclusions are primarily positive for SOD1 and ub iquitin (125-127). However, some inclusions have varied reactivity with Hsp40, Hsp60, Hsp70, Hsp90 (125), and small h eat shock proteins (Hsp25 and B crystallin) (43, 115). Filamentous st ructures bound to SOD1 have morphology similar to amyloid fibrils (128-130). These stru ctures also bind Congo red and Thioflavin T and S (43, 104, 108), markers for extensive -sheet stacking. Wild-type SOD1 transgenic mice do not develop neuronal loss or inclusions (113, 119). In all SOD1 transgenic mice, the appearan ce of symptoms is associated with an accumulation of sedimentable structures that are detergent-insoluble, which is diagnostic for protein aggregation (43, 104, 114, 115, 117, 123, 131, 132). Detergent-insoluble SOD1 protein is exclusively located in the affected tissues (brainstem and spinal cord) (114). Because inclusion bodies differ in abunda nce and cause difficulty for co rrelation to disease in mouse models, we prefer to measure aggregation by detergent-insolubility of SOD1 protein. Alternative mouse models of SOD1-linked FALS Several groups of created a lternative mouse m odels of FA LS that produce varying in expression patterns in order to address key questi ons that remain in the field. Recently, mice were generated that express mutant SOD1 (G93A) under the control of the Thy1.2 promoter, which produces mutant SOD1 expression specifically in neuronal cells (133). Heterozygous Thy1.2-G93A mice do not develop the ALS phenotype; however, when these mice are bred for

PAGE 25

25 homozygousity or crossed to SOD1-wild-type mi ce, mice develop disease by 20 months with detectable aggregates of the mutant protein in affected tissue (133). Other groups have also demonstrated that breeding mutant SOD1 transg enic mice (ubiquitously expressing, traditional models) to wild-type SOD1 transgenic mice pr oduces accelerated disease with more robust levels of SOD1 aggregates accumulating in affected tissues (134). SOD1 toxicity is non-cell autonomous Several groups have generated evidence to suggest that m utant SOD1 toxicity affects multiple types of cells in the nervous system. Re cent studies have examined the role of mutant SOD1 toxicity in motor neurons (135), microglia (135), muscle (136), astrocytes (137), and Schwann cells (137) in SOD1 transgenic mice. Results from these studies demonstrate that the expression of mutant SOD1 in motor neurons is a primary determinant of disease onset and progression. However, the aberrant function of mutant SOD1 in other cells may also influence the progression of the disease. Chimeric mice th at selectively express mutant SOD1 in motor neurons develop ALS; however, chimeric mice that have a majority of non-mutant cells survive and mutant expressing cells do not appear sick (aggregates, inclusions or vacuoles) (138). Lowering levels of mutant SOD1 in microglia (135) or astrocytes (137) sl ows the later stages of disease progression, while the onset remains unc hanged. Lowering mutant SOD1 in muscle, however, has no effect on disease onset, progression, or surviv al (136). Furthermore, mice selectively expressing mutant SOD1 (G93A) in muscle tissue develop muscle abnormalities without developing ALS symptoms (139). Together, these studies suggest that damage to cells other than motor neurons, mediated by mutant SOD1, may be an impor tant feature of FALS disease progression. This feature of the disease provides an inte resting avenue for the future studies of aggregates and the factors eff ecting their formation in specific tissues.

PAGE 26

26 Cell Culture Model of Mutant SOD1 Aggregation Cell culture models provide a valuable t ool for screening for SOD1 aggregation, a hallm ark feature of FALS in human and mouse models. In cell aggregation assays, mutant human SOD1 cDNA is expressed in human immo rtalized cell lines (H EK293-FT), and protein fractions are isolated by high-speed centrifugation and non-ioni c detergent extraction to isolate the detergent-insoluble SOD1, representing aggregating SOD1 protein (43, 104, 114, 115). This cell aggregation assay accurately models aggrega tion in mutant SOD1 transgenic mice. All of the FALS mutants expressed in cell culture result in a heightened potential to form sedimentable structures that are detergent-insoluble (43, 104, 114, 115, 117, 123, 131, 132). As we continue to study more SOD1 mutants in cell cu lture, it is evident that all na tural FALS mutants have some propensity to aggregate; however, the rate of a ggregation can differ greatly (140). Furthermore, our group has shown that the rate of aggregation of individual mu tants in cell culture inversely correlates with disease duration of that mutant in FALS patients (Prudencio M, Hart PJ, Borchelt DR, Andersen P, submitted). Because we are searching for a common mech anism for toxicity, and all mutants possess some propensity to form detergent-insoluble, sedimentable species, we will focus on studying SOD1 aggregation in this cell culture model. The application of compounds to mutant SOD1expressing cells can provide a means for identif ying inhibitors of aggregation and for studying the factors effecting SOD1 aggregate formation.

PAGE 27

27 CHAPTER 2 A LIMITED ROLE FOR DISULFIDE CROSSL INKING IN THE AGGREGATION OF MUTANT SOD1 LINKED TO FAMILIAL AMYOTROPHIC LATERAL SCLEROSIS Introduction One proposed toxic property of m utant SOD1 is aggregation of SOD1 protein. Our group and others have found that eight FALS mutants (i n transgenic mouse mode ls) and thirteen FALS mutants (in cell culture models) exhibit a height ened potential to form aggregates (43, 104, 114, 115, 117, 123, 131, 132). These detergent-insoluble (aggregated) forms of SOD1 proteins contain molecules that are cr oss-linked by intermolecular disulfide bonds (23, 141, 142). Each SOD1 subunit possesses four cysteine resi dues at amino acids 6, 57, 111, and 146, and an intramolecular disulfide bond, betw een cysteine residues at 57 and 146, is found in the natively folded holo-enzyme (38). In symptomatic SOD1 transgenic mice, high-molecular-weight forms of SOD1 are visible by SDS-PAGE when detergen t-insoluble protein is electrophoresed in the absence of reducing agents (23, 141). These hi gh-molecular-weight, disulfide-linked, forms of SOD1 become more abundant as ALSlike symptoms progress (23, 141, 142). There has been considerable attention focuse d recently on the role of disulfide crosslinking in the aggregatio n of SOD1 (both FALS-mutant and wild-type protein) (23, 141-145). Initial studies of purified SOD1 in vitro suggest that all four cyst eine residues of SOD1 are capable of forming intermolecular disulfide bonds, with cysteines at 57 and 146 perhaps playing more important roles (141). However, in the past year, there have been se veral studies that used cell culture models or in vitro aggregation studies to examine the role of individual cysteine residues in mutant SOD1 aggregation (143-145). Collectively, these studies have focused This work is adapted from a manuscript published in The Journal of Biological Chemistry 283(20):13528-37 (2008).

PAGE 28

28 attention on cysteines 6 and 111 of SOD1 as play ing important roles in modulating mutant SOD1 aggregation with the putative mechanism invol ving formation of aberrant intermolecular disulfide bonds. In the present study, we have systematically exam ined the role of the four cysteine residues in SOD1 in the formation of aberrant disulfide bonds and protein aggregates. Our data suggest that although cysteines 6 and 111 may play critical roles in the formation of SOD1 aggregates, the mechanism of mutant protein aggregation does not appear to require extensive intermolecular disulfide linkages. Analysis of a series of experimental mutants le d us to conclude that cysteines 6 and 111 may modulate structural features of the protein, apart from disulfide linkages, that influence aggregate formation. Methods Construction of SOD1 Expression Vectors FALS and e xperimental mutations were creat ed in the cDNA of human SOD1 or mouse SOD1 using standard PCR strategies with oligonucleotides that introduce specific point mutations (Table 2-1). All mu tant cDNAs created in this manner were sequenced in their entirety to verify the presence of the desired mutations and the absence of undesired mutations. SOD1 mutants were expressed in the pEF-BOS vector (146). Tissue Culture Transfection and Transgenic Mice Hum an embryonic kidney 293 cells with a T antigen (HEK 293-FT) were cultured in 60mm poly-D lysine coated dishes. Upon reaching 95% confluency, cells we re transfected with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) and then harvested after 24 hours. The SOD1 transgenic mice have been previous ly characterized: the G93A variant [B6SJLTgN (SOD1-G93A)1Gur; Jackson Laboratory, Bar Harbor, ME, USA] (119), the G86R variant

PAGE 29

29 [FVB-Tg(Sod1-G86R)M1Jwg/J; Jackson Laboratory, Bar Harbor, ME, USA] (121), and the wild type (WT) protein (line 76) (147). Table 2-1. Aggregation pr opensity of SOD1 mutant s at cysteine residues See Table 1-1 SOD1 Aggregation Assay by Differential Extraction The procedu res used to assess SOD1 aggregation by different ial detergent extraction and centrifugation were similar to previous descriptio ns (43). Spinal cords were homogenized with a probe sonicator (Microson XL2000; Misonix, Fa rmingdale, NY 2W at 22.5 kHz) in 1:10 w/v of 1x TEN (10mM Tris, 1mM EDTA, 100mM NaCl). In cell culture experiments, cells were scraped from the culture dish in phosphate bu ffered saline (PBS) and centrifuged to pellet the cells before the pellets were resuspended in 100 l 1xTEN. Spinal cord homogenates and resuspended cell culture pellets were then mixe d with an equal volume of 2x extraction buffer 1 (10 mM Tris, 1 mM EDTA, 100 mM NaCl, 1% Nonidet P40, and 1x protease inhibitor cocktail) and sonicated as described above. The resul ting lysate was centrifuged for 5 minutes at >100,000g in a Beckman AirFuge to separate a nonionic detergent-insoluble pellet (P1) from SOD1 mutants Average aggregation propensity Production of aggregates WT 0.10 G85R 0.99 +++ C6G 0.92 +++ C6F 1.08 +++ C111Y 0.51 +++ C111S 0.07 C146R 1.3 +++ C6G/C111S 0.07 C6G/C111Y 0.06 C6F/C111S 0.45 ++ G85R/C111S 0.12 CSYR 0.84 +++ GCYR 0.04 GSCR 0.98 +++ GSYC 0.01 GSYR 0.06 FSYR 0.89 +++

PAGE 30

30 the supernate (S1). The supernate (S1) was decanted and saved for analysis. The pellet (P1) was resuspended in 200 l of 1x extraction buffer 2 (10 mM Tris, 1 mM EDTA, 100 mM NaCl, 0.5% Nonidet P40, and 1x protease inhibitor cocktail) and sonicated to resuspend. The extract was then centrifuged for 5 minutes at >100,000g in a B eckman AirFuge to separate a pellet (P2) from the supernate. The P2 fraction was resuspende d in buffer 3 (10 mM Tris, 1 mM EDTA, 100 mM NaCl, 0.5% Nonidet P40, 0.25% sodium dodecyl su lfate (SDS), 0.5% deox ycholic acid, and 1x protease inhibitor cocktail) by sonication and saved for analysis. Protein concentration was measured in S1 and P2 fractions by BCA met hod as described by the manufacturer (Pierce, Rockford, IL, USA) (Table 2-1). Assay for Disulfide Cross-Linked SOD1 In variations of this procedure, buffer 1 wa s m odified to include 100 mM iodoacetamide as noted in figure legends and the text. Additionally in one set of experiments (see Fig. 2-8) SDS was substituted for NP40 in all extraction buffers; 2x SDS buffer (10 mM Tris, 1 mM EDTA, 100 mM NaCl, 1% SDS, and 1x protease inhibitor cocktail) was substituted for buffer 1, and a 1x SDS buffer (10 mM Tris, 1 mM EDTA, 100 mM NaCl, 0.5% SDS, and 1x protease inhibitor cocktail) was substituted for buffers 2 and 3. Immunoblotting Standard sodium dodecyl sulfate-polyacryl amide gel electrophoresis (SDS-PAGE) was performed in 18% or 4-20% Tris-Glycine gels (Invitrogen, Carlsbad, CA, USA). Samples were boiled for 5 minutes in Laemmli sample buffer prior to electrophoresis (148). In some experiments, reducing agent (5% -mercaptoethanolME) was omitted from the sample buffer. Immunoblots were probed with rabbit polyclona l antibodies termed hSOD1 or m/hSOD1 at dilutions of 1:2500. The hSOD1 antibody is a pep tide antiserum that bind s to amino acids 24-36

PAGE 31

31 (not conserved between mouse a nd human SOD1 proteins), and m/h SOD1 antibody is a peptide antiserum that recognizes amino acids 124-136 (conserved between mouse and human SOD1 proteins) (24). Quantitative Analysis of Immunoblots Quantification of the SOD1 protein in deterg ent-insoluble and deterg ent-soluble fractions was performed by measuring the band intensity of SOD1 in each lane using a Fuji Imaging system (FUJIFILM Life Science, Stamford, CT USA). The untransfected control served as background. SOD1 aggregation propensity was a function of the ratio of the band intensity in the detergent-insoluble fraction to that of the detergent-soluble fraction. The mean and standard error of the mean (SEM) were ca lculated for the aggregation prope nsity of each sample in each experiment. A homoscedastic students T-te st was used to calculate significance. Results Our initial study focused on exam ining known FA LS mutations at cysteine residues to determine whether loss of any single cysteine resi due diminishes SOD1 aggregation. This initial study also provided a reference point to which to compare our next set of studies when these mutations were experimentally combined in re combinant SOD1 proteins. FALS-linked SOD1 cysteine mutants (C6G, C6F, C111Y, and C146R) were expressed in HEK293-FT cells, and the cell lysates were separated into detergent-soluble and detergent-insol uble fractions following previously established protoc ols (43). All four mutants produced detergen t-insoluble, sedimentable forms of mutant protein. In this assay, the detergent-soluble fractions (Fig. 2-1B) are representative of the steadystate levels of each mutant (an indication of the efficiency of transfection and level of expression). These experiments and those that follow include the G85R mutant as a positive control; this mutant exhibi ts anomalous migration in SDS-PAGE, migrating slightly faster than expected for its molecular weight. The aggregation potential, a measure

PAGE 32

32 comparing the band intensity of the detergent-insoluble fraction (Fig. 2-1A) to the detergentsoluble fraction (Fig. 2-1B), fo r each of the cysteine mutants was significantly di fferent from WT protein (Fig. 2-1C). The detergent-insoluble fractions from cells transfected with SOD1 -G85R often contained a form of the protein that migrat es at a size expected for a dimer. Interestingly, lysates from cells that expressed the C6F mutant (alone or in combination with other mutations see below) invariably contained forms of mutant SOD1 th at migrated at higher-molecular-weight; often a Figure 2-1. SOD1 aggregation of FALS cysteine mutants in transfected cells. Mutants were expressed in HEK293-FT cells and aggregate levels were determined as described in Methods. UT, untransfected cells. WT, cells transfected with vectors for WT SOD1. SOD1-G85R, a robustly aggreg ating FALS mutant, provides a positive control. SDSPAGE was performed in the pr esence of a reducing agent in an 18% Tris-Glycine gel. Immunoblots were probed with an antiserum specific for human SOD1 (hSOD1). A) Detergent-insoluble protein fraction (20 g). B) Detergent-soluble protein fraction (5 g). C) Relative aggregation is a function of the amount of SOD1 found in the pellet fraction as compared to the supernatant (see Methods). The gr aph represents the mean ( SEM error bars) of at l east three different experime nts. All FALS mutants were statistically different from WT SOD1: (*) p 0.001. C111S was not statistically different from WT SOD1. Open arrowhead ; dimer-sized SOD1 molecules. Closed arrowhead; monomeric SOD1 molecules.

PAGE 33

33 laddering effect was noted, indicative of the assembly of some type of rep eating structure. These structures persist despite boili ng in the presence of SDS and ME; thus, they are presumably either molecules that are covalently cross-li nked by mechanisms other than disulfide or are assemblies of mutant SOD1 that are resistant to denaturation. In the case of C6F, these multimer-like structures were also detected in detergent-soluble fractions. The origin and relative importance of these structures is presently unclear. In analyzing the data from the cysteine mutants, we noted variability in the amount of aggregated mutant protein produced from one experiment to the ne xt, possibly due to variation in transfection efficiency between experiments. For example, the immunoblot shown in Fig. 2-1A suggests that the C146R mutant produces less detergent-insoluble protein than the C6G or C6F mutants. However, when measurements of mu ltiple immunoblots from replicate experiments were compared, then the C146R mutant was not reproducibly different fr om the other cysteine mutants (Fig. 2-1C). The mutation of cysteine 111 to serine, a non-FALS mutation, did not produce a SOD1 protein that spontaneously aggregat es (Fig. 2-1A, lane 8; Fig. 2-1C). In the analysis of these mutants and in the studies that follow, the principal outcome measure was whether the amount of mutant protein in the detergent-insoluble fraction at ~24 hours posttransfection was statistically di fferent from WT SOD1, an indi cation of aggregation. More subtle differences in aggregation levels be tween mutants (e.g. C6F vs. C111Y, aggregation potentials of 1.08 and 0.51, respectively) are difficult to interpret and were not the focus of this study. The only measure of interest was that both mutants score as statistically different from wild-type protein. Overall, we can conclude that FALS mutations at cysteine residues 6, 111, or 146 induce, possibly to varying degrees, aggregation of the protein.

PAGE 34

34 To assess the roles of individual cysteine residues in mutant SOD1 aggregation, four SOD1 constructs were created, each of which cont ained one intact cysteine residue with the other three cysteines mutated (C 6G, C57S, C111Y, or C146R). In this experiment, the C57S mutation is an experimental mutation. When we began this work, no FALS mutations at cysteine 57 were known, and the C57S mutation serves only as a means to eliminate the cysteine residue. These mutants were expressed in HEK293-FT cells and cell lysates were assayed for Figure 2-2. Role of cysteines 6 and 111 in SOD1 aggregation. Mutants were expressed in HEK293-FT cells and aggregate levels were determined as described in Methods. SDS-PAGE was performed in the presence of a reducing agent in an 18% TrisGlycine gel. Immunoblots were probed with m/hSOD1 antiserum. A) Detergentinsoluble protein fraction (20 g). B) Detergent-solu ble protein fraction (5 g). C) Quantification of the ratio of SOD1 in pe llet fractions relative to supernatant was assessed as a measure of relative aggreg ation. The data repr esents the mean ( SEM error bars) of at least three different expe riments. Mutants where the ratios were statistically different from WT SOD1 we re marked by (*): p<0.0001. Some mutants were not statistically different from WT SOD1: GCYR, GSYC, GSYR. CSYR=C6/C57S/C111Y/C146R, GCYR=C6G/C57/C111Y/C146R, GSCR=C6G/C57S/C111/C146R, GSYC =C6G/C57S/C111Y/C146. Closed arrowhead; monomeric SOD1 molecules.

PAGE 35

35 aggregate formation as described above. Wh en cysteine 6 (C57S/C111Y/C146R; labeled CSYR) or cysteine 111 (C6G/C57S/C146R; labeled GSCR) were intact, detergent-insoluble mutant protein was consistently detected within 24 hours (Fig. 2-2A, lanes 5 and 7). In all mutants, a considerable fraction of the mutant protein was so luble in detergent (Fig. 2-2B), an indicator of the overall level of expression of the different mutants. From replicate experiments, the aggregation potentials for mutants with cysteine 6 or cysteine 111 intact were determined to be significantly different from WT SOD1 protein (Fig. 2-2C). Much less aggregated SOD1 was detected when cysteine 57 (C6G/C111Y /C146R; labeled GCYR) or cysteine 146 (C6G/C57S/C111Y; labeled GSYC) were the only cysteines intact (Fig. 2-2A, lanes 6 and 8). Neither of these later two mutant s differed from the aggregation potential for WT SOD1 (Fig. 22C). Notably, the two mutants that failed to form aggregates (C6G/C111Y/C146R and C6G/C57S/C111Y) produced detergent-soluble forms of these proteins at le vels equivalent to those that produced aggregates (Fig. 2-2B, lanes 6 and 8), indi cating that similar levels of expression were achieved. These re sults indicate that cysteine 6 and cysteine 111 play important roles in inducing the aggr egation of human SOD1. Because our studies indicate that proteins c ontaining only a single cysteine at 6 or 111 alone will rapidly aggregate, we sought to de termine if these residues are required for SOD1 aggregation. Two constructs were generated in which all four cysteine residues (C6G or F, C57S, C111Y, C146R) were mutated and expres sed in HEK293-FT cells for assessment of aggregation. One of these constructs in which cysteine 6 was mutated to glycine in the context of three other mutants (C6G /C57S/C111Y/C146R) failed to produ ce aggregates within 24 hours (Fig. 2-3A, lane 4). However, when cysteine 6 was mutated to phenylalanine (C6F/C57S/C111Y/C146R), the de tergent-insoluble fraction contained significant amounts of

PAGE 36

36 aggregated protein (Fig. 2-3A, lane 5). The aggregation potential for the C6F version of the four-cysteine variant (C6F/C57S/C111Y/C146R) was significantly different from WT SOD1 and roughly equivalent to the natural FALS mutant SOD1-G85R (Fig. 2-3C). The difference in the propensity of the two 4-Cys mutants to aggregate did not appear to be due to a lower expression of the C6G version, as the level of this mutant protein in the deterg ent-soluble fraction was similar to that of the C6F version (Fig. 2-3B, la nes 4 and 5). Thus, while cysteine 6 and 111 may play a role in modulating the rate or propensity of mutant human SOD1 to aggregate, the presence of a cysteine residue is not required for rapid aggregate formation. To further explore the role of cysteine re sidues 6 and 111 in aggregation, we produced a series of constructs in which these two cyst eine residues were mani pulated. Using the cell aggregation assay described a bove, we observed that when th e C111S mutation was added to C6G, C6F, or G85R, less detergent-insoluble protein was form ed as compared to the FALS mutants alone (Fig. 2-4A, lanes 4, 6, 8, and 9). The level of aggregat ed protein in cells transfected with C6G/C111S a nd G85R/C111S double mutants was not different from that of cells transfected with WT SOD1 (Fig. 2-4C). Next, the FALS mutation at cysteine 111 (C111Y) was combined with the FALS mu tant at cysteine 6 (C6G). Although each of these FALS mutants aggregated when mutated alone (see Fig. 2-1A), when combined the amount of detergent-insoluble SOD1 detected was not different from WT SOD1 (Fig. 2-4A and C). All combination mutants were detected in the deterg ent-soluble fraction at similar levels, indicating that the protein was stably e xpressed and only a fraction of th e total mutant SOD1 protein aggregates (Fig. 2-4B). These data provide evidence that cystei nes 6 and 111 play a role in the formation of SOD1 aggregates with the rate of aggregation slowing significantly when cysteine 111 was mutated to serine.

PAGE 37

37 Although the foregoing studies implicate a ro le for cysteines 6 and 111 in promoting aggregation, with the mutation of cysteine 111 to serine appearing to strongly suppress aggregation [also see (144) and (145)], it is noteworthy that mouse S OD1 naturally encodes serine at position 111 and possesses only 3 cysteine residues in total (equivalent to positions 6, 57, 146 in human protein). If the mutation of cy steine 111 to serine re duces aggregation, then the prediction would be that mouse SOD1 enc oding FALS mutations should not be prone to aggregate. Previous studies have established a transgenic mouse model of FALS by the expression of mouse SOD1 encoding the equivale nt to the human G85R mutation (121). These mice exhibit a rapidly progressing paralytic disorder. To determ ine the importance of cysteine Figure 2-3. Cysteine residues ar e not required for SOD1 aggregate formation. Mutants were expressed in HEK293-FT cells and aggregate levels were determined as described in Methods. SDS-PAGE was performed in the presence of a reducing agent in an 18% Tris-Glycine gel. Immunoblots were probed with m/hSOD1 antiserum. A) Detergent-insoluble protein fraction (20 g). B) Detergent-soluble protein fraction (5 g). C) Quantification of relative aggregation potential (mean ratio SEM). Mutants G85R and FSYR were significantly di fferent from the aggregation potential of WT SOD1: (*) p<0.0009. GSYR=C6G/C57S/C111Y/C146R. FSYR=C6G/C57S/C111Y/C146R. Open arro whead; monomeric SOD1 molecules.

PAGE 38

38 111 in vivo spinal cord tissue from symptomatic SOD1-G86R mice was assayed for SOD1 aggregation and found to accumulate detergent-inso luble SOD1 (Fig. 2-5A). Compared to mice expressing high levels of WT human SOD1, the symptomatic G 86R mice contained high levels of detergent-insoluble mouse SOD1. In the spin al cords of the WT human SOD1 mice, the vast majority of the SOD1 was soluble in detergen t, whereas in the G86R spinal cord tissue it appeared as though the majority of accumulated S OD1 was insoluble in de tergent (compare Figs. 2-5A and B). This finding demonstrates that encoding a serine at 111 does not block the aggregation of mouse SOD1. Because our assessment of the experimental mutation of human SOD1 (G85R/C111S) in which aggregation was slowed relied on th e HEK293-FT cell model, we examined the aggregation of mouse SOD1-G86R in cell culture. The WT and G86R variants of mouse SOD1 were compared to the G85R variant of hu man SOD1. Both mouse SOD1-G86R and human SOD1-G85R formed detergent-insoluble aggregates at similar propensities (Fig. 2-6A). Similar to human WT SOD1 (Fig. 2-1A), mouse WT SOD1 was not readily detected in the detergentinsoluble fraction (Fig. 2-6A). As expected, a ll three SOD1 variants were detected in the detergent-soluble fraction (Fig. 2-6B). These data indicate that amino acid sequence differences between human and mouse SOD1 modulate the requirement for cy steine at 111 in promoting aggregation. In the context of the mouse prot ein, a cysteine at 111 is not required and the presence of serine does not reduce aggregation.

PAGE 39

39 To determine the extent to which disulfide-linked multimers occur in lysates from HEK293-FT cells expressing human SOD1 with the A4V, G85R, and G93A natural FALS mutations, detergent extraction buffers were supp lemented with iodoacetamide, a non-reversible sulfhydryl-blocking agent, as dete rgent-soluble and dete rgent-insoluble fractions were isolated. When these fractions were anal yzed by SDS-PAGE in the absence of reducing agents, most of the detergent-insoluble protein migrated at the size expected for monomeric SOD1 (Fig. 2-7A, Figure 2-4. Role of cysteine 111 in mutant SOD1 aggregation Mutants were expressed in HEK293-FT cells and aggregate levels were determined as described in Methods. SDS-PAGE was performed in the presence of a reducing agent in an 18% TrisGlycine gel. Immunoblots were probed with m/hSOD1 antiserum. A) Detergentinsoluble protein fraction (20 g). B) Detergent-solu ble protein fraction (5 g). C) Relative aggregation ratios are graphed (mean ratio SEM) of at least three different experiments. All FALS mutants were statis tically different from wild-type SOD1: (*) p<0.0005. The ratios of insoluble to solubl e SOD1 for all FALS mutants that were combined with mutated C111 were not st atistically different from WT SOD1: G85R/C111S, C6F/C111S, C6G/C111S, C6 G/C111Y. (+) C6G was significantly different from C6G/C111S and from C6 G/C111Y p<0.009. C6F did not differ from C6F/C111S. (#) G85R was st atistically different from G85R/C111S p<0.004. Open arrowhead; dimer-sized SOD1 molecules. Closed arrowhead; monomeric SOD1 molecules.

PAGE 40

40 lanes 3, 4, and 5). To a variable degree, a portio n of the detergent-insoluble SOD1 in each cell lysate migrated at a higher than expected mol ecular weight (Fig. 2-7A). Distinct bands were detected at 40kDa, which are a size (in this gel system) that is consiste nt with disulfide-linked dimers of SOD1. When the mutants expressed in cell culture were prepared in the absence of iodoacetamide and examined by SDS-PAGE with reducing agent, nearly all mutant protein detected in the detergent-insoluble and deterg ent-soluble fractions migrated at a position expected of monomeric protein (F ig. 2-7C and D). Collectively, th ese data indicate that in the HEK293-FT cell culture model, a portion of the detergent-insolubl e protein forms intermolecular disulfide bonds, forming primarily structures dimeri c in size, but a large portion of the SOD1 in the detergent-insoluble fraction is not involved in in termolecular disulfide bonding. Figure 2-5. Symptomatic G86R mice form detergen t insoluble species in spinal cord tissue. Spinal cords from WT and G86R mice were extracted in buffers containing NP40 as described in Methods. SDS-PAGE was performed in the presence of a reducing agent in an 18% Tris-Glycine gel. Immunoblots were probed with m/hSOD1 antiserum. Experiments were replicated twice; a representative example is shown. A) Detergent-insolubl e protein fraction (20 g). B) Detergent-soluble protein fraction (5 g). In a previous study, Furukawa and colleagues (141) noted that extr action of tissues from symptomatic G93A mice in 0.5% SDS solubilized disulfide cross-linked forms of mutant SOD1. To determine whether 0.5% SDS solubilizes all aggregated forms of mutant SOD1 in these mice, we used our extraction/centrifugation protoc ol, substituting 0.5% SDS for 0.5% NP40 (see

PAGE 41

41 Methods). Extraction of spinal cord tissues from asymptomatic WT and symptomatic G93A transgenic mice in buffers with 0.5% SDS and iodoacetamide, which irreversibly blocks free sulfhydryl residues, completely so lubilized all mutant SOD1 that accumulated (Fig. 2-8A). No protein was detected in the high-speed pellet after centrifugation (the detergent-insoluble fraction) (Fig. 2-8A, middle panel). In SDS-PAGE of these samples in the absence of reducing agent, high-molecular-weight SOD1 proteins were detected in the G93A tissue but not WTSOD1 mice (Fig. 2-8A, left panel). When the detergent-soluble fraction was treated with reducing agent before SDS-PAGE, all of the disulfide bonds were br oken and only monomeric SOD1 was detected (Fig. 2-8A, right panel). As a positive co ntrol, the same spinal cord homogenates were extracted with buffers contai ning 0.5% NP40 and were r un in the presence of reducing agent (without iodoacetamide). As exp ected, the tissues from the symptomatic G93A mouse contained large amounts of mutant protein in the detergen t-insoluble fraction (Fig. 2-8B, right panel P2) with both WT and G93A SOD1 present in the detergent-soluble fractions (Fig. 2-8B, left panel S1). Together, these data indi cate that in a setting in which disulfide crosslinks are preserved, ionic detergents are sufficient to disrupt aggregate stru cture. Hence disulfide cross-linking alone is not responsible for the main tenance of detergent-inso luble structures that are distinguished by sedimentation upon ultracentrifugation. Discussion Our study sought to examine how disulfide bond s, formed between specific cysteine residues of mutant SOD1, may mediate the formation of detergent-insoluble, sedimentable structures (termed aggregates). From our findings, we conclude that disulfide-cross-linking of mutant SOD1 is not critical in aggregate formation nor do thes e cross-links appear to be a sufficient bonding force in maintain ing aggregate structure. By yet to be defined mechanisms,

PAGE 42

42 cysteine residues 6 and 111 were found to exert significant influen ce over mutant protein aggregation. A critical element in considering our work is how protein aggregat es are defined. In histological studies of tissues, aggregates are usually defi ned by the formation of discernable inclusion-body structures. However, in FALS mice, such structures are not necessarily prominent pathologic features ( 43, 149). Biochemically, aggregates isolated from tissues and cells are defined by several criteri a (for review see (150)). In general, aggregates are derived from assemblies of monomeric protein that at tain relatively high-molecular-weight (examples include filamentous aggregates as well as sma ller oligomeric structures). In many cases, pathologic protein aggregates resi st dissociation in detergent, and larger aggregates are of a size that allows for sedimentation upon centrifugation. In the FALS mice, previous work from our laboratory has shown that spinal cord tissues of symptomatic mice accumulate substantial levels of detergent-insoluble, sedimentable, mutant SOD1 (43, 115, 151). Thus, our study focused on the biochemistry of these sedi mentable aggregates. Figure 2-6. Mouse SOD1-G86R aggregates in cell culture. Mutants were expressed in HEK293FT cells and aggregate levels were determ ined as described in Methods. SDS-PAGE was performed in the presence of a reduci ng agent in an 18% Tr is-Glycine gel. Immunoblots were probed with m/hSOD1 antis erum. Experiments were replicated three times; a representative example is shown. A) Detergent-insoluble protein fraction (20 g). B) Detergent-solu ble protein fraction (5 g).

PAGE 43

43 Previous work, from our group and others, has established that detergent-insoluble aggregates of mutant SOD1 that accumulate in sp inal cords of FALS mice are extensively crosslinked by disulfide bonds (23, 141). Recent studies have investigated the role of disulfide bonding in SOD1 aggregation, and data reported to date have been interpreted as evidence that disulfide bond formation between mutant SOD1 prot eins could either initiate oligomerization or provide the major bonding force to stabilize aggreg ate structures (141, 144, 145). Indeed, one of Figure 2-7. Intermolecular disulfide bonding by S OD1 mutants in expressed cultured cells. FALS mutants (A4V, G85R, G93A) were expressed in HEK293-FT cells and detergent extracted in buffers containing i odoacetamide (see Methods). A) and B). SDS-PAGE was performed in the absence of reducing agent in a 4-20% Tris-Glycine gel. Immunoblots were probed with m/hSOD1 antiserum. Experiments were replicated four times; a representative ex ample is shown. A) Detergent-insoluble protein fraction (20 g). B) Detergent-solu ble protein fraction (5 g). C) and D) The same sets of samples as depicted in pane ls A and B were analyzed by SDS-PAGE in the presence of a reducing agent. C) Detergent-insoluble protein fraction (20 g). D) Detergent-soluble protein fraction (5 g). Open arrowhead; dimer-sized SOD1 molecules. Closed arrowhead; monomeric SOD1 molecules.

PAGE 44

44 the authors of the present study participated in a study suggesting a role for disulfide linkage in mutant SOD1 aggregation (23). However, through examination of multiple mutants and combinations of mutations, the pr esent study demonstrates that the role of cysteine residues in SOD1 in modulating the aggreg ation of mutant SOD1 is complex and likely to involve mechanisms other than disulfide cross-linking. First, we found that SOD1 encoding FALS-linked mutations at cyst eine residues 6, 111, or 146 aggregate when expressed in cell culture, indicating that the loss of any one of these cysteines does not block mutant protein aggregation. These fi ndings, regarding cysteine 6 and 111, are in agreement with a study by Cozzolino and colleagues (144). Second, we noted that cysteines 6 and 111 appear to play important ro les in promoting SOD1 aggregate formation as mutant proteins that possess either one of these residues retain the abilit y to rapidly aggregate while proteins that retained onl y cysteines 57 or 146 di d not rapidly aggregat e [also see (144)]. Similar to Cozzolino, we found that elimination of both cysteine 6 and 111 (most specifically by Figure 2-8. Disulfide bonds are not sufficient to maintain SOD1 a ggregate structure. A) Spinal cord tissue from asymptomatic WT and symptomatic G93A SOD1 transgenic mice was extracted in 0.5% SDS detergent and i odoacetamide. B) Standard extraction in NP40 and centrifugation of the same tissues util ized in panel A. Samples were run in the presence or absenc e of reducing agent ( ME), as noted, in 4-20% Tris-Glycine gels. Immunoblots were probed with m/hSOD1 antiserum. S1, de tergent-soluble (5 g). P2, detergent-insoluble (20 g).

PAGE 45

45 mutating cysteine 111 to serine) reduced or dramatically slowed aggregation. However, we now demonstrate that this outcome is unique to hum an SOD1. Mouse SOD1 encoding the equivalent of the human G85R mutation, and which naturally encodes serine at position 111, retains a high propensity to aggregate in cell and mouse models. We also identified experimental mutants that retained only one cysteine resi due while retaining hi gh propensity to aggregate; these mutants are incapable of forming disulfide-linked structur es larger than dimers (Fig. 2-9) and hence extensive cross-linking by disulf ide bonding cannot be critical for aggregate formation or stability. Finally, and most importantly, we iden tify an experimental SOD1 mutant lacking all cysteine residues (3 of 4 repla ced by FALS-linked mutations) that retained capacity to rapidly aggregate. From this body of evidence, we concl ude that if disulfide cross-linking has any role in promoting or stabilizing mutant SOD1 aggreg ation, such cross-linking is not required and may not be of major importance. Although one could argue that these experiment al mutants and the cell culture system do not reflect natural events, we demonstrate that aggregates of SOD1-G93A found in the spinal cords of symptomatic mice dissociate completely in 0.5% SDS, despite extensive preservation of disulfide cross-linking. Collectivel y, these data demonstrate that disulfide cross-linking is not responsible for the structure adopted by mutant SOD1 that accounts for sedimentation upon ultracentrifugation; a hallmark feature of SOD1 aggregates. Cysteines 6 and 111 Modulate Aggregation of Mutant SOD1 Sim ilar to a recent series of studies (141, 143-145), we find that cysteines 6 and 111 in human SOD1 play important role s in modulating mutant SOD1 a ggregation. However, the C6G, C6F, and C111Y mutants each rapidly formed a ggregates in our cell m odel, indicating that neither of these residues is indispensable. Alth ough we identified experimental mutants in which the need for cysteine at any of the natural pos itions is obviated, some of our experimental

PAGE 46

46 mutants suggested important roles fo r cysteines 6 and 111 in promoting aggregation. Combining C6G and C111Y mutations (C 6G/C111Y), or C6G and C6F with C111S (C6G/C111S and C6F/C111S), produced mutants with low propensities to aggregate. One of the most striking effects on aggregation was noted when the cystei ne 111 to serine mutation was combined with the G85R mutation in human S OD1, producing a double mutant that failed to aggregate within 24 hours; similar to WT SOD1. Cozzolino r ecently demonstrated that combining a C111S mutation with the A4V, G93A, and C146R FALS mutations produced similar reductions in aggregation (144). Collectively, these studies im plicate cysteine 111 as a potentially crucial residue in promoting aggregation. However, in mouse SOD1, position 111 is serine. Thus, the mouse SOD1-G86R animal, wh ich develops motor neuron disease marked by hindlimb paralysis, possesses an equivalent of our experimental human G 85R/C111S mutant. In contrast to human SOD1, mouse SOD1-G86R readily forms de tergent-insoluble, sedimentable species in both cell culture models and most importantly in the spinal cords of symptomatic mouse SOD1G86R mice. Hence, the effect of serine ve rsus cysteine at position 111 on SOD1 aggregation appears to be dependent upon the species from which the protein is derived (mouse and human SOD1 differ in sequence at 25 positions). Another key observation that lead s us to believe that structural features of cysteines 6 and 111 modulate aggregation, rather th an disulfide bonding, derives from the comparison of our two experimental mutants that alte r all four cysteine residues. The C6G/C57S/C111Y/C146R variant showed an aggregation potential no diffe rent from WT SOD1, whereas the C6F/ C57S/C111Y/C146R variant rapi dly aggregated similar to SOD1-G85R. The simplest explanation for this outcome is that the phe nylalanine at position 6 imparts an important structural feature to the protein that restores its ability to rapidly aggregate.

PAGE 47

47 Structural Features of Cysteine 111 In m ost species, the amino acid homologous to position 111 is serine rather than cysteine: humans and chickens are the only species in whic h position 111 is occupied by cysteine (23). In crystal structures of human SOD1, position 111 is located in the Greek-key loop near the dimer interface (38). Cysteine 111 is not obviously involved in crucial structural elements of the enzyme, but it is notable that serine is highly co nserved at this position with the two exceptions noted above. Recent studies have demonstrated that cysteine 111 may strongly bind metal ions, including copper (152). Cystei ne 111 may also bind other lig ands such as glutathione, thioredoxins, or other molecules that utilize di sulfide linkages to mediate binding (153, 154). A recent study by Fujiwara (155) reported that cystei ne 111 is a target for oxidative modification Figure 2-9. Intermolecular disulfide bonding by S OD1 mutants in expressed cultured cells. SOD1 mutants, each with only one cystei ne residue intact, were expressed in HEK293-FT cells and were treated with i odoacetamide to irreversibly block free sulfhydryl groups before extraction in dete rgent and high speed centrifugation. SDSPAGE was performed in the absence of reduc ing agent in a 4-20% Tris-Glycine gel. Immunoblots were probed with m/hSOD1 antis erum. Experiments were replicated four times; a representative example is s hown. The labels above lanes 5-9 use the same nomenclature as is used in Figure 22. A) Detergent-in soluble protein fraction (20 g). B) Detergent-soluble protein fraction (5 g). Small amounts of SOD1 migrating at a size consistent with dimeri c enzyme is visible when mutants encoded cysteine at position 6 or 111 (lanes 4, 5 and 7). Open arrowhead; dimer-sized SOD1 molecules. Closed arrowhead; monomeric SOD1 molecules. A nonspecific protein band that migrates just below the dimer-si zed SOD1 band (MW ~37 kDa) is visible in all lanes.

PAGE 48

48 and that some of the mutant SOD1 that accumula tes in pathologic structures in G93A mice is oxidatively modified at this pos ition. Notably, Cozzolino et al (144) reported that modulating the redox potential of NSC-34 cells influenced mutant SOD1 aggregation. It is possible that mechanisms involving a modification of the cysteine residues, part icularly 111, impart structural alterations in the protein to promote aggregation. Disulfide Bonding in Mutant SOD1 Aggregation Our analysis of m utant SOD1 aggregates fo rmed in both cell culture and the G93A mouse model indicates that aggregates of SOD1-G93A in symptomatic mice appear to be dissociated (converted from structures that sediment upon ultr acentrifugation to structures that do not) in 0.5% SDS, despite retaining significant disulfide cross-linking. Thus, whatever role disulfide bonding plays in aggregate formation, such bonds ar e not sufficient to ma intain the aggregate structure as a sedimentable entity. We were initially surprised to find that 0.5% SDS was sufficient to so lubilize (fra ctionate to high speed supernate) mutant SOD1. In prev ious studies, using a fi lter-trap assay, we had interpreted the retention of significan t fractions of mutant SOD1 in 0.22 m cellulose acetate filters as evidence that mutant SOD1 aggregates ar e SDS-resistant (114). In light of our current study, and the study by Furukawa and colleagues de scribed above (141), we now interpret the retention of mutant SOD1 in these filters as en trapment of disulfide cr oss-linked lattices that persist after solubilization in SDS rather than re tention of structures that are identical to the sedimentable species of mutant pr otein described here. In this la tter scenario, the disulfide crosslinks maintain an extended netw ork of intermolecular interactions, some of which would be large enough to be retained in the filter. Importantly, a disulfide cross-linked ne twork is distinct from

PAGE 49

49 the structures that are distingui shed by resistance to solubili zation in non-ionic detergent and sedimentation upon ultr acentrifugation. Prior studies of mutant SOD1 in tissues fr om G93A mice had identified high-molecularweight structures in denaturing SDS-PAGE that were interpreted to represent SDS-resistant complexes (117, 156, 157). The SDS-resistant highmolecular-weight SOD1 seen in SDS-PAGE generally ranges in size from dimers to molecu les of a mass no larger than relatively small oligomers (no more than 10-20 subunits). Such structures do not appear to possess sufficient mass to sediment in ultracentrifugation (see Fig. 2-9). We propose that these SDS-resistant structures that have been de scribed in SDS-PAGE may largely represent covalently-linked adducts to SOD1, formed by modifications su ch as ubiquitination or sumolyation, or may represent SOD1 proteins that are covalently cross-linked by mechanisms other than disulfide linkages (43). We have previously demonstrat ed high-molecular-weight forms of mutant SOD1 in denaturing SDS-PAGE analysis of detergent-insoluble, sedimentable st ructures (43). These high-molecular-weight, SDS-resistant, SOD1 prot eins are a relatively mi nor fraction of total detergent-insoluble protein. Notably, a recent st udy of mutant SOD1 that fractionates into the NP40 insoluble fraction demonstrated that the majority of the detergent-insoluble SOD1 display a mass consistent with an unmodified monomer; which is in some manner assembled into higher order sedimentable structures (151). Conclusions We conclude that disulfide bonding is likely to be of lim ited importan ce in maintaining the structure of aggregated forms of mutant SOD 1. Instead, it appears that the aggregates are stabilized by molecular interac tions that are dissociable by ionic detergent. We note that strand elements are a prominent feature of S OD1 and that these elements in the core -barrel

PAGE 50

50 structure of the protein are aligned adjacent to one another forming numerous hairpin folds (38). The stacking of -sheets is a characteristic feature of am yloids (158-162). In previous studies, we have noted the appearance of am yloid-like structures (stained with Thioflavin-S) in tissues of some but not all FALS mouse models (43, 115). Whether FALS mutant SOD1 assembles into amyloid-like structures in vivo is, however, unclear. Notably, in vitro amyloid-like structures that are detergent-sensitive have been demons trated for the H46R and S134N FALS mutants (163). We are convinced by our data and that of others (141, 143-145) that cysteines 6 and 111 participate in some crucial component of human SOD1 aggregation, however we believe that our study provides abundant evidence that the mechan ism does not require disulfide cross-linking. Instead, we propose that some other structural feature of these resi dues is critical in promoting aggregate formation. Whether modifications to cy steine residues, partic ularly cysteine 111, are critical in promoting human S OD1 aggregation is clearly a topic deserving further study. Similarly, a better understanding of the structural features of human SOD1 aggregates and the role cysteine residues play in maintena nce of such structures is required.

PAGE 51

51 CHAPTER 3 AGGREGATION MODULATING ELEMEN TS IN HUM AN SOD1 PROTEIN Introduction Normally folded SOD1 forms a homodimer of two 153-amino acid subunits. Each subunit contains eight -strands, copper bound in the activ e site, zinc binding site, an electrostatic loop, and an intr amolecular disulfide bond between cysteine 57 and cysteine 146 (38). The majority of FALS mutations are point mutations that occur predominantly in -strand regions. However, a subset of FALS muta tions produce reading frameshifts and early termination codons. The effects of FALS mutati ons on enzyme activity, tu rnover, and folding of the SOD1 protein vary considerably (24-26). Enzyme activity ranges from undetectable to normal (24, 27-30), and many mutants increase the su sceptibility of SOD1 to disulfide reduction (31). However, all FALS mutants that have be en studied in cell culture share a propensity to form detergent-insoluble SOD1 species in cel l culture, which are defined as aggregates (43)(Prudencio M, Hart PJ, Borchelt DR, Andersen P, submitted). It is possible that FALS mutations in SOD1 alter structur al aspects of the protein that enhance aggregation, and it is possible that specific regions with in the protein mediate misfolding. Considerable attention has been placed on the role of disulfide cross-linking in the formation of SOD1 aggregates (see Chapters 2 and 4) (23, 46, 141, 144). In symptomatic SOD1 transgenic mice, high-molecularweight, disulfide cross-linked forms of SOD1 are prominent in the detergent-insoluble protein fraction, which b ecome more abundant as mice approach disease endstage (23, 141)(Karch CM, Prudencio M, Wi nkler D, Hart PJ, Borchelt DR, In Press). However, in Chapter 2, we demonstrate that SO D1 aggregates are not stabilized by disulfide cross-linking alone (46). Studies in vitro and in cell culture suggest th at cysteine residues 6 and

PAGE 52

52 111 are important for modulating mutant SOD1 aggregation (46, 128, 144). However, my studies found that no single cyst eine residue is required for aggr egation (46)(see Chapter 2). In this study, we examined the role that spec ific regions in the prot ein might play in the formation of SOD1 aggregates using a mutagenesis approach and cell culture assay for aggregation. The FALS mutati on G85R and the mouse SOD1 G86R mutation have similar aggregation propensities. Howeve r, we discovered that chimeric human-mouse proteins with the FALS mutation (G85R) have strikingly different aggregation propensities. We exploited this finding to develop a system to identify sequences in human SOD1 that are important for the formation of aggregates. We found that am ino acids 42-50 and 109-123 in human SOD1 are important for aggregation and th at species-specific interaction is required between these two regions for aggregation to occur. Thus, it is li kely that aggregation occurs by global misfolding of the protein and that aggregation is enhan ced by several regions normally adjacent in the folded structure as opposed to a singl e residue in the human protein. Methods Construction of SOD1 Expression Vectors FALS and e xperimental muta tions were created in the cDNA of human SOD1, mouse SOD1, and chimeric proteins of human and mous e sequence (wild type (WT) chimeric proteins were generated by Genscript, Piscataway, NJ, USA) using standard PCR strategies with oligonucleotides that in troduce the specific point mutations (Table 3-1). All mutant cDNAs created in this manner were sequ enced in their entirety to verify the presence of the desired mutations and the absence of undesired mutations SOD1 mutants were expressed in the pEFBOS vector (146).

PAGE 53

53 Table 3-1. Aggregation propensit y of chimeric SOD1 mutants See Table 1-1 Tissue Culture Transient Transfection HEK 293-FT cells were cultured in 60-mm polyD lysine coated dishes. Upon reaching 95% confluency, cells were transfected with Lipofectam ine 2000 (Invitrogen, Carlsbad, CA, USA) and then harvested after 24 hours as previously described (46). SOD1 Aggregation Assay by Differential Extraction The procedu res used to assess SOD1 aggregation by different ial detergent extraction and centrifugation were similar to previous descriptions (43, 46). Protein concentration was measured in S1 and P2 fractions by BCA met hod as described by the manufacturer (Pierce, Rockford, IL, USA) (Table 3-1). SOD1 mutants Average aggregation propensity Production of aggregates H G85R 1.36 ++ M/H WT 0.00 M/H G85R 0.64 ++ M/H G85R/C111S 0.38 ++ H/M WT 0.01 H/M G85R 0.06 H/M G85R/S111C 0.41 ++ H/MG85R/S111C/G90D 0.32 ++ H/M G85R/S111C/R102S 0.33 ++ H/M G85R/ 90-102 0.13 H/M G85R/ 109-123 1.91 +++ H/M G85R/S111C/M117L 0.23 H/M G85R/S111C/E109D 0.05 H/M G85R/ 40-52 0.09 H/M G85R/ 1-40/ 90-102 0.08 H/M G85R/S111C/Q123A 0.07 H/M G85R/ 40-52/ 109-123 0.15 H A4V 0.99 +++ M A4V 0.29 ++ H/M A4V 0.35 ++ H/M A4V/ 109-123 1.82 +++ H V148G 0.79 +++ M V148G 0.04 H/M V148G 0.11 H/M V148G/ 109-123 0.50 ++

PAGE 54

54 Immunoblotting Standard sodium dodecyl sulfate-polyacryl amide gel electrophoresis (SDS-PAGE) was performed in 18% or 4-20% Tris-Glycine gels (Invitrogen, Carlsbad, CA, USA). Samples were boiled for 5 minutes in Laemmli sample buffer prio r to electrophoresis (148 ). Immunoblots were probed with a rabbit polyclonal antibody termed m/hSOD1 at dilutions of 1:2500. The m/h SOD1 antibody is a peptide antiserum that reco gnizes amino acids 124-136 (conserved between mouse and human SOD1 proteins) (24). Quantitative Analysis of Immunoblots Quantification of the SOD1 protein in deterg ent-insoluble and deterg ent-soluble fractions was performed by measuring the band intensity of SOD1 in each lane using a Fuji Imaging system (FUJIFILM Life Science, Stamford, CT USA). The untransfected control served as background. SOD1 aggregation propensity was a function of the ratio of the band intensity in the detergent-insoluble fraction to that of the detergent-soluble fraction. The mean and standard error of the mean (SEM) were ca lculated for the aggregation prope nsity of each sample in each experiment. A homoscedastic students T-te st was used to calculate significance. Figure 3-1. Human and mouse SOD1 differ at 25 amino acids. Alignment of human and mouse SOD1. Unconserved amino acids are marked in bold. Blue, human residues. Red, mouse residues. *, residues that form the di mer interface. Green, metal binding sites. Gold S, residues forming the intramolecular disulfide bond.

PAGE 55

55 Figure 3-2. Cysteine 111 plays a role in SOD1 aggregation. Mutant s were expressed in HEK293-FT cells, and aggregate levels were measured as described in Methods. UT, untransfected cells. WT, cells transfected with vectors containi ng WT SOD1 protein, which does not aggregate. G85R is a robustly aggregating FALS mutant. SDSPAGE was performed in the pr esence of reducing agent in an 18% Tris-Glycine gels. Immunoblots were probed with an antiser um recognizing a conserved region in mouse and human SOD1 (m/hSOD1). A) Detergent-insoluble protein fraction (20 g). B) Detergent-solu ble protein fraction (5 g). Image is representative of 3 repetitions of this experiment. Results SOD1 aggregation is a pathologic hallm ark of FALS. We have demonstrated that the cysteine encoded at amino acid 111 is important in modulating SOD1 aggregation; however, we also showed that we could create a SOD1 recombin ant protein in which all four cysteine residues were mutated (C6F/C57S/C111Y/C 146R) that still re tains a high propensity to aggregate (46)(see Chapter 2). These findings lead us to search for other regions of the SOD1 protein that enhance mutant SOD1 aggregation. To examin e regions of the protein that may enhance aggregation, SOD1 mutants were expressed in HEK293-FT cells, and the cell lysates were separated into detergent-insoluble and detergent-soluble fractions.

PAGE 56

56 The human FALS G85R mutation and the mous e G86R mutation have similar aggregation propensities in cell culture a nd both mutants cause hindlimb pa ralysis when the protein is overexpressed in mice (46, 113, 121). Human and m ouse SOD1 differ at 25 amino acids (Fig. 31). Human SOD1 encodes a cysteine at ami no acid 111. Alternatively, mouse SOD1 encodes a serine at amino acid 111. In human SOD1, when amino acid 85 encodes an arginine and amino acid 111 encodes a serine (G85R/C111S), the aggreg ation potential is si milar to WT protein (46)(see Chapter 2)(Fig. 3-2). However, in mouse SOD1, when amino acid 85 encodes an arginine and amino acid 111 encodes a serine, the protein aggregates robustly (46). Because a serine encoded at amino acid 111 has different effects in mouse SOD1 and in human SOD1, we exploited this finding to develop a system to identify sequences in human SOD1 that are important for the formation of a ggregates in FALS mutants. When the cysteine at amino acid 111 is substitu ted for a serine in the highly aggregating FALS mutant G85R, SOD1 aggregation is reduced to near wild-type (WT) SOD1 levels, which is considered non-aggregating (46)(see Chapter 2)(F ig. 3-2A). To determine if this effect is unique to the serine, cysteine 111 was mutated to an alanine in the context of the G85R mutation. Vectors containing SOD1 with intergenic mutatio ns at amino acids 85 (an FALS mutation) and 111 (an experimental mutation) were expresse d in HEK293-FT cells, and cell lysates were extracted in non-ionic detergent to isolate detergentinsoluble and deterg ent-soluble protein fractions. Note that a serine encoded at ami no acid 111 is highly conserved, with the exception of human and chicken SOD1 (23). An alanin e encoded at amino acid 111 is a structurally conservative mutation. The G85R/C111S and th e G85R/C111A mutants had similarly low levels of SOD1 in the detergent-insoluble fractio n (Fig. 3-2A). The amount of detergent-soluble material, representing stably folded protein, wa s similar in all mutants (Fig. 3-2B). Thus,

PAGE 57

57 substituting the cysteine residue at amino acid 111 in the context of a FALS mutation has a dramatic effect on aggregation. Figure 3-3. Differences in a ggregation propensity between m ouse and human SOD1 chimeric proteins. Mutants were e xpressed in HEK293-FT cells, and aggregate levels were measured as described in Methods. UT, unt ransfected cells. WT, cells transfected with vectors containing WT SOD1 protein, which does not aggregate. H/M, human SOD1 between amino acids 1-80 and mouse SOD1 between 81-153. M/H, mouse SOD1 between amino acids 1-80 and human SOD1 between 81-153. G85R is a robustly aggregating FALS mutant. SDSPAGE was performed in the presence of reducing agent in 18% Tris-G lycine gels. Immunoblots were probed with m/hSOD1 antiserum. A) Detergent-inso luble protein fraction (20 g). B) Detergent-soluble protein fraction (5 g). C) Relative aggregation pr opensity is a function of the amount of SOD1 found in the detergent-inso luble fraction compared to the detergentsoluble fraction. The graph represents the mean ( SEM (error bars)) of at least three different experiments. *, Statistically di fferent from all WT variants (H WT, H/M WT, M/H WT): p<0.05. Statistically different from H/M G85R: p <0.05. Statistically different from human G85R: p<0.05. Chimeric proteins were created to iden tify the residues that mediate human SOD1 aggregation. In the H/M SOD1 protein, ami no acids 1-80 were encoded as human SOD1 and amino acids 81-153 were encoded as mouse SOD1. In the M/H SOD1 protein, amino acids 1-80 were encoded as mouse SOD1 and amino aci ds 81-153 were encode d as human SOD1.

PAGE 58

58 Chimeric SOD1 proteins were expressed in HEK293-FT cells and extracted in non-ionic detergent. Chimeric proteins that encode WT sequence of human and mouse SOD1 (H/M WT and M/H WT) had an aggregation propensity that was similar to the human WT protein (Fig. 33C, lanes 2, 6, 8). Thus, the fusion of mouse a nd human sequence did not overtly alter protein folding. When the highly aggregating FALS mutant G85R was introduced into the H/M and M/H chimeric proteins, the aggregation propensities for both cons tructs were significantly less than human G85R (Fig. 3-3C). The H/M G85R mutant had very little detergent-insoluble protein (Fig. 3-3A) and did not di ffer from WT protein in its prope nsity to aggregate (Fig. 3-3C); however, the M/H G85R mutant formed detectable amounts of detergent-insoluble protein (Fig. 3-3A). The M/H G85R mutant had an aggregation propensity that was significantly different from WT protein (Fig. 3-3C). Thus, the M/H G85R mutant aggregates but at lower levels than the human protein. Despite the differences in aggregation propensity, th e levels of detergentsoluble SOD1 remained consistent (Fig. 3-3B). These chimeric proteins demonstrate that differences occur in human and mouse sequences that alter the propens ity of the FALS-G85R mutant to misfold and aggregate. Due to the apparent importance of amino acid 111 in human SOD1 aggregation, we used chimeric proteins to understand how this residue mediates aggregation. The chimeric protein H/M G85R encodes a serine at amino acid 111, which results in an aggregation propensity that is similar to human G85R/C111S (compare Figs. 32A and 3-3A). To more closely examine the role of cysteine 111 in SOD1 aggregation, amino acid 111 was altered to encode either cysteine or serine in the presence of the FALS-G85R mutant in the H/ M and M/H chimeric proteins (termed H/M G85R/S111C and M/H G85R/C111S). Encoding a cysteine at amino acid 111 in the H/M G85R mutant (H/M G85R /S111C) resulted in a slight increase in aggregation (Fig. 3-

PAGE 59

59 4A). A cysteine encoded at amino acid 111 prod uced a protein with an aggregation propensity that was significantly different from H/M G 85R; however, it was also different from human G85R (Fig. 3-4C). Thus, restoring a cystei ne to amino acid 111 enhanced mutant SOD1 aggregation; however, it was not su fficient to restore aggregation to the levels of human G85R. Alternatively, when amino acid 111 encoded a serine (the mouse residue) in the M/H G85R mutant (M/H G85R/C111S), the ag gregation levels of this mutant decreased only slightly (Fig. 3-4A). A serine encoded at amino acid 111 prod uced a protein with an aggregation propensity that was not significantly differe nt from M/H G85R (Fig. 3-4C). Thus, the presence of a cysteine at amino acid 111 in eith er chimeric protein does not ha ve robust restorative effects on aggregation. Together, thes e findings are consistent with evidence from our group, which suggests that cysteine 111 is important but not required for SOD1 aggregation (46). It is possible that cysteine 111 works in conc ert with other neighbori ng residues to mediate aggregation. To determine if residues between amino acids 81 and 153 are capable of enhancing SOD1 aggregation, we focused on the H/M SOD1 protein by humanizing sp ecific amino acids or blocks of sequence. Humanizing the regi on encoding amino acids 109-123 (E109D, S111C, M117L, and Q123A) in the H/M G 85R mutant (termed H/M G85R/ 109-123) resulted in a protein that robustly aggregated in cell culture (Fig. 3-5A). When amino acid 109-123 encoded human residues, the aggregation propensity was si milar to human G85R (Fig. 3-5C). Thus, this region appears to be important for aggregation. To determine if individual residues between amino acids 109-123 are responsible for enhancing aggregation, each residue was substituted for the corresponding human residue (eg: H/M G85R/E109D/S111C, H/M G85R/S111C/M117L, H/M G85R/S111C/Q123A). Each of these mutant s had an aggregation propensity similar to H/M G85R/S111C (Fig. 3-5C). Thus, individu al amino acids encoding human SOD1 do not

PAGE 60

60 enhance the effect of cystei ne 111 on aggregation. With the exception of H/M G85R/ 109-123, all mutants had similar levels of SOD1 in the detergent-soluble fraction (Fig. 3-5B). The H/M G85R/ 109-123 protein was less stable than other muta nts used in this study. Together, this evidence suggests that no single re sidue is responsible for aggreg ation but that the region of amino acids 109-123 may be important for enhancing human SOD1 aggregation. Figure 3-4. Amino acid 111 does not predict aggr egation alone. Mutants were expressed in HEK293-FT cells, and aggregate levels were measured as described in Methods. UT, untransfected cells. WT, cells transfected with vectors containi ng WT SOD1 protein, which does not aggregate. H/M, human S OD1 between amino acids 1-80 and mouse SOD1 between 81-153. G85R is a robustly aggregating FALS mutant. SDS-PAGE was performed in the presence of reduci ng agent in 18% Tris -Glycine gels. Immunoblots were probed with m/hSOD1 antiserum. A) Detergent-insoluble protein fraction (20 g). B) Detergent-solu ble protein fraction (5 g). C) Relative aggregation propensity is a function of th e amount of SOD1 f ound in the detergentinsoluble fraction compared to the detergen t-soluble fraction. The graph represents the mean ( SEM (error bars)) of at least three di fferent experiments. *, Statistically different from all WT variants (H WT, H/M WT, M/H WT): p<0.05. Statistically different from H/M G85R: p <0.05. Statistically differe nt from H G85R: p<0.05.

PAGE 61

61 Upon closer examination of the normally folded SOD1 protein, we conf irmed that residues 109-123, which reside in -strand 7, are in close proximity to the unconserved region of residues 42-50, which reside in -strand 6 (38). In the H/M G85R mutant, -strands 6 and 7 are not homologous and the highly aggregating H/M G85R/ 109-123 mutant restores homology in strands 6 and 7. To determine if -strands 6 and 7 undergo species-s pecific interactions that are required for aggregation, residues 42-50 were substituted for amino acids encoding mouse SOD1 (L42Q, E49Q, and F50Y) in th e context of the H/M G85R/ 109-123, creating a mutant in which -strands 6 and 7 differed in species (termed H/M G85R/ 42-50/ 109-123). Encoding mouse SOD1 at amino acids 42-50, when human SOD1 was encoded at amino acids 109-123 results in little detergent-insoluble SOD1 (F ig. 3-5A) and an aggregation pr opensity similar to WT (Fig. 35C). If -strands 6 and 7 undergo species-specific inte ractions for aggregati on to occur in FALS mutants, we predicted that s ubstituting residues 42-50 for amino acids encoding mouse SOD1 in the context of H/M G85R would enhan ce SOD1 aggregation (termed H/M G85R/ 42-50). H/M G85R/ 42-50, in which -strands 6 and 7 encoded mouse SOD1, had a low propensity to aggregate, similar to WT (Fig. 3-5C). The amount of detergent-soluble SOD1 was similar in all the mutants examined (Fig. 3-5B). Thus, our evidence suggests that SOD1 aggregation is enhanced when -strands 6 and 7 encode human SOD1 in the context of the mutant protein. To determine if amino acids 109-123 are univers ally important for aggregation of mutant SOD1, two FALS mutations located at the beginning of the protei n (A4V) and at the end of the protein (V148G) were introduced into the chimeric H/M SOD1 protein. H/M A4V produced significantly less aggregates than human A4V (Fi g. 3-6C); however, the aggregation propensity of H/M A4V was also different from WT protein. Humanizing amino acids 109-123 in the H/M A4V protein (termed H/M A4V/ 109-123) produced a robustly ag gregating mutant that was

PAGE 62

62 significantly different from H/M A4V; yet, simila r in its propensity to aggregate to human A4V (Fig. 3-6C). A similar pattern of aggregati on was observed with the V148G mutation, whereby the H/M variant aggregated at low levels and humanizing 109-123 enhanced aggregation to human levels (Fig. 3-6C). Thus, the effect of amino acids 109-123 on aggregation is shared among a variety of FALS mutants. Discussion This study was designed to identify regions in the hum an SOD1 protein that enhance mutant SOD1 aggregation. We utilized chimeric SOD1 proteins that combine human and mouse SOD1 sequence and found that amino acids 109-123 of human SOD1 are important for enhancing mutant SOD1 aggregation. Additionally, we provide evidence that species-specific interactions between -strand 6 (including amino acids 42-50) and -strand 7 (including amino acids 109-123) result in hi gh levels of human SOD1 aggregation. This effect is similar in FALS mutants that occur at the beginning, middle, and end of the protein. Thus, we suggest that aggregation occurs by global misfol ding of the protein and is enhanc ed by specific regions in the human protein. Our results demonstrate that regions 4250 and 109-123 are important for enhancing mutant human SOD1 aggregation (Fig. 3-7). These residues reside within -strands 6 and 7, respectively. In the normally folded SOD1 prot ein, these two elements neighbor one another. Normal structural components of -strand 6 and 7 are critical to structures that support aggregation. Residues within this region form the dimer interface (Fig. 3-1). Additionally, three of the four copper binding s ites are located between ami no acids 42-50 and 109-123. These copper binding sites are well conserved. Inte restingly, dimer formation and copper binding promote stability in the WT protein, which account s for a protein that is stable in 1% SDS and

PAGE 63

63 8M urea (39). Thus, it may be possible that FALS mutations disrupt these important interactions, destabilize the prot ein, and promote aggregation. Figure 3-5. Amino acids 42-50 and 109-123 in human SOD1 are important for aggregation. Mutants were expressed in HEK293-FT cells, and aggregate levels were measured as described in Methods. UT, untransfected ce lls. WT, cells transfected with vectors containing WT SOD1 protei n, which does not aggregate. H/M, human SOD1 between amino acids 1-80 and mouse SOD1 between 81-153. G85R is a robustly aggregating FALS mutant. SDS-PAGE was performed in the presence of reducing agent in 18% Tris-Glycine gels. Immunoblots were probed with m/hSOD1 antiserum. A) Detergent-inso luble protein fraction (20 g). B) Detergent-soluble protein fraction (5 g). C) Relative aggregation pr opensity is a function of the amount of SOD1 found in the detergent-inso luble fraction compared to the detergentsoluble fraction. The graph represents the mean ( SEM (error bars)) of at least three different experiments. *, Statistically di fferent from all WT variants (H WT, H/M WT, M/H WT): p<0.05. Statistically different from H/M G85R: p <0.05. Statistically different from human G85R: p<0.05. Statistically different from H/M G85R/ 109-123: p<0.05.

PAGE 64

64 Figure 3-6. Human amino acids 109-123 enhance aggregation in FALS mutants throughout the protein. Mutants were expre ssed in HEK293-FT cells, a nd aggregate levels were measured as described in Methods. UT, untransfected cells. A4V and V148G are robustly aggregating FALS mutants. SDSPAGE was performed in the presence of reducing agent in 18% Tris-G lycine gels. Immunoblots were probed with m/hSOD1 antiserum. A) Detergent-inso luble protein fraction (20 g). B) Detergent-soluble protein fraction (5 g). C) Relative aggregation pr opensity is a function of the amount of SOD1 found in the detergent-inso luble fraction compared to the detergentsoluble fraction. The graph represents the mean ( SEM (error bars)) of three different experiments. Statistically different from H A4V: p<0.05. Statistically different from H/M A4V/ 109-123: p <0.05. Statistically different from H V148G: p<0.05. Statistically different from H/M V148G/ 109-123: p<0.05. If strong interactions are required between -strands 6 and 7 for mutant human SOD1 aggregation to occur, we would predict that FALS mutants within th is region would disrupt interactions between -strands 6 and 7. Disrupted interactions between -strands 6 and 7 might then result in lower aggregation propensity. Because aggregation inve rsely correlates with disease duration in FALS patients (Prudencio M, Hart PJ, Borchelt DR, Andersen P, submitted), we would predict that FALS mutants occuri ng between amino acids 42-50 and 109-123 would

PAGE 65

65 produce a long disease duration. However, FA LS mutations between amino acids 42-50 and 109-123 are moderate to highly aggr egating in cell culture and have short disease duration (with 2 exceptions of long disease duration: G41D and H46R)(see also Table 1-1)(Prudencio M, Hart PJ, Borchelt DR, Andersen P, submitted). Thus, it is possible that a single mutation in one of these regions does not alter inter action or that tight binding is not required for aggregation. Figure 3-7. Chimeric proteins have differential aggregation pr opensities. Diagrams of the chimeric proteins used in this study are re presented on the left panel. Blue, human SOD1. Red, mouse SOD1. Black, highl y conserved between human and mouse (amino acids 124-153). Relative aggregation propensity is a function of the amount of SOD1 measured in the detergent-insoluble fraction compared to the detergentsoluble fraction. The graph represents the mean ( SEM (error bars)) of at least three different experiments.

PAGE 66

66 Our study identifies sequences in -strand 6 (residues 40-52) and -strand 7 (residues 109123) of SOD1 that appear to modulate aggrega tion of the human protein. Shaw and colleagues report similar findings, using hydrogen exchange a nd SOD1 peptides, that a portion of the dimer interface (residues 50-53) and a region containing -strand 7, loop VII, and a portion of -strand 8 (residues 117-144) are de stabilized (164). Together these studies suggest that mutant ions in SOD1 promote destabilization in the -strand regions to enhance SOD1 aggregate formation. However, the mechanisms by which these resi dues enhance aggregat ion are yet to be determined.

PAGE 67

67 CHAPTER 4 THE ROLE OF MUTANT SOD1 DISULFID E OXIDATI ON AND AGGREGATION IN THE PATHOGENESIS OF FAMILIAL ALS Introduction One proposed gain of toxic property is the pr opensity of SOD1 to m isfold and for these misfolded SOD1 proteins to interact to form increasingly high-molecular-weight oligomers (105). Mutations in SOD1 destabilize the nativ e state protein and possibly promote aggregation by diminishing metal binding and altering the sec ondary, tertiary, or quate rnary structures (106111). SOD1 immunoreactive inclusions are detected in tissues from FALS patients (112) and from SOD1 transgenic mice (113). Aggregated forms of mutant S OD1 can also be detected in SOD1 transgenic mice and cell culture based on dete rgent insolubility or size exclusion filtration (23, 43, 114, 115). All mouse models of FALS th at overexpress variants of mutant human SOD1 share a similar phenotype of motor neuron loss, muscle wasting, and hindlimb paralysis. FALS-linked SOD1 mutants that have been expressed in mi ce include: G93A (119), G37R (120), G85R (113), L126Z (115), G86R (121), D90A (123), Gins127T GGG (123), and H46R (166). Mouse models have also been developed to express forms of mutant SOD1 that combine disease-linked mutations and include experime ntal mutations to study the mechanism of SOD1 toxicity: H46R/H48Q (104) and H46R/H48Q/H63G/H120G ( 43). FALS-linked SOD1 mutant mice are predominantly characterized by the loss of motor ne urons and the presence of aggregates in the spinal cord and brainstem (119, 124, 125). In a ll SOD1 transgenic mice, the appearance of symptoms is associated with an accumulation of sedimentable structures that are detergentinsoluble, which is diagnostic for protein aggregation (43, 104, 114, 115, 117, 123, 131, 132). This work is adapted from a manuscript published in the Proceedings of the National Academy of Sciences In Press.

PAGE 68

68 Together, this evidence illustrates that SOD1 aggr egates are a prominent pathological feature in ALS mice and are possible mediators of the disease. Several groups have described the appearance of SOD1 positive inclusions (130, 167) and high-molecular-weight, detergent-inso luble SOD1 protein complexes (23, 117, 123, 127, 142) in SOD1 transgenic mice prior to the onset of paralysis. These studies have also documented an increase in the abundance of these sp ecies as the mice age. A number of recent studies have examined the role of aberrant intermolecular disulfide cross-linking in the formation and stabilization of mutant S OD1 aggregates (23, 46, 141, 144, 145)(see Chapter 2). Mutant protein isolated from transgenic mouse models form aberrant intermolecu lar disulfide bonds that generate extensively cross-linked high-molecular weight proteins (23, 46, 123, 141) (see Chapter 2). However, the role of disulfide bonding in ag gregate formation is unclear. Several studies have suggested that such cross-linking is cruc ial to aggregate formation (141, 143), whereas we recently demonstrated that, in cell culture mode ls, aggregated forms of mutant SOD1 can be generated in the absence of disulfide cross-linking (see Chapter 2) (46). In the present study, we have examined the so lubility and extent of disulfide cross-linking through the course of disease in four different mutant SOD1 mouse models (G37R, G93A, and H46R/H48Q, L126Z). The questions asked were the following: 1) Do disulfide-cross-linked forms of mutant SOD1 accumulate prior to the fo rmation of aggregates; 2) Is there a correlation between the formation of disulfide cross-linked SOD1 and the appearance of other pathologic or symptomatic features of disease; 3) What is th e status of the normal in tramolecular disulfide bond of SOD1 (C57 to C146) in aggregated forms of the protein. Using a highly sensitive detergent extraction assa y coupled to immunoblot analysis, we traced the appearance and abundance of detergent-insoluble and disulfide-cross-linked SOD1 species throughout the

PAGE 69

69 disease course of SOD1 transgenic mice Our findings demonstrate that the accumulation of disulfide cross-linked mutant protein is co-incident with the accumulation of detergent-insoluble aggregates of mutant protein, with both of these events occurr ing well after the appearance of multiple pathologic abnormalities but concurrent with the onset of symptoms. Though aggregates formed in vivo are extensively disulfide cross-linke d, complete reduction of disulfide cross-linkages does not dissociate SOD1 aggregates In both cell culture and the mouse models, we find that mutant protein lacking any type of disulfid e linkage, including its normal intramolecular disulfide bond, is the major component of th e detergent-insoluble SOD1 aggregates. Because formation of the intr amolecular disulfide bond is associated with maturation of SOD1, including acquisition of metal cofactors, it appears that mutant SOD1 that fails to mature may be disproportionately prone to aggregation. Methods Transgenic Mice The SOD1 t ransgenic mice used in this st udy have been previously characterized: the G93A variant [B6SJL-TgN (SOD1-G93A)1Gur; di sease onset 4-5 mo; Jackson Laboratory, Bar Harbor, ME, USA] (31), the G37R variant [lin e 29 (disease onset at 7-8 mo)] (147), the H46R/H48Q variant [line 139 (disease onset at 6-7 mo)] (104), the L126Z variant [line 45 (disease onset at 8-9 mo)] (115), and the wild type (WT) variant (line 76) (147). SOD1 Aggregation Assay by Differential Extraction The procedu res used to assess SOD1 aggreg ation by differential de tergent extraction and centrifugation in cell culture and mouse models were similar to previous descriptions (see Chapter 2) (43, 46). Spinal co rds are homogenized with a probe sonicator (Microson XL2000; Misonix, Farmingdale, NY 2W at 22.5 kHz) in 1:10 w/v of 1x TEN (10mM Tris pH 7.5, 1mM EDTA, 100mM NaCl). A crude supernate was is olated by centrifugation at 800g in an HS-4

PAGE 70

70 rotor. A fraction of the spinal cord homogenate (100 l) was then extrac ted in 0.5% NP40 as described previously (46). Assay for Disulfide Cross-Linked SOD1 In variations of the extraction, samples were extracted in th e presence of 100 mM iodoacetamide (as noted in the figure legends an d text). SDS was substituted for NP40 in all extraction buffers: 10x SDS buffer (10 mM Tris pH 7.5, 1 mM EDTA, 100 mM NaCl, 10% SDS, and 1x protease inhibitor cocktail) was substituted for buffer 1, and 1x SDS buffer (10 mM Tris, 1mM EDTA, 100 mM NaCl, 1% SDS, and 1x protease inhibitor cocktail) was substituted for buffers 2 and 3. Protein concentration was measured in both fractions by BCA method as described by the manufacturer (Pierce, Rockford, IL, USA). Immunoblotting Standard sodium dodecyl sulfate-polyacryl amide gel electrophoresis (SDS-PAGE) was performed in 18% Tris-Glycine gels (Invitrogen, Calsbad, CA, US A). Samples were boiled for 5 minutes in Laemmli sample buffer prior to elec trophoresis (148). Immunoblots were probed with rabbit polyclonal antibodies termed hSOD1 and m/hS OD1 at dilutions of 1:2500. The hSOD1 antiserum was raised against a synthetic peptide that is specific to human SOD1 (aa 2436) (113). The m/hSOD1 antiserum was raised against a synthetic peptide conserved between mouse and human SOD1 (aa 124-136) (24). Assay for Detection of Reduced and Oxidized SOD1 In som e experiments, the reducing agent (5% -mercaptoethanol [ ME]) was omitted from the sample buffer. In experiments requiring in -gel reduction, prior to electrotransfer to nitrocellulose membranes, gels were incubate d in transfer buffer in the presence of 2% ME for 10 minutes. Immunoblots were probed as described above.

PAGE 71

71 Quantitative Analysis of Immunoblots Quantification of SOD1 protein in detergent-insoluble and de tergent-soluble fractions was perform ed by measuring the band intensity of SOD1 in each lane using a Fuji imaging system (FUJIFILM, LifeScience, Stamford, CT, USA) (see chapter 2). Results Previous studies of m ice that model SOD1 -linked FALS have established that the brainstems and spinal cords of symptomatic mi ce contain relatively high levels of detergentinsoluble mutant SOD1 mice (46, 114, 117, 141, 142). A substantial fraction of the detergentinsoluble mutant SOD1 that accumulates in these tissues is aberrantly cross-linked via intermolecular disulfide linkages (46, 114, 117, 141, 142). In the present study, we determined the rela tive presence of disulfide cross-linked and detergent-insoluble mutant SOD1 through the course of disease in four models of SOD1-linked FALS (G93A, G37R, H46R/H48Q, and L126Z). Spinal cords were taken at time points throughout the lifespan of these transgenic SOD1 mice, and spinal cord homogenates were extracted in non-ionic detergent (0.5% NP40). In this assay, th e detergent-insoluble fraction represents aggregated SOD1 protein. In the fi rst 3 months of diseas e course, in SOD1-G93A mice, the levels of detergent-insoluble SOD1 remained low (Fig. 4-1A, lanes 2, 3, and 4) and roughly equivalent to the level of detergent-in soluble SOD1 in contro l mice expressing WT SOD1 (Fig. 4-1A, lane 1); rising only after 120 to 150 days (Fig. 4-1A, lanes 5 and 6), at which time the mice were paralyzed. Throughout the lifespan of the SOD1-G93A mice, the amount of detergent-soluble SOD1, which is presumed to represent the fr action of mutant protein that acquires a near normal conformation, remained cons tant (Fig. 4-1B). Similarly to SOD1-G93A mice, SOD1 aggregates were dete cted only near the endstage of disease in SOD1-G37R, SOD1H46R/H48Q, SOD1-L126Z mice, which have similar ages of disease onset (Fig. 4-1C, lanes 3

PAGE 72

72 and 4; Fig. 4-1E, lanes 5 and 6; Fig 4-1F, lanes 4 and 5; respectiv ely). Prior to the presence of overt symptoms for these mutant lines of mice, th e levels of detectable detergent-insoluble SOD1 were not different from that of control WT SO D1 mice (Fig. 4-1C, lanes 1 and 2; Fig. 4-1E, lanes 1-4; Fig. 4-1F, lanes 1 and 2). The levels of G37R and H46R/H48Q SOD1 in the detergent-soluble fr action remained constant as the mice aged (Figs. 4-1D and F). SOD1-L126Z has a short half-life; s o, little detergent-soluble SOD1 accumu lates throughout the disease course (Fig. 4-1 H). Thus, each SOD1 variant illustra ted similar aggregation profiles throughout their lifespans in that the levels of de tergent-insoluble SOD1 aggregates rise only near the endstage of disease. Because all SOD1 variants il lustrate similar aggregate a ccumulation over time, we sought to compare the variants at presymptomatic and symptomatic stages of the disease. At a presymptomatic time point, the spinal cords of each line of SOD1 transgenic mice (G93A, G37R, H46R/H48Q, and L126Z) contained no greater leve l of detergent-insolu ble SOD1 than mice expressing WT SOD1, with the exception of SOD1-L 126Z (Fig. 4-2A). Thus, at the early stages of the disease, the levels of detergent-insolubl e SOD1 are low with the vast majority of the protein found in the detergent-solubl e fraction (Fig. 4-2B). At disease endstage for each of these mutants, detergent-insoluble SOD1 was readily detected (Fig. 4-2C, lanes 2-9). While the degree of accumulation of detergent-insoluble SOD1 was similar in the G93A and H46R/H48Q mice, the levels of detergent-insoluble SOD1 in endstage G37R mice were consistently lower than the other two lines (Fig. 4-2C). SOD1 in the detergent-soluble fraction at disease endstage (Fig. 4-2D) was similar to that present at the pr esymptomatic stage (Fig. 4-2B), which indicates that overall SOD1 levels do not change as the mice age. It is interesting to note that the amount of detergent-insoluble SOD1 detected at disease endstage (Fig. 4-2C) may be related to the age

PAGE 73

73 that the mouse develops paralysis. SOD1-G 37R mice, which develop symptoms between 210 and 240 days, have the lowest amount of deterg ent-insoluble SOD1 species (Fig. 4-2C, lanes 2 and 3). These findings indicate that the levels of accumulated SOD1 aggregates at endstage disease vary to some degree among the different lines of mice. Figure 4-1. The levels of detergent-insoluble SOD1 increase drama tically late in the course of disease in SOD1 transgenic mice. Detergent (NP40) insoluble (20 g) and soluble proteins (5 g) were isolated from spinal cord as described in Methods and then electrophoresed in the presence of reduc ing agent before immunoblotting (hSOD1 antiserum). A, B) G93A de tergent-insoluble (A) and sol uble (B) proteins. C, D) G37R detergent-insoluble (C) and soluble (D) proteins. E,F) H46R/H48Q detergentinsoluble (E) and soluble (F) proteins. G,H) L126Z detergent-insoluble (G) and soluble (H) proteins. WT, 180 days. Denotes mice harvested at disease endstage (paralysis). In SOD1 transgenic mice, extensive patholog ical abnormalities have been described prior to the onset of paralysis. For each SOD1 varian t, aggregation was quantified and expressed as a percentage of the amount of detergent-insoluble SOD1 at disease endstage. In SOD1-G93A mice, aggregation dramatically in creased as the mice approach di sease endstage (Fig. 4-3A). When comparing the aggregation potential in these mice with other pathology that has been previously described, it was clear that a number of abnormalities occur pr ior to the accumulation

PAGE 74

74 of aggregated SOD1 (Fig. 4-3A). Simila rly, in SOD1-G37R mice, significant pathology, including vacuole formation and gliosis, occurred prior to the robust accumulation of detergentinsoluble SOD1 (Fig. 4-3B). Less is known about the progression of disease in SOD1H46R/H48Q mice, but we have dete cted evidence of reactive gliosi s in these mice as early as 60 days of age (Fig. 4-4B), well before the appear ance of significant levels of aggregated mutant SOD1 (Fig. 4-3C). Similarly, in SOD1-L126Z mice, we have detected evidence of gliosis at 150 days (Fig. 4-5) prior to aggreg ate accumulation (Fig. 4-3D), and B crystallin activation in astrocytes at 240 days (Fig. 4-6), at disease endstage. Note that variation in the age at which mice reach criteria for endstage (obvious hind limb paralysis) introduces variability in the lifespan of the mice. For this reason, the animals sacrificed for timed harvest may be at different stages of disease, leading to variability in ag gregate loads in tissues at pre-symptomatic ages. SOD1-G93A mice, which have the shortest lifespan, also had the most rapid increase in aggregate levels (Fig. 4-3A). Correspondi ngly, SOD1-H46R/H48Q mice, which have the longest lifespan, had the slowest in crease in aggregate levels (Fig. 4-3C). Thus, it appears that the earlier the mice developed symptoms, the more quickly the levels of a ggregated SOD1 rise. However, in all cases, significant pathologic abno rmalities occurred in spinal cord prior to the accumulation of significant levels of mutant SOD1 aggregates. It has been proposed that oxidative stress may enhance the formation of non-native, intermolecular disulfide bonds th at contribute to the forma tion of high-molecular-weight disulfide cross-linked aggregates (141), a prom inent feature in symptomatic SOD1 transgenic mice (123, 141, 142). We sought to determine if disulfide cross-linking between misfolded SOD1 species occurs as a precursor to aggregate formation. Spinal cords were taken at time points throughout the lifes pan of SOD1-G93A transgenic mice a nd were extracted in 1% SDS in

PAGE 75

75 the presence of iodoacetamide (IA), a thiol modifying agent that irreversibly prevents air oxidation and disulfide bond scramb ling. Treating spinal cord hom ogenates with IA allows for a more accurate representation of the disulfide status of SOD1 in th e spinal cord at the time of extraction, without experimental alteration. A high concentration of SDS (1%) was sufficient to solubilize most of the SOD1 protein (Fig. 4-7A), which allowed for the detection of the total amount of high-molecular-weight, disulfide cross-li nked species in the spinal cord at each time point. High-molecular-weight, disulfide crosslinked SOD1 species were most prominent at disease endstage in SOD1-G93A mice (Fig. 4-7B, lanes 7 and 8). At 105 days, 2-3 weeks prior to the appearance of overt hindlimb weakness, high-molecular-weight disulfide cross-linked species were detectable (Fig. 47B, lane 6). These high-molecu lar-weight species correspond to SOD1 multimers, representing dimers, trimers, te tramers, and larger multimers. Prior to 105 days, high-molecular-weight species were virtually undetectable (F ig. 4-7B, lanes 3-5). Treating the SDS-solubilized fractions w ith reducing agent, pr ior to gel electrophoresis, produced SOD1 monomer (Fig. 4-7C), indicating that the high -molecular-weight species detected (Fig. 4-7B, lanes 6, 7, 8) were composed of disulfide cross-linked SOD1 species. For comparison to the appearance of disulfid e cross-linked mutant pr otein, we show the time course of the appearance of NP40-insoluble mutant SOD1 at the ages examined (Figs. 47D, E, and F). High-molecular-weight, disulf ide cross-linked, SOD1-G 93A was most abundant at disease endstage (Fig. 4-7D, lanes 7 and 8), with a smaller fraction of high-molecular-weight species present prior to paralysis at 105 days (F ig. 4-7D, lane 6). Before 105 days, virtually no detergent-insoluble, disulfide cross-linked, SOD1-G93A was detect ed (Fig. 4-7D, lanes 3-5). The detergent-soluble fraction contained only SOD1 monomer until disease endstage (120 and 150 days), when faint high-molecular-weight SOD1 immunoreactive bands were detected (Fig.

PAGE 76

76 4-7E, lanes 7 and 8). Collectively, these data de monstrate that the levels of disulfide crosslinked mutant SOD1 rise in parallel with th e levels of detergentinsoluble protein. The logistics of harvesting tissues from multiple mice at different stages of disease requires storage of tissues, homogena tes, or individual tissue extrac ts for some period of time. The studies here included tissues th at had been in storage for exte nded intervals as well as tissues harvested and used within a few days. In our experience, storage of the tissues by flash freezing on dry ice followed by storage at -80 C introduces the least artifact in analysis of mutant SOD1 aggregation. By this approach, tissues from diffe rent animals at different ages can be directly compared by side-by-side extraction and im munoblotting. We observed no obvious indication that tissues stored for longer intervals contained higher levels of disulfid e cross-linked material. In one example, we compare the extraction of tissues from 60-day-old animals stored at -80 C to Figure 4-2. Variation in the levels of detergent-insoluble SOD1 in the spinal cords of paralyzed FALS mice. Spinal cords were extracted in buffers containing 0.5% NP40 to isolate detergent-insoluble and sol uble protein fractions and an alyzed by immunoblot as described in Figure 1. A) De tergent-insoluble SOD1 (20 g total protein from insoluble fraction) from presymptoma tic mice: L126Z (120 d), G93A (60 d), H46R/H48Q (60 d), G37R (120 d). B) Detergent-soluble SOD1 (5 g total protein from soluble fraction). C) Detergent-insoluble SOD1 (20 g total protein from insoluble fraction) from endstage (paral yzed) mice: G37R (240 d), H46R/H48Q (210 d), L126Z (270 d), and G93A (150 d). D) Detergent-soluble SOD1 (5 g total protein from soluble fr action). WT, 180 days.

PAGE 77

77 tissues that were harvested and extracted in the same day (Fig. 4-7, compare lanes 3 and 4 all panels). Freezing and storage did not induce the formation of disulfide cross-linked material. Thus we are confident that the storage of tissue at -80 C did not artificially reduce disulfide cross-links and that such cro ss-linking do not extensively occur prior to the appearance of detergent-insoluble mutant SOD1. Because high-molecular-weight, disulfidelinked SOD1 species were prominent at disease endstage in SOD1-G93A mice, we sought to determine whethe r the maintenance of structures that are de tergent-insoluble is dependent upon di sulfide cross-linking. Spinal cords from SOD1-G93A mice (at disease endstage) were extracted in 0.5% NP40 and in the presence Figure 4-3. Detergent-insoluble SOD1 accumulates to high levels near disease endstage. The amount of SOD1 present in the detergentinsoluble protein fraction was quantified by measuring the band intensity of SOD1 in each lane using a Fuji imaging system (FUJIFILM, LifeScience, Stamford, CT USA). Aggregation was quantified by calculating the ratio of band intensity in th e insoluble fraction at each time point to the insoluble fraction at disease endstage (set at 1). The standard error of the mean (SEM) was calculated for aggregation of each sample and graphed (A) G93A; B) G37R; C) H46RH48Q; D) L126Z). Timelines of pathologic changes were graphed with aggregation.

PAGE 78

78 of varying concentrations of reducing agent ( ME). In comparison to extractions performed in the absence of ME, extraction of the spin al cord tissue in 30% ME did not substantially change the amount of mutant SOD1 that partitioned into the dete rgent-insoluble fraction (Fig. 48A). Similarly, extraction in 30% ME did not alter the amount of mutant SOD1 that partitioned into the detergent-soluble fraction (Fig. 4-8B). Thus, while high-molecular-weight species of SOD1 are extensively disulfide cross-linked (F ig. 4-7D), these bonds are not responsible for maintaining the structure that i nduces detergent-insol ubility, which we equate with aggregation. Figure 4-4. Astrogliosis occurs early in disease progression in SOD1-H46R/H48Q mice. At the ages specified, SOD1-H46R/H48Q mice were perfused with 4% paraformaldehyde. Tissue was stored in 30% sucrose prior to cr yostat sectioning (14 microns). Sections were blocked in 5% normal goat serum (Invitr ogen, Calsbad, CA), and tissue sections were stained with antibodies to GFAP (Chemicon) followed by secondary anti-mouse IgG antibodies conjugated to Alexafluor 488 F(ab)2 fragment (green) (Invitrogen). NTg, age matched littermates. Ventral horn, spinal cord. Magnification of image at capture 40x.

PAGE 79

79 Recent studies have demonstrated that SOD1 that has acquired its normal intramolecular disulfide bond can be distingui shed from reduced forms by el ectrophoretic migration in SDSPAGE (in the absence of reducing agents) ( 142, 168). The oxidized form of the enzyme migrates slightly faster than the reduced form; possibly as a result of the more compact structure of the oxidized form of the protein. In non-re ducing SDS-PAGE of NP40 -insoluble fractions from the spinal cords from endstage G93A mice, we detected a form of mutant protein that appeared to be completely reduced with an a pparent molecular mass of about 20 kDa (see Fig. 47D, lanes 7 and 8). In the detergent-soluble fractions of mice across all ages, in these nonreducing gels, we observed mutant SOD1 proteins of 20 kDa and approximately 16 kDa (see Fig. 4-7E, lanes 2-8). This banding pattern matched that described by Jonsson et al, as reduced Figure 4-5. Astrogliosis occurs at 150 days in L126Z mice. At the ages specified, SOD1L126Z mice were perfused with 4% paraformaldehyde. Tissue was stored in 30% sucrose prior to cryostat s ectioning (14 microns). Sec tions were blocked in 5% normal goat serum (Invitrogen, Calsbad, CA), and tissue sections were stained with antibodies to GFAP (Chemicon) followed by secondary anti-mouse IgG antibodies conjugated to Alexafluor 488 F(ab)2 fragment (green) (Invitrogen). Ventral horn, spinal cord. Magnification of image at capture 40x. Scale bar 10 microns.

PAGE 80

80 (labeled R) and disulfide-oxi dized (labeled O) SOD1 prot ein (142). The presence of the single higher-molecular-weight band in the detergen t-insoluble fraction sugge sted that all of the monomeric SOD1 protein in this fraction is completely reduce d. However, in the study by Zetterstrom et al, the authors demonstrated that the detection of oxidized forms of SOD1 by nonreducing SDS-PAGE and immunoblo tting is greatly enhanced when the gels are incubated in reducing agent prior to electrophor etic transfer to immunoblotti ng membranes (168). Treating the SDS and NP40 solubilized fractions with reducing agent, prior to gel electrophoresis, produced a single SOD1 band that migrated at about 20 kDa (Fig. 4-7C, F). Therefore, we concluded that the spinal cords of G93A mice contained both reduced and oxidized forms of SOD1. Figure 4-6. B crystallin is upregulated in astrocytes at dise ase endstage in L126Z mice. At the ages specified, SOD1-L126Z mice were perf used with 4% paraformaldehyde. Tissue was stored in 30% sucrose prior to cryosta t sectioning (14 microns). Sections were blocked in 5% normal goat serum (Invitroge n, Calsbad, CA), and tissue sections were stained with antibodies to B crystallin (Stressgen) followed by secondary antimouse IgG antibodies conjugated to Alexafluor 568 F(ab)2 fragment (green) (Invitrogen). Ventral horn, spinal cord. Magnification of image at capture 40x. Scale bar 40 microns.

PAGE 81

81 To confirm this finding and eliminate any pot ential influence of sample preparation or storage, we sacrificed paralyzed animals and performed the detergent extraction and SDS-PAGE in the same day. We also included a treatment in which the gels were incubated in reducing agent prior to electrophoretic transfer to immunoblotting me mbranes. Zetterstrom and colleagues demonstrated that the detection of oxidized forms of SOD1 by non-reducing SDSPAGE and immunoblotting is enhanced by in-g el reduction of disulfide bonds prior to immunoblot transfer (168). In detergent-soluble fractions of ti ssues from both WT and endstage G93A mice, we detected both reduced and oxi dized forms of monomeric SOD1. In the detergent-soluble fractions the oxidized form of the protein represented the vast majority of detectable SOD1 protein (Fig. 4-9B, lanes 2-4). In the detergent-insoluble fractions, no WT SOD1 was observed and the major detectable sp ecies of mutant G93A SOD1 migrated as expected for reduced protein (Fig. 4-9A, lanes 2-4). A second minor species of detergentinsoluble mutant SOD1 in these fractions migrated to size similar, but not identical to, oxidized SOD1. This species was not detected when in-gel reduction before transfer was omitted (Fig. 49C, lanes 3-4). Whether this apparently oxidi zed, detergent-insoluble form of mutant SOD1 possesses a disulfide bond between cysteines 57 and 146 is uncertain. It is possible that an intramolecular disulfide between other cysteine residues produ ced this species of mutant, detergent-insoluble SOD1. Similarly, we found si gnificant levels of reduced mutant SOD1 in the detergent-insoluble fraction of tissues from endstage G3 7R and H46R/H48Q mice (Fig. 410A). For reasons that are unclear, in-gel reduction noticeably reduced the amount of highmolecular-weight SOD1 species detected in the detergent-insoluble fractions in these experiments. It appears that in-gel reduction increases either the transfer efficiency or antigenicity of SOD1 (with the normal 57 to 146 disulfide bond) for immunoblot detection (168);

PAGE 82

82 however, this treatment reduces one or both of these parameters for high-molecular-weight SOD1 with abnormal intermolecular disulfide bo nds. Whether the disulfide cross-linked highmolecular-weight mutant SOD1 possesses a normal 57 to 146 disulfide bond is unclear. Overall, we conclude that a portion of the mutant S OD1 that accumulates as detergent-insoluble is derived from protein that either never acquired or lost the normal intramolecular disulfide bond. Figure 4-7. The appearance of disulfide cross-linked SOD1 is coincident with the accumulation of detergent-insoluble mutant protein. G 93A spinal cords were extracted in 1% SDS (panels A,B,C) or in 0.5% NP40 (panel s D,E,F) in the presence of 100 mM iodoacetamide. Fractions were electrophores ed in the presence (C, F) or absence (A,B,D,E) of ME. Immunoblots were probed with m/hSOD1 antiserum. NTg, nontransgenic. WT, WT SOD1. A) Insoluble in 1% SDS (20 g). B) Soluble in 1% SDS (5 g). C) Soluble in 1% SDS electrophoresed in the presence ME. D) Insoluble in 0.5% NP40. E) Soluble in 0.5% NP40. F) Soluble in 0.5% NP40 electrophoresed in the presence ME. R, reduced disulfide bond (C57-C146). O, oxidized disulfide bond (C57-C146). Denotes disease endstage (paralysis). Denotes fresh spinal cord tissue. Ntg, 90 days. WT, 180 days.

PAGE 83

83 We re-examined the detergent-insoluble form s of mutant SOD1 that accumulate in the spinal cords of G93A, G37R, and H46R/H48Q mice using the in -gel reduction procedure to enhance detection of oxidized form s of SOD1. In spinal cords of endstage mice that express each of these three mutants, in the detergentinsoluble fraction we detected both reduced and oxidized forms of monomeric SOD1 (Fig. 4-10A, lanes 3-5). The relative ratio of reduced to oxidized was reproducibly in favor of the reduced form being most prominent. By contrast, in the detergent-soluble fractions, the ratio of reduc ed to oxidized protein was far in favor of the oxidized form in mice that express WT, G93A and G37R SOD1 (Fig. 4-10B, lanes 2-5). However, in mice that express H46R/H48Q SOD1, th e ratio of reduced to oxidized protein in the detergent-soluble fraction remained in favor of the reduced form of the protein (Fig. 4-10B, lane 5). Thus, these results are consiste nt with the cell culture findings in that, with regard to forms of detergent-insoluble mutant SOD1 that migr ate near the position expected for monomeric enzyme, reduced forms of mutant SOD1 app ear to preferentially acquire detergentinsolubility. To further explore the role of reduced a nd oxidized mutant SOD1 in aggregation, we turned to a cell culture model of aggregation that utilizes HE K293-FT cells (23, 43, 46). The high levels of expression that are achieved in th ese cells induced rapid aggregation of mutant SOD1 (A4V, G37R, G93A) and the formation of detergen t-insoluble, sedimentable forms of the protein (Fig. 4-11) [also see ( 23, 43, 46)]. When we applied the non-reducing SDS-PAGE, with or without, in-gel reduction prior to transfer, we noted a number of interesting features of the cell culture model. First, we determined that in the cell culture model, the majority of detergentinsoluble mutant SOD1 that accumulates appear s to be completely reduced (compare Figs. 411A and C). In these gels we included purified dimeric, metallated, WT human SOD1 (isolated

PAGE 84

84 from yeast) as a control (lanes 1 and 8 of each gel these pr oteins were not subjected to detergent extraction or centrif ugation). Reduction and denatura tion of this purified protein produced a single band that migrated at approx imately 20 kDa (Fig. 4-11A, lane 1), where as denaturation and SDS-PAGE in the absence of reducing agent produced a single band that migrated at approximately 16 kDa (Fig. 4-11A, lane 8). Note that the oxi dized form of purified WT SOD1 was not as readily detected when the in-gel incubation with reducing agent was omitted prior to transfer (compare Figs. 4-11A and C, lane 8 in each) [also see (168)]. Figure 4-8. SOD1 aggregates re sist dissociation in high concen trations of reducing agents. Spinal cord tissue from a symptomatic G93A transgenic mouse was extracted in 0.5% NP40 and increasing concentrations of ME (noted on Figure). Immunoblots were probed with hSOD1 antiserum A) Detergent-insoluble (20 g). B) Detergentsoluble (5 g).

PAGE 85

85 Applying this technique to the detergent-in soluble forms of the A4V, G37R, G93A, and H46R/H48Q mutants produced in ce ll culture demonstrated an elec trophoretic migration to a size closer to the reduced forms of purified WT SOD1 (Figs. 4-11A and C). Note that the relative detection of reduced and oxidized mutant protein in these fractions was similar between gels that were or were not incubated in reducing agents prio r to transfer. In these cell culture experiments and similar to previous studies (43), the H46R /H48Q mutant produced le ss detergent-insoluble protein in cell culture. By contrast in the detergent-soluble fractions of these cell extracts, we detected both reduced and oxidized forms of SOD1 (Fig. 4-11B, lanes 3-7). Omission of the ingel reduction demonstrated the upper band to be the reduced forms of SOD1 (Fig. 4-11D, lanes Figure 4-9. SOD1 aggregates are largely composed of disulfide-reduced forms of SOD1 in fresh spinal cord tissue. Spinal cords from freshly harvested mice were extracted in nonionic detergent and iodoacetamide. A and B) Gels were incubated in reducing buffer prior to transfer. C and D) Gels were processed without reducing agent. A and C) Detergent insoluble (40 g protein). B and D) Detergent soluble (40 g protein). Immunoblots were probed with m/hSOD1 antiserum. R, reduced. O, oxidized. #, Possible non-natively oxidized bond.

PAGE 86

86 3-7). In analyzing the detergen t-soluble fractions in the blots that included the in-gel reduction, we noted that the relative ratio of reduced to oxidized protein differed between the mutant and WT SOD1 proteins. In a very reproducible patt ern, the predominant form of WT SOD1 in these transfected cells migrated to the same positi on as purified oxidized SOD1 (Fig. 4-11B, lanes 3 and 8). By contrast, for each of the mutants, th e predominant form of detergent-soluble protein migrated to the position of reduced WT SOD1 (Fi g. 4-11B, compare lane 1 to lanes 4-7). For the A4V and H46R/H48Q mutants, the majority of detergent-soluble protein migrated to the same position as reduced WT protein. Together, thes e findings demonstrate th at reduced forms of mutant SOD1 (lacking the normal intramolecula r disulfide bond and presumably any other type of intramolecular disulfide linkage) are more prone to form detergent-insoluble complexes, which is consistent with evidence (see Chapte r 2) which suggests that cysteine residues are important for aggregation via a mechanism alternative to disulf ide cross-linking. Discussion High-m olecular-weight, disulfide-linked SOD1 a ggregates accumulate to high levels in paralyzed transgenic mice that express mutant forms of SOD1 linked to familial ALS (123, 141, 142). In the present study, we sought to trace the accumulation of these species throughout the lifespans of four lines of mu tant SOD1 transgenic mice (G 93A, G37R, H46R/H48Q, L126Z). We demonstrate, in four lines of mutant SOD1 transgenic mice (G93A, G37R, H46R/H48Q, L126Z), that the formation of SOD1 aggregates follo ws a similar time course. In all four lines of mice, SOD1 aggregates accumulate to high levels just prior to hindlimb paralysis but well after the appearance of several patholog ic features in spinal cord. We also demonstrate that the appearance of disulfide cross-linked mutant SOD1 parallels the appearance of detergentinsoluble (NP40-insoluble) protein. We find no evidence that disulfide cross-linked forms of SOD1 are precursors to more hi ghly structured aggregates. Mo reover, we demonstrate that

PAGE 87

87 disulfide cross-linking is not responsible for the maintenan ce of structures that induce insolubility in NP40 detergent. In both cell and mouse models, a sign ificant fraction of the NP40-insoluble mutant SOD1 dissociates upon SD S and heat denaturation (in the absence of reducing agent) to what appears to be monomeric protein. These species of mutant SOD1 must be organized into high-molecular-weight structures, which results in their sedimentation upon centrifugation, that are bound via for ces other than disulfide crosslinking. We demonstrate that the majority of these non-cross-linked forms of mu tant protein in SOD1 aggregates appear to lack the normal intramolecular disulfide bond. Moreover, our work in cell culture models suggests that completely reduced forms of mutant protein are more prone to produce aggregates. This finding is in agreement with a recent study of the in vitro aggregation of purified mutant SOD1 (169). Collectively, our data supports a m odel in which the mutant SOD1 that fails to acquire the normal intramolecular disulfide bond (or suffers reduction of this bond) is prone to aggregate. These events seem to occur to the greatest extent late in disease progression. The implications of these data in regard to the role of aggregation in disease pathogenesis are discussed below. Redox Chemistry and SOD1 Aggregation in SOD1 Transgenic Mice High-m olecular-weight, detergent-insoluble, SOD1 complexes (termed aggregates) have been repeatedly detected in spinal cords of tran sgenic mice that model SOD1-linked FALS (see Chapter 2) (23, 46, 117, 141, 142). Recently, seve ral groups have published studies that implicate aberrant disulfide cross-linking between cysteine residues, particularly cysteines 6 and 111, of mutant SOD1 in the formation of thes e complexes (23, 46, 141, 144, 145). However, we have provided evidence to indi cate that disulfide bonding is not the only force stabilizing these aggregates (46)(see Chapter 2). In this study, we sought to determine whether disulfide crosslinking precedes the formation of detergent-insoluble SOD1 aggregates. We determined that

PAGE 88

88 disulfide cross-linked forms of mu tant protein are detected concu rrently with detergent-insoluble mutant SOD1, suggesting that it is unlikely that an oxidative event occurs early in the disease course to induce disulf ide cross-linking, which then induces the formation of aggregates. This line of thinking is consistent with a previous st udy from our group that demonstrated that mutant forms of SOD1 that are incapable of forming disu lfide cross-links retain the ability to aggregate into high-molecular-weight specie s (46)(see Chapter 2). We also demonstrate that these highmolecular-weight structures are dissociat ed by high levels of reducing agent (30% ME), leaving SOD1 monomer intact, which indicates that these aberrant disulfide cross-links are not directly involved in the formation or ma intenance of these structures. Instead, it is likely that disulfide cross-links form after the mutant SOD1 protei ns come into close proximity as a result of assembly into high-molecular-weight structures. Post-Translational Modification of Mutant S OD1 and Aggregation In addition to high-molecular-weight, disu lfide cross-linked, forms of mutant SOD1, detergent-insoluble fractions also contain forms of mutant protein that migrate at the size expected for monomeric protein. Using a protocol developed by Jonsson and colleagues (142), we demonstrate that the majority of the muta nt SOD1 from G37R, G93A, and H46R/H48Q mice that is NP40-insoluble but dissociable into monomeric species by heat and SDS denaturation shows the electrophoretic characte ristics of completely reduced WT SOD1. This outcome is consistent with the study of Jonsson and colleagues (142); which demonstr ated that a significant fraction of the total SOD1 expressed in tr ansgenic mice producing D90A, G93A, and G127X variants of SOD1 lacks the normal intr amolecular disulfide bond (C57-C146). Our study of SOD1 aggregation in cell culture provided more definitive evidence for the contribution of reduced forms of mutant SOD1 in aggregation. In our cell model, the vast

PAGE 89

89 majority of the NP40-insoluble mutant SOD1 th at dissociates upon heat and SDS denaturation to a monomeric size protein exhibite d electrophoretic mobilit ies consistent with reduced protein. Moreover, when we examined the mutant proteins that were NP40-solubl e, we noted that the majority of the protein produced in these cells showed electrophoretic mob ilities consistent with reduced protein, suggesting that in our cell cultur e model the high level expression of the mutant protein overwhelms the ability of the cells to provide the necessary factors to correctly fold most of this protein. In contrast, the majority of WT SOD1, even when overexpressed, displayed electrophoretic properties consistent with the presence of the normal intramolecular disulfide bond. These findings are informative in three ways First, we can be confident that reduced forms of mutant SOD1 are prone to form NP40-insoluble aggreg ates. Second, we demonstrate that the failure to form the normal disulfide linkage is not suffici ent to induce aggregation. Finally, a significant fraction of disulfide-reduced mutant SOD1 can achieve conformations that remain soluble in detergent. In spinal cords of the WT, G37R, and G93A mice, we found that the majority of SOD1 that achieves a conformation that is soluble in detergent shows electrophoretic characteristics consistent with the presence of the normal intr amolecular disulfide bond. Importantly, in these mice, the levels of expression, though high, are still an order of magnitude or more lower than what is achieved in cell culture Moreover, the time scales in the mouse and cell systems are vastly different. Thus, in the animals, there is either sufficient capacity or sufficient time to allow some mutant SOD1 to achieve detergen t-soluble conformations with normal disulfide linkages. In the spinal cords of H46R /H48Q mice, the detergent-so luble forms of the protein appeared to be largely reduce d. Evidence suggests that coppe r is required to form the

PAGE 90

90 intramolecular disulfide bond ( 170); our data on the H46R/H48 Q mutant, which does not bind Cu (171), is consistent with this hypothesis. Previous work has shown that FALS mutants of SOD1 are more susceptible to proteolytic dige stion in the presence of reducing agents (31), suggesting that mutant SOD1 is destabilized under reducing conditions in the cell. Because reduced SOD1 monomers have lo wer stability and higher confor mational freedom (172), these species are more likely to contribute to the formation of aggregates. WT SOD1 is less susceptible to disulfide reduction than mutant SOD1 (31), which may explain why WT SOD1 is less prone to misfold. A recent study of the aggregation of purified SOD1 in vitro demonstrated that completely reduced forms of mutant S OD1 were most prone to produce filamentous aggregates (169). Figure 4-10. Reduced forms of mutant SOD1 are components of NP40-insoluble aggregates. Spinal cords from three lines of SOD1 mice at endstage were extracted in 0.5% NP40 in the presence of 100 mM iodoacetamide. Immunoblots were probed with m/h SOD1 antiserum. A) Detergent-insoluble (20 g). B) Detergent-soluble (5 g). To enhance detection of oxidized forms of hSOD1, gels were incubated in transfer buffer containing 2% ME for 10 min before electrotransfer to nitrocellulose membranes. R, reduced disulfide bo nd (C57-C146). O, oxidized disulfide bond (C57-C146). #, Possible non-native ly oxidized disulfide bond.

PAGE 91

91 Figure 4-11. Reduced hSOD1 protein is preferen tially incorporated into detergent-insoluble aggregates. HEK293-FT cells were transfected with vectors for WT, A4V, G37R, G93A and H46R/H48Q hSOD1 proteins, harv ested after 48 hours, and extracted in buffers with 0.5% NP40 and 100 mM iodoaceta mide. Detergent-insoluble (A, C) and soluble fractions (B, D) from cells transfected with each construct were electrophoresed in the absence of ME (lanes 2-8). Immunoblots were probed with m/h SOD1. Lane 1 of each panel contains purified WT SOD1 that was reduced prior to electrophoresis (*). This lane was separated from the remaining samples by a lane containing marker proteins and an empt y space to prevent reducing agent from diffusing into other samples (these interven ing lanes have been cropped out of the gel image). Lane 8 of each panel contains purified WT SOD1 that was verified to have an intact intramolecular disulfide bond ( ). Purified WT SOD1 proteins were treated with 100 mM iodoacetamide prior boiling in Sample buffer and electrophoresis. To enhance detection of oxidized forms of hSOD1, gels were in cubated in transfer buffer containing 2% ME for 10 min before electrotransfer to nitrocellulose membranes (A, B). UT, untransfected cells. R, reduced disulfide bond (C57-C146). O, oxidized disulfide bond (C57-C146). Image is representative of 3 repetitions of the experiment. While a fraction of reduced SOD1 protein may remain soluble in non-ionic detergent, the presence of reduced SOD1 in aggregates produced in mouse and cell models supports the hypothesis that immature (with respect to acquis ition of Cu or Zn cofactors and formation of intramolecular disulfide bonds) mutant SOD1 is more prone to misfold and produce conformations that support aggr egation [reviewed in (173)].

PAGE 92

92 We note that some of the tissues used in this study were stored at -80 C for substantial intervals, which could have provided an opportunity for ex-vivo oxidation and artifactual generation of disulfide linkages. We did not obser ve an obvious pattern in which tissues stored for longer periods of time produce higher amounts of disulfide cross-linked forms of SOD1. It is possible, however, that some oxidation of cysteine residues occurs during tissue storage or sample preparation. We have no indication of disulfide reduction with storage or sample preparation. Thus, we can conclude that the reduced forms of mu tant SOD1 we describe here represent the minimum levels of this form of SOD1 in tissues of symptomatic mice. The best estimation of the relative amount of completely reduced mutant SOD1 in detergent-insoluble fractions is obtained by analysis of the data in Fig. 4-7, in whic h we estimate that completely reduced protein may account for ~10% of total detergent-insoluble muta nt protein in mouse tissues. However, we do not know what proportion of the mutant protein that is associated with high-molecular-weight, disulfide cross-linked species possesses a normal 57 to 146 disulfide linkage. Mutant SOD1 Aggregation in Disease Pathogenesis In this study, aggregates are defined as form s of mutant SOD1 that are insoluble in nonionic detergents (NP40) and sediment upon cent rifugation. When exposed to heat, SDS, and reducing agents, these aggregates dissociate a nd predominantly show electrophoretic mobilities consistent with a monomeric protein. Mass spectrum analysis of NP40-insoluble SOD1 aggregates have recently demonstrated that unmodified mutant SOD1 proteins are major components of the aggregate (151). Notably, in this later study, aggregates were dissociated in guandinium isothiocyanate and DTT prior to an alysis, which would have reduced all oxidized cysteine modifications. Our study here, along with several previous studies (23, 46, 141) demonstrates that intermolecular disulfide crosslinking is a major modification to mutant SOD1

PAGE 93

93 in the aggregates. However, the timing of the ap pearance of this modification in mice, and the lesser degree of such modifications in cell culture models of aggrega tion, lead us to conclude that oxidative cross-linking of disulfide residues is likely to be a sec ondary event in the formation of the aggregate structures. Several groups have described SOD1 aggr egation in mutant SOD1 transgenic mice through various stages of the lifespans (117, 142). Our study builds on the cu rrent literature. In a previous study from our group, a filter trap assay was used to detect and measure the aggregation of SOD1-G37R in mutant mice-findi ng that aggregates are most abundant when mice are overtly paralyzed (210-240 days) with a lesser degree of SOD1 aggregation noted at ages prior to symptoms at 180 days (114). We ha ve recently determined that the methods used in this study may have primarily detected the disulfide crosslinked forms of mutant protein (46)(Chapter 2). However, the outcome of the study is essentially unchanged in regards to the timing of the appearance of mutant SOD1 aggr egates. In G93A mice, detergent-insoluble protein complexes have been de scribed to increase linearly thr ough the lifespan of the animal and were detected as early as 30 days of age (117). In this particular study, aggregates were defined as high-molecular-weight sp ecies that were resistant to heat, SDS, and reducing agents, and they were unique to mice expressing SOD1-G93A (117). Recently, however, we have established that these forms of mutant SOD1 ar e distinct from the NP40-insoluble protein we have studied here (46)(Chapter 2). It is unc lear whether the high-mol ecular-weight species of mutant SOD1 that are resistant to SDS, heat, and reducing agents repr esent forms of mutant protein which are covalently cross-linked to othe r proteins (such as ubiquitin (43, 104)) or some type of highly stable oligomeric structure. Howe ver, it is clear that these SDS, heat, reducing

PAGE 94

94 agent resistant species of mutant protein ar e ultimately components of NP40-insoluble aggregates (43, 46, 104). The study that most closely resembles our pr esent report examined mice expressing G93A, G127X, and G85R FALS variants of SOD1 (142). In this study, spinal cords were extracted in low concentrations of non-ionic detergent (0.1% NP40), which resulted in the detection of detergent-insoluble species of mutant SOD1 as early as 50 days (123, 142). However, similar species were detected in WT SOD1 mice at all time points (142). At terminal stages of disease, the levels of mutant protein that were NP40-inso luble increased dramatica lly, including forms of mutant protein that were resistant to SDS, heat, and reduci ng agent (142). Moreover, in mice expressing WT SOD1 at very adva nced ages (400 to 600 days), th e levels of NP40-insoluble WT SOD1 were observed to increase with the appe arance of high-molecular-weight species were SDS, heat, and reducing agent resistant (142). Thus, the methodology employed in the Jonsson study indicates that both mutant and WT SOD1 adopt similar conformatio ns as a function of age. While our assay is similar to Jonsson and colleagues (142, 168), it differs in several significant ways. First, we use 5-fold higher co ncentrations of NP40 in our extraction buffers (0.5%), with a 20 to 1 ratio of buffer to tissue (vol to wt). The Jonsson method uses 0.1% NP40 (the volume is not specified). Second, we cen trifuge tissue extracts at >100,000 xg in an Airfuge for 5 minutes followed by one wash step in which the pellet is resuspended in buffer with NP40, disrupted by sonication, and th en centrifuged at >100,000 xg to produce the final pellet. The Jonsson method uses 20,000 xg for 30 minutes, fo llowed by 5 washes in double the original extraction volume with 20,000 xg centrifugation to separate detergent-in soluble material. Comparison of these methods is difficult because of uncertainty regarding so me of the details of the protocols. We use sonicati on to disrupt the tissue in the absence of detergent, whereas

PAGE 95

95 Jonsson describes homogenization of the tissue in detergent as the method of tissue disruption. After initial disruption of the tissue, we us e a low speed centrifugation to remove tissue fragments that have not been completely disrup ted. This step is absent from the Jonsson protocol. Because we use higher detergent conc entrations, we conclude that we extract the tissues at higher detergent to protein ratios than that of the Jonsson method. We conclude that this difference in methodology, alone or in conjun ction with other differences, allows us to distinguish the forms of mutant SOD1 that are most distinct from WT protein. Despite the methodological differences between our approach and that of comparable studies, it appears to us that th e general consensus is that the levels of NP40-insoluble mutant SOD1 increase rapidly towards the terminal stages of disease. All variants of SOD1 transgenic mice (G93A, G37R, H46R/H48Q, L126Z) examined here illustrated a similar time course of SOD1 aggregate accumulation, where aggregates were most prominent when mice developed hindlimb paralysis. Our inclusion of the G93A mice provides a point of reference with other published studies (117, 142). In all of these studies, the levels of conformationally distinct forms of G93A SOD1 increase dramatically as paralysis develops. What is less clear is the nature and abundance of stable, misfolded, mutant protein earl ier in disease. The identification of SDS, heat, and reducing agent resist ant forms of high-molecular-weight SOD1-G93A by Johnston and colleagues as early as 30 days of age implies that conformational abnorma lities in mutant SOD1 occur early in disease pa thogenesis (117). The effects of FALS mutations on the normal enzyme activit y, turnover, and folding of SOD1 vary considerably (24, 26). In cell culture and in vitro models, enzyme activity ranges from undetectable to near normal (24, 27-30); mo st mutations accelerate the rate of protein turnover (24, 29); and many mutations increase the susceptibility of SOD1 to disulfide reduction

PAGE 96

96 (31). The G93A variant has an activity similar to WT SOD1 protein ( 174) and has a half-life similar to WT SOD1 protein (175); the G37R vari ant also has an activity similar to WT SOD1 (24, 147) and a half-life that is estimated to be half of th at of WT SOD1 (24); and the H46R/H48Q variant has no activity (104, 171). Despite this vast range of biophysical properties, all of the mutants, in mice and in humans, produce a similar phenotype of paralysis (104, 119, 147). Here, we demonstrate that SOD1 mutants with different properties also show a similar time course of SOD1 aggregate formation in rela tion to the development of hindlimb paralysis. SOD1 Aggregation and Toxicity Previous studies have esta blished that FALS mice devel op significant pathology prior to the appearance of hindlimb paralysis. In S OD1-G93A mice, approximately 40% of the motor endplates are denervated at 47 days, 80 days before the appearance of hindlimb paralysis and SOD1 aggregation (176). At this early time poi nt, motor unit loss [40 da ys (177)] and a decline of contractile force in fast twit ch muscles [40 days (178)] also oc cur, illustrating that dramatic loss of muscle function and moto r neuron innervation may occur ea rly in the disease course. Similarly, SOD1-G37R mice show vacuolar change s in spinal cord [36 days (179)], motor neuron loss [133 days (125)], and reactive astroglio sis [77 days (125)] before these mice develop hindlimb paralysis (210 days). The SOD1-H46R/H48Q mice that we re included in this study are less well characterized, but we have noted the app earance of astrogliosis in spinal cord as early as 60 days (Fig. 4-4). Our data indicate th at at time points when each of these mice have significant pathologic abnormalities, including as trogliosis, motor neuron loss, denervation, and vacuolation, only low levels of detergent-insoluble SOD1 can be detected. This outcome implies that other forms of mutant SOD1 may mediate toxic events at early stages of disease or that very low levels of aggregates are su fficient to mediate toxicity.

PAGE 97

97 It is difficult to discard SOD1 aggregation as entirely secondary to disease pathogenesis. In a study of mice that express SOD1-G37R via th e mouse PrP vector, we created a line of mice in which expression of the mutant protein, when animals were heterozygous for the transgenes, was below the threshold for inducing diseas e (114). However, upon mating to homozygousity, which resulted in a 2-fold increase in mutant protein expression, ALS symptoms were observed to occur by 9 months of age (149). In th e homozygous paralyzed mice, we detected NP40insoluble forms of mutant SOD1 whereas in the disease-free heterozygous mice there was no evidence of aggregated SOD1. More recentl y, Jaarsma and colleagues generated mice that express SOD1-G93A via a vector generated from the Thy1.2 promoter (133). In the heterozygous state, these animals do not develop disease and do not develop detergent-insoluble aggregates of mutant SOD1. When crossed to homozygousity, disease is produced at about 20 months of age, and paralyzed mice accumulate detergent-insoluble forms of mutant protein (133). A third example involves matings of mice that express mutant SOD1-A4V at levels too low to induce disease with mice that express high levels of WT SOD1 (134). Mice expressing the low level of SOD1-A4V do not accumulate detergent-insoluble SOD1 aggregates, but such aggregates along with paralytic disease are produced in mice that co-express A4V and WT SOD1 (134). Jaarsma and colleagues reported similar outcomes when Thy1.2-G93A mice were crossed to mice expressing high levels of WT SO D1 (133). Collectively, these studies provide a compelling argument for the notion that the thres hold of SOD1 burden (mutant alone or mutant with WT) that is required to induce disease is similar to the threshold required to induce aggregation. The late appearance of high-molecular-weigh t SOD1 aggregates in the progression of disease in FALS mice leaves open three possi bilities for SOD1 aggregate function in ALS

PAGE 98

98 pathology. First, it is possible that the accumulati on of SOD1 aggregates me diates all aspects of toxicity with early pathologic even ts mediated by very low levels of aggregates (below the level of detection). Under this scen ario, some of the early patholog y described in FALS transgenic mice prior to aggregate formation could be the result of overexpression of mutant SOD1. For example, the early vacuolar pathology is a majo r feature of mice expressi ng high levels of G93A and G37R variants of mutant SOD1, but much less apparent in mice expressing C-terminal truncation mutants or inactive mutants (104, 115, 119, 147). There is one example in transgenic mice where mutant SOD1 induces disease without a concurrent accumulation of SOD1 aggregates: coexpression of human copper chaperone for SOD1 (CCS) with SOD1-G93A, hastens the onset of disease and blocks aggregate formation (157). In these animals, vacuolar degeneration of the spinal cord is the primary pathology. It remains to be established whether CCS overexpression accelerates disease in mice that normally lack th e vacuolar pathology. Second, all toxic events are medi ated by detergent-soluble forms of mutant SOD1 (monomeric or oligomeric). In this scenario aggregates form because the toxic processes that occur in affected cell types leave them less able to properly m odify nascent mutant SOD1 by loading structure stabilizing metal cofactors, which induce normal intramolecular disulfide bonding. These immature SOD1 proteins then become the buildi ng blocks of aggregates. Under this scenario, the aggregates are simply biomarkers for extrem e cell stress. Lastly, some unidentified toxic event mediated by detergent-soluble forms of mu tant SOD1 (monomeric or oligomeric) triggers early pathologic abnormalities (gliosis, vacuol ation, denervation, etc) with the resulting stress on the cell inducing the formation of SOD1 aggregat es, which impart additional toxicity and induce paralytic phenotypes. Although there is one ex ample in which aggregation and toxicity of mutant SOD1 are clearly separate d (157), it is not entirely clear whether this example extends to

PAGE 99

99 other FALS mutants of SOD1 or involves toxi c mechanisms relevant to human disease. Ultimately, unraveling the true nature of SOD1 aggregation in SOD1-li nked FALS may require the identification of small molecules that selectively inhibit SOD1 aggregation. Conclusions The presen t study demonstrates that the accumula tion of detergent-insoluble aggregates of mutant SOD1 increases dramatically as FALS mi ce develop paralysis. Extensive disulfide crosslinking occurs in these aggregat es, but we do not find evidence that such cross-linking is a prerequisite to aggregation or is critical to aggregate stability (see also Chapter 2). We also provide evidence that SOD1 that fails to acquire the normal disulfide linkage is most prone to aggregation. Overall, these data suggest that oxidative stresses, which diminish cellular redox potential, are unlikely to provide the trigger that induces mutant SOD1 aggregation. Instead, it appears that failure of mutant SOD1 to mature is more critical. The fact that high-molecularweight forms of detergent-insoluble mutant SOD1 do not accumulate to high levels until very late in disease limits the role of such structures in disease pathogenesis.

PAGE 100

100 CHAPTER 5 B CRYSTALLIN IS A MODEST MODIFIER OF DISEASE PROGRESSION IN MOUSE MODELS OF ALS Introduction All m ouse models of FALS that overexpr ess mutant human SOD1 share a similar phenotype of motor neuron loss, muscle wasti ng, and hindlimb paralysis (180). In mouse models of FALS, mutant SOD1 is ubiquitously expressed; ho wever, cell death is largely restricted to motor neurons. In all SOD1 transgenic mice, the appearance of symptoms is associated with an accumulation of sedimentable st ructures that are detergent-insoluble, which is diagnostic for protein aggregation (43, 104, 114, 115, 117, 123, 131, 132)(Karch CM, Prudencio M, Winkler D, Hart PJ, Borchelt DR, In Press) (Chapter 4). SOD1 aggr egates are selectively detected in the spinal cord, and these aberrant st ructures are absent in muscle and other tissues (114). While SOD1 aggregation is absent in muscle tissue, muscle pathology, including denervation of the motor endplate (181), has been reported to occur early in the disease in FALS mice. Previous studies suggest that heat shock proteins may be important in modulating SOD1 aggregation (182). It has been proposed that a common feature of FALS-linked SOD1 mutants is an increased propensity to misfold (26, 106-111). An in vitro study of aggregation demonstrated that B crystallin inhibits mutant SOD1 aggr egation (23). This small heat shock protein binds to exposed hydrophobic surfaces on denatured or misfolded proteins (183-186), which allows B crystallin to inhibit protein aggr egation (187, 188). Oligomerization of B crystallin is associated with its activation as a chaperone (189). In spinal cords of FALS mice (G37R, G85R, G93A, H46R/H48Q, Quad, L126Z), increased levels of detergent-insoluble B crystallin are detected after the onset of symptoms (43, 115) In symptomatic L126Z mice, mutant SOD1 appears to specifically accumulate in motor neurons where B crystallin

PAGE 101

101 expression is absent (115). Together, this evidence suggests that, in vivo B crystallin may suppress SOD1 aggregation in select cell types. B crystallin is unique in that it is highly expressed in select tissues: skeletal muscle and oligodendrocytes (190, 191). B crystallin knock-out mice have been developed in which B crystallin gene and a portion of the HSPB2 gene are removed (HSPB2 was unidentified at the time of development) (192). HSPB2 (also termed myotonic dystrophy kinase bind protein) is highly expre ssed in the heart and muscle, specifically in Z membranes and neuromuscular junctions, and is ubiquitously expressed at lower levels (193, 194). HSPB2 acts as a molecular chaperone for m yotonic dystrophy protein kinase (193). B crystallin knock-out mice have no overt pathology until 40 weeks of age, at which time they develop muscular degeneration, kyphosis, a nd osteoarthritis (192). However, mice that retain one B crystallin allele do not develop an adverse phenotype (192). In this study, we used thr ee lines of SOD1 transgenic mice to study the effect of B crystallin on disease course, aggregate accumulation, and aggreg ate localization. Two of the models used, Gn.G37R and Gn.L 126Z mice, have mutations in human genomic SOD1 and are under the control of the huma n SOD1 promoter (23, 115, 147), producing ubiquitous expression of human SOD1. Both mice develop the charact eristic ALS phenotype of hindlimb paralysis, motor neuron loss, and accumulation of detergent-insoluble SOD1 in the spinal cord. In both lines of mice, misfolded forms of mutant SOD1 appear to accumulate selectively in motor neurons (115, 125, 147). Gn.L126Z mice differ fro m Gn.G37R mice in that the L126Z mutant has a short half-life: SOD1 protein that is not immediately degraded is converted into aggregates (43, 115). PrP.G37R mice have mutations in the human SOD1 cDNA and are under the control of the mouse prion promoter, which is predomin antly expressed in neuronal and muscle tissue (149). Mice that are heterozygous for PrP.G37R do not develop pathology by two years of age

PAGE 102

102 and do not form SOD1 aggregates; however, mi ce that are homozygous for PrP.G37R develop hindlimb paralysis, motor neuron loss, and dete rgent-insoluble SOD1 species, indicating that threshold levels of mutant SOD1 expression are required to induce disease in this mouse model (149). To study the role of B crystallin in SOD1 pathology, we crossed the three models of SOD1 transgenic mice described above to B crystallin knock-out mice and asked the following questions: (1) does the elimination of B crystallin alter the clinical appearance of disease in mutant SOD1 transgenic mice; (2) does elimination of B crystallin alter the abundance and tissue distribution of detergent-insoluble S OD1 species; and (3) doe s the elimination of B crystallin alter the cellular loca lization of pathologic lesions. The B crystallin knock-out mice develop a late onset muscular degeneration and to our surprise the presence of this second insult had little impact on progression of neurologic phenotypes of SOD1 mice. The levels and tissue distribution of detergent-insoluble SOD1 species which represent aggregated mutant SOD1, did not change in mice lacking B crystallin. Finally, the distribu tion of pathologic features was not changed by the elimination of B crystallin. We conclude that B crystallin is not normally involved in modulating the rate or subcellular distribution of mutant SOD1 aggregation or pathologic accumulation. Methods Tissue Culture Transfection and Transgenic Mice SOD1 mutants have been prev iously characterized (43). B crystallin cDNA (Clontech, Mountain V iew, CA, USA) was expressed in th e pEF-BOS vector (146) GFP cDNA (Clontech, Mountain View ,CA, USA) was expressed in the pcDNA3.1(A)-Myc vector (Invitrogen,

PAGE 103

103 Carlsbad, CA, USA). Transient transfection was performed in HEK293-FT cells as previously described (43, 46). The SOD1 transgenic mice used in this study have been previously characterized. Three mice were created in genomic SOD1 sequence unde r the control of the human SOD1 promoter: the G37R variant [line 29 (onset at 7-8 mo)] (147), the L126Z variant [line 45 (onset at 7-9 mo)](149), and the wild type (W T) variant (line 76) (147). One SOD1 transgenic mouse was created using SOD1 cDNA under the control of the mouse prion promoter: G37R [line 39 (asymptomatic)] (149). Mice homozygous null for B crystallin were obtained from Dr. Eric Wawrousek at the National Eye Institute and have been described previously (192). Mice were identified by PCR of tail DNA using primers described previously (115, 147, 149, 192). All procedures involving mice were reviewed an appr oved by the University of Florida Institutional Animal Care and Use Committee. SOD1 Aggregation Assay by Differential Extraction The procedu res used to assess SOD1 aggreg ation by differential de tergent extraction and centrifugation in spinal cords and muscle were si milar to previous descriptions (see Chapter 2) (43, 46). Protein concentration was measured in the detergentinsoluble and detergent-soluble fractions by BCA method as desc ribed by the manufacturer (Pie rce, Rockford, IL, USA). Immunoblotting Standard sodium dodecyl sulfate-polyacryl amide gel electrophoresis (SDS-PAGE) was performed in 18% Tris-Glycine gels (Invitrogen, Calsbad, CA, US A). Samples were boiled for 5 minutes in Laemmli sample buffer pr ior to electrophoresis (148). Quantitative Analysis of Immunoblots Quantification of SOD1 protein in detergent-insoluble and de tergent-soluble fractions was perform ed by measuring the band intensity of SOD1 in each lane using a Fuji imaging system

PAGE 104

104 (FUJIFILM, LifeScience, Stamford, CT, USA) Aggregation propensity was measured by comparing the ratio of band intensity in the dete rgent-insoluble fraction to the band intensity in the detergent-soluble fraction. The standard error of the mean (S EM) was calculated for aggregation of each sample in e ach experiment. Statistical significance was measured using an unpaired t-test. Antibodies Immunoblots were probed with rabbit polycl onal antibodies term ed hSOD1 and m/hSOD1 at dilutions of 1:2500. The hSOD1 antiserum was raised against a synthetic peptide that is specific to human SOD1 (aa 2436) (113). The m/hSOD1 antiserum was raised against a synthetic peptide conserved be tween mouse and human SOD1 ( aa 124-136) (24). Immunoblots were also probed with a mouse polyclonal antibodyB crystallin (Stressgen, Ann Arbor, MI, USA) at dilutions of 1:10 00; rabbit polyclonal antibody-Hs p40 (Stressgen, Ann Arbor, MI, USA) at dilutions of 1:2000; and rabbit polyc lonal antibody-Hsp70 (Stressgen, Ann Arbor, MI, USA) at dilutions of 1:1000. Tissue was stained with antibodies to hSOD1 (1:2500), GFAP (1:500 Chemicon, Billerica, MA, USA), NeuN (1:500Chemicon, Billerica, MA, USA), and B crystallin (1:1000Sressgen, Ann Arbor, MI, USA). Secondary antibodies used included goat-anti-mouse AlexaFluor 488 (Invitrogen, Calsbad, CA) and goat-anti-rabbit AlexaF luor 568 (Invitrogen, Calsbad, CA).

PAGE 105

105 Figure 5-1. B crystallin reduces mutant SOD1 aggreg ation in cell culture. HEK293-FT cells were transfected with mutant SOD1 (4 g) alone, with mutant SOD1 (2 g) and B crystallin (2 g), or with mutant SOD1 (2 g) and GFP (2 g). Cell lysates were extracted in non-ionic dete rgent and run on 18% Tris-Glycine gels. Immunoblots were probed with m/hSOD1 antiserum. A) Detergent-insoluble (20 g). B) Detergent-soluble (5 g). C) Quantification of aggreg ation propensity measured as a ratio of detergent-insoluble to soluble and expressed as SEM. *, significantly different from mutant SOD1 (p<0.05). significantly different from WT SOD1 (p<0.05). The image shown is representative of 3 repetitions of the experiment. Data from at least 3 experiments were used to quantify aggregation propensities in panel C. Note: due to low expression levels at 24 hours, cells expressing G37R, G37R/ B, and G37R/GFP were harvested 48 hours after transfection. Histology After inducing deep anesthesia, m ice were transcardially perfused with cold PBS and then 4% paraformaldehyde in 1x PBS. After 48 hours of post-fixation, tissues were transferred to 30% sucrose for at least 48 hours prior to cryostat sectioning (Microm HM550). Coronal sections of spinal cord (14 microns) were cut and stored in anti-freeze solution (100 mM sodium acetate, 250 mM polyvinyl pyrolidone, 40% ethylene glycol, pH 6). Sections were stored at 20C. To stain tissue, sections were washed in PBS prior to blocking in 5% normal goat serum

PAGE 106

106 (Invitrogen, Calsbad, CA). Sections were then incubated in primary and secondary antibodies with PBS, 5% normal goat serum, and 1% triton. Images were captured using an Olympus IX81-DSU Spinning Disk c onfocal microscope. Results Sm all heat shock proteins, such as B crystallin, aid in the folding of misfolded protein in the cell (195). Because mutations in SOD1 predispose the protein to misfold and form aggregates, we believed that a sm all heat shock protein such as B crystallin could reduce the aggregation of mutant SOD1 protei n. Additionally, there is evidence in vitro that B crystallin can modulate mutant SOD1 aggregation (23). To assess the effect of B crystallin on mutant SOD1 aggregation, vectors encoding FALS mu tants (G85R, A4V, G37R) were transfected alone, with vectors encoding B crystallin (ratio of 1:1), or w ith vectors encoding GFP (ratio of 1:1) in HEK293-FT cells. The cell lysates we re fractionated into detergent-insoluble and detergent-soluble fractio ns by sonication and high speed cen trifugation in non-ionic detergent (46). In this study, SOD1 aggregates are defi ned as protein that is insoluble in non-ionic detergent and sediments upon centrifugation. All FALS-linked SOD1 mutants formed detergentinsoluble species (Fig. 5-1A, la nes 5, 8, 11) and had aggregat ion propensities significantly different from WT protein (Fi g. 5-1C). The co-expression of B crystallin with each SOD1 mutant (G85R, A4V, G37R) reduced aggregation (Fig. 5-1A, lanes 6, 9, 12) to levels that were not significantly different from WT (Fig. 5-1C). To control fo r non-specific effects on mutant SOD1 expression that could be caused by co-transfection of the B-crystallin expression vectors, vectors expressing mutant SOD1 were co-trans fected with vectors expressing GFP cDNA. There was a slight reduction in mutant SOD1 protein in the dete rgent-insoluble protein fraction due to the reduced amount of mu tant SOD1 starting material (2 g versus 4 g) (Fig. 5-1A, lanes

PAGE 107

107 7, 10, 13); however, the aggregation propensity of mutant SOD1 (ratio of soluble to insoluble protein) expressed with GFP was similar to mutant SOD1 expressed alone (Fig. 5-1C). Under all conditions, a significant amount of mutant SOD1 was found in th e detergent-soluble fraction, which likely represents the protein that is folded in a normal or near normal conformation (Fig. 5-1B). Thus, B crystallin specifically reduces mutant SOD1 aggregation in cell culture. The active form of B crystallin is oligomeric and in soluble in detergent (189). Cell lysates used to measure SOD1 aggregation were also immunoblotted wi th a polyclonal antibody to human B crystallin. In cells co-transfected w ith vectors expressing mutant SOD1 (G85R, A4V, and G37R) and B crystallin, B crystallin was highly expre ssed in the detergent-soluble protein fraction (Fig. 5-2B, lanes 4, 7, 10) and in the detergent-insoluble protein fraction (Fig. 52A, lanes 4, 7, 10). However, B crystallin was undetectable in cells transfected with vectors expressing mutant SOD1 alone or in cells co-transfected with vectors expressing mutant SOD1 Figure 5-2. Overexpressed B crystallin is upregulated in response to mutant SOD1 in cell culture. Detergent-insoluble protein fracti ons and detergent-sol uble protein fractions from Figure 1 were run on 18% Tris-Glycine gels and immunoblotted with B crystallin antiserum. A) Detergent-insoluble (20 g). B) Detergent-soluble (5 g). Note: due to low expression levels at 24 hours, cells expressing G37R, G37R/ B, and G37R/GFP were harvested 48 hours after transfection.

PAGE 108

108 and GFP (Fig. 5-2). Thus, the robus t expression and oligomerization of B crystallin accounts for the low levels of mutant SOD1 aggregates in HEK293-FT cells. Also note, HEK293-FT cells do not constitutively express B crystallin nor can these cells be induced to express B crystallin in response to mutant SOD1 expressi on. To more extensivel y assess the heat shock response in HEK293-FT cells, we incubated the cells at 42 C for 30 minutes and collected cells at multiple times points post heat shock. Im munobloting for Hsp70 and Hsp40 revealed that Hsp70 was highly induced prior to heat shock and remained highly induced 24 hours post-heat shock (Fig. 5-3). Hsp40 was cons titutively induced at levels sli ght lower than Hsp70 (Fig. 5-3). In spite of high levels of Hsp70 and Hsp40 in th is cell culture system, mutant SOD1 robustly aggregates. Figure 5-3. Hsp40 and Hsp70 are constitutively indu ced in HEK293-FT cells. Cells were not exposed to heat shock or heat shocked at 40 C for 30 minutes and harvested at 5 hours or 24 hours post heat shock. Cell lysate s were sonicated in PBS and run on an 18% Tris-Glycine gel. Immunoblots were probed with Hsp40 and Hsp70 antibodies. Image is representative of 3 repetitions of the experiment. Using available mice with the targeted deletion of B crystallin, we asked whether reducing or eliminating B crystallin in mutant SOD1 transgenic mice would alter the disease course and, or, change the location of SOD1 aggregates. Because B crystallin null mice develop myopathy at 40 week s (192), we asked whether muscle degeneration could compound the ALS phenotype in mice overexpressing mutant SOD1. Denervation of the motor endplate occurs early in the disease course (176), prior to SOD1 aggregation and hindlimb paralysis (Karch CM, Prudencio M, Winkler D, Hart PJ, Borchelt DR, In Press). Mutant SOD1 transgenic

PAGE 109

109 mice (Gn.G37R, Gn.L126Z, PrP.G37R) were cro ssed to mice that were homozygous null for B crystallin (termed B crystallin knockout (KO) mice here after) to produce mutant SOD1 mice in which B crystallin was reduced (heterozygous; +/-) or eliminated (KO: -/-). In all three lines of mutant SOD1 transgenic mice (Gn.G37R, Gn .L126Z, PrP.G37R), the reduction (+/-) or the elimination (-/-) of B crystallin resulted in a modest ch ange in disease course (Fig. 5-4). Gn.G37R and Gn.L126Z mice develope d characteristic hindlimb para lysis. Gn.G37R mice with reduced B crystallin (n=17, mean: 204 days) and no B crystallin (n=12, mean: 195 days) had lifespans significantly different from Gn.G37R mice with WT levels of B crystallin (n=15, mean: 214.5 days; p<0.0001, p=0.001, respectively) (F ig. 5-4A). The lifespans of Gn.G37R mice with reduced B crystallin were not differe nt from Gn.G37R mice with no B crystallin (p=0.6919). Gn.L26Z mice with reduced B crystallin (n=8, mean: 225 days) had lifespans similar to Gn.L126Z mice e xpressing WT levels of B crystallin (n=8, mean: 210 days; p= 0.5552) (Fig. 5-4B). Gn.L126Z mice without B crystallin (-/-) died earlier than Gn.L126Z mice expressing WT levels of B crystallin and Gn.L126Z mice expressing reduced levels of B crystallin (n=8, mean: 180 days; p<0.0001, p=0.0145, re spectively) (Fig. 5-4B). Thus, reduction or elimination of B crystallin is capable of altering dis ease course in SOD1 transgenic mice; however, this was not a robust effect. It is possible that B crystallin is only a minimal modifier in disease course. To more stringently test the role of B crystallin in FALS disease course, we used mice that express G37R under the control of the mouse prion promoter (PrP.G37R). Mice that are heterozygous for the PrP.G37R transgene do not develop FALS symptoms, whereas a 2-fold increase in expression by gene rating homozygous animals produces disease (149). Eliminating B crystallin in heterozygous PrP.G37R mice did not induce FALS symptoms (mice were

PAGE 110

110 harvested at 40 weeks due to symptoms associated with loss of B crystallin). Thus, in PrP.G37R mice, the second insult provided by the elimination of B crystallin did not create enough of a burden to alter the disease progression in asymptomatic mutant SOD1 mice. We conclude that B crystallin may play a role as one m odifier of mutant SOD1 misfolding and toxicity, but it does not appear to be a critical modifier. Figure 5-4. Reduction or elimination of B crystallin in mutant SOD1 transgenic mice does not substantially alter survival. Kaplan-Meier survival curves of mutant SOD1 transgenic mice with WT (black), heterozygous (red), or homozygous null (blue) B crystallin. A) Gn.G37R mice. B) Gn.L126Z mice. Note: PrP.G37R mice did not develop symptoms (n=10 per condition). Because the elimination of B crystallin altered the disease course, we were interested in the effects of B crystallin on aggregate accu mulation and aggregate location in vivo In mutant SOD1 mice, SOD1 aggregates are absent in muscle and are only detected in the spinal cord and brainstem (Fig. 5-5, lanes 2 and 3) (115, 147, 149) We hypothesized that the high level of expression of B crystallin in muscle could protect this tissue from the accumulation of SOD1 aggregates. To test this hypothesis, hindlimb muscle from Gn.G37R, Gn.L126Z, and PrP.G37R ( B +/+, B +/-, B -/-) was extracted in non-ionic de tergent and centrifuged at 100,000xg to isolate detergent-insoluble and detergent-soluble fractions. In muscle from symptomatic mice (Gn.G37R and Gn.L126Z), SOD1 was not detected in the detergent-insoluble fraction when B crystallin was reduced (+/-) or eliminated (-/-) (F ig. 5-6A, lanes 3-4, 7-8). In these tissues, all

PAGE 111

111 mutant SOD1 was found in the detergent-soluble fraction (Fig. 5-6B, lanes 3-6); however, only faint bands were apparent for Gn.L126Z, due to th e short half-life of th is protein (Fig. 5-6B, lanes 7-8). In asymptomatic mice, (PrP.G37R ), SOD1 was not detected in the detergentinsoluble fraction when B crystallin was reduced (+/-) or elim inated (-/-) (Fig. 5-6A, lanes 5-6); however, detergent-so luble SOD1 was present in muscle tiss ue (Fig. 5-6B, lanes 5-6). Thus, the reduction or elimination of B crystallin in SOD1 transgenic mice does not alter the tissue specific accumulation of detergent-insoluble SOD1. Figure 5-5. SOD1 aggregation is restricted to the brainstem and spinal cord in Gn.L126Z mice. Organs and nervous tissues were extracted in non-ionic detergent and run on 18% Tris-Glycine gels. Immunoblots were probe d with hSOD1 antiserum. A) Detergentinsoluble (20 g). B) Detergent-soluble (5 g). Image is representative of 2 repetitions of the experiment. Because the disease course wa s slightly accelerated without a change in localization of mutant SOD1 aggregates, it was still possibl e that the reduction or elimination of B crystallin could alter the abundance of SOD1 a ggregates in the spinal cord tis sue. The spinal cord is the prominent site of pathology in mutant SOD1 transgenic mice (115, 147, 149). In symptomatic mutant SOD1 transgenic mice, detergent-insolubl e SOD1 aggregates are detected exclusively in the brainstem and spinal cord (180) (Fig. 5-5A, la nes 2 and 3). We measured detergent-insoluble SOD1 in spinal cord tissue in each variant usi ng the detergent extraction assay described above.

PAGE 112

112 Gn.G37R and Gn.L126Z, which normally form dete rgent-insoluble SOD1, showed little change in the overall levels of de tergent-insoluble SOD1 when B crystallin was reduced (+/-) or eliminated (-/-) (Fig. 5-7A, lanes 3-5 and 911, respectively). PrP.G37R, which does not form aggregates when heterozygous for the transgene, produced levels of dete rgent-insoluble SOD1 similar to or less than WT SOD1 (Fig. 5-7A, co mpare lanes 1 and 6-8). Aggregation propensity of the mutant SOD1 did not differ among B crystallin genotypes in all three lines of mice (Fig. 5-8; Fig. 5-9). SOD1 WT mice and B crystallin KO mice containe d little or no detectable SOD1 in the detergent-insoluble fraction (Fig. 5-7A lanes 1 and 2). Levels of detergent-soluble protein were detected at s lightly varying levels in G n.G37R and PrP.G37R in all B crystallin variants; however, no consistent change was detected when B crystallin was reduced or eliminated (Fig. 5-7B, lanes 3-8). Levels of detergent-soluble SOD1 in Gn.L126Z mice were virtually undetectable due to th e short half-life of this varian t (115) (Fig. 5-7B, lanes 9-11). Thus, the reduction or elimination of B crystallin does not alter the abundance of detergentinsoluble SOD1 in the spinal cords of mutant SOD1 transgenic mice. When B crystallin is activated, it forms a large complex that is insoluble in detergent (43, 115, 196) (Fig. 5-2). Using the detergent extracti on assay described above, detergent-insoluble and detergent-soluble protein was isolated and immunoblotted to det ect the presence of B crystallin in spinal cord tissue. In all three lines of mice, B crystallin was predominantly detected in the detergent-insoluble fraction when wild-type (+/+) or re duced (+/-) levels of B crystallin were present in the tissue (Fig. 5-10A, lanes 3-4, 6-7, 910). Only in the presence of wild-type levels of B crystallin could B crystallin be prominently detected in the detergentsoluble fraction (Fig. 5-10B, lanes 3,6,9). The levels of B crystallin were the highest in Gn.G37R mice, despite similar load ing levels (Fig. 5-10A, lanes 3,4) This is consistent with

PAGE 113

113 previous studies that measured B crystallin in mutant SOD1 transgenic mice (43, 115, 196). Our results further suggest that mutant SOD1 can activate B crystallin and that some mutants are better substrates for B crystallin than other mutants. Figure 5-6. SOD1 aggregation is absent in muscle ti ssue. Muscle tissue from hindlimbs were isolate from mice and extracted in non-i onic detergent and run on 18% Tris-Glycine gels. Immunoblots were probed with hSOD1 antiserum. A) Detergent-insoluble (20 g). B) Detergent-soluble (5 g). The image shown is representative of 2 repetitions of the experiment. To examine possible changes in the types of neural cells that accumulate misfolded SOD1 when B crystallin levels were reduced (+/-) or elim inated (-/-), spinal cords from each variant were fixed, and frozen sections were stained using cell-type speci fic markers. In this study, we used an antiserum that recognizes a peptide from amino acids 24-36 that is unique to human SOD1 (113). To determine if this antibody exclus ively recognizes misfolded SOD1 protein or if it is also able to recognize normally folded SOD1, we immunoprecipita ted purified WT SOD1 protein (made by D. Winkler in th e laboratory of Dr. P. John Hart University of Texas Health Science Center at San Antonio) with the hS OD1 antibody (Fig. 5-11). We found that purified WT SOD1 protein could only be captured with the hSOD1 antibody when the protein was

PAGE 114

114 denatured (Fig 5-11). Thus, it is likely that th is antibody recognizes protein that is not in its native state; however, in fixed tissues, we do not know whether natively folded SOD1 could be immunoreactive. Thus, our study sought to dete rmine whether the general pattern of SOD1 immunoreactivity differed between mice that express wild-type levels of B crystallin and mice that express reduced or no B crystallin. To identify cell types within these tissues, we used standard cell markers (GFAP and NeuN) to id entify astrocytes a nd neurons, respectively. Figure 5-7. Reduction or elimination of B crystallin in SOD1 transgenic mice does not alter aggregation in spinal cord tissue. Spinal cords were extracted in non-ionic detergent and run on 18% Tris-Glycine gels. Immunoblots were probed with hSOD1 antiserum. A) Dete rgent-insoluble (20 g). B) Detergent-soluble (5 g). The image shown is representative of 3 repetitions of the experiment. In Gn.G37R (Fig. 5-12) and Gn.L 126Z mice (Fig. 5-13), and in each B crystallin variant (+/+, +/-, -/-), no novel appearance of SOD1 immunoreactivity in activated astrocytes was detected. Astrocytes appeared to surround SOD1 positive motor neurons (Figs. 5-12 and 5-13, panels A-C, G-I, M-O, S-U). In all variants of PrP.G37R mice, li ttle GFAP staining was detected, which was consistent with the lack of overall pathology, including astrocytosis, in these mice (Fig. 5-14, panels A-C, G-I, M-O, S-U). To determine whether SOD1 accumulates in motor neurons, tissue from each SOD1 mutant with each B crystallin variant (+/+, +/-, -/-) were co-stained for human SOD1 and NeuN, a neuronal marker. In Gn.G37R (Fig. 5-12, panels D-F,

PAGE 115

115 J-L, P-R, V-X) and Gn.L126Z (Fig. 5-13, panels D-F, J-L, P-R, V-X) (in mice expressing all levels of B crystallin), SOD1 predominantly co-lo calized with NeuN positive neurons. In asymptomatic PrP.G37R B crystallin variants (Fig. 5-14, panels D-F, J-L, P-R, V-X), human SOD1 immunoreactivity co-l ocalized with NeuN positive neurons. Thus, when B crystallin is reduced (+/-) or eliminated (-/-), there are no obvious changes in the location of SOD1 immunoreactivity compared with SOD1 transgenic mice expressing wild-type levels of B crystallin. Figure 5-8. SOD1 aggregation propensity in mice expressing varying levels of B crystallin. The ratio of detergent-insol uble to detergent-soluble SOD1 was measured in spinal cord tissue (see Fig. 5-7) and each measurement was graphed. significantly different from WT SOD1 (p<0.05) A) Gn.G37R. B) Gn.L126Z. Previously, B crystallin has been described to become induced in astrocytes once mice become symptomatic (115). Thus, we sought to examine the location of B crystallin in all the variants of mice. Consistent with what we have previously shown, in symptomatic Gn.L126Z mice, B crystallin was upregulated primarily in astrocytes, with some staining of oligodendrocytes (Fig. 5-15, panels GH). However, in Gn.G37R mice, B crystallin was primarily present in oligodendrocytes (Fig. 5-15, panels A-B). Similarly, in PrP.G37R mice, B crystallin was only detected in oligodendrocytes (Fig. 5-15, panels D-E). Only background B crystallin immuostaining was detected in SOD1 transgenic mice in which B crystallin was eliminated (Fig. 5-15, panels C, F, I). Thus, induction of B crystallin in astrocytes is not

PAGE 116

116 universal to all the SOD1 mouse models, indicating that there must be a unique signal presented by the accumulating mutant L126Z SOD1. Figure 5-9. SOD1 aggregation propensity in PrP.G37R mice expressing varying levels of B crystallin. The ratio of detergent-insolubl e to detergent-soluble SOD1 was measured in spinal cord tissue (see Figures 4 and 5) and measured as SEM. Discussion In this study we examined the role of B crystallin as a potential modifier of mutant SOD1 misfolding. First, in cell culture, we found that B crystallin selectively reduced mutant SOD1 aggregation, indicating that B crystallin possesses the cap acity to modulate mutant SOD1 misfolding. Using available mi ce with the targeted deletion of B crystallin, we asked whether reducing or eliminating B crystallin in mutant SOD1 transgenic mice would alter the disease course and, or, change parameters of mutant SOD1 aggregation. Elimination of B crystallin in Gn.G37R and G n.L126Z mice modestly shortened the interval (statistically validated) to which these mice reach human e ndpoints (obvious paralysis). Eliminating B crystallin in mice that express G37R SOD1 at levels below th e threshold for inducing disease was not sufficient to produce FALS-like sympto ms. Muscle tissue, which does not accumulate aggregates in mutant SOD1 mice, contained no detectable aggregates in the absence of B crystallin. Similarly, the absence of B crystallin did not alter the amount of detergentinsoluble, sedimentable SOD1 that accumulated in the spinal cord. We did not observe an obvious change in the distribution of SOD1 imm unoreactivity that accumulated in spinal cords

PAGE 117

117 of mutant mice lacking B crystallin. Together, this study illustrates that B crystallin is not a major factor in modulating the formation of t oxic forms of mutant SOD1 or in moderating mutant protein aggregation. Figure 5-10. B crystallin is upregulated in SOD1 transg enic mice. Spinal cords were extracted in non-ionic detergent and run on 18% Tris -Glycine gels. I mmunoblots were probed with B crystallin antiserum. A) Detergent-insoluble (20 g). B) Detergent-soluble (5 g). The image shown is representative of 3 repetitions of the experiment. Myopathy Associated with ALS Disease Course Muscle pathology has been described early in the disease course in m utant SOD1 transgenic mice (181). The B crystallin KO mice we used develop a profound myopathy (192), which we anticipated could synergize with the activ ities of mutant SOD1 to hasten the onset of disease. Gn.G37R and Gn.L126Z B crystallin KO mice developed symptoms slightly earlier than mutant mice expressing wild-type levels of B crystallin without a corresponding change in overt phenotype and without a change in aggregate accumulation This slight change in disease course may be due to an interaction between compromised muscle and neurotrophic disease induced by mutant SOD1. The absence of B crystallin in SOD1 transgenic mice did not alter the accumulation of detergent-insoluble SOD1 aggr egation in muscle, and it did not appear that the myopathic phenotypes were hastened by the e xpression of mutant SOD1. However, it is

PAGE 118

118 possible that the absence of B crystallin increases oxidative stre ss or other pathology that could hasten disease course. Dobrow olny and collegues recently show ed that the specific expression of mutant SOD1 in muscle resulted in a myopa thy without motor neuron di sease (139). If the loss of B crystallin interacted with mutant SOD1 to worsen the myopathic disease, the effect was not sufficiently robust to be noticeable. Figure 5-11. SOD1 antibodies recognize denatu red forms for SOD1. WT purified protein was incubated with hSOD1, m/hSOD1, and whole protein SOD1 antiserum in detergent (0.25% SDS, 0.5% NP40, 0.5% DOC) and in th e presence or absence of heat (95 C). Immunoblots were probed with whole protein SOD1 antiserum. A) Binding fraction (10 l). B) Non-bindi ng fraction (10 l). The image shown is representative of 3 repetitions of the experiment. The Role of Heat Shock Protei ns in the ALS Disease Course When activated, B crys tallin forms large heterogeneous multimers that range in size from 300 to 1000 kDa and can contain as many as 40 subunits (197). By binding to exposed hydrophobic surfaces on denatured or misfolded proteins, B crystallin inhibits protein aggregation (183-188). Our fi nding in cell culture that B crystallin reduced mutant SOD1 aggregation and that B crystallin is upregulated into the detergent-insoluble fraction when cotransfected with mutant SOD1 suggests that B crystallin binds to misfolded mutant SOD1 and

PAGE 119

119 prevents aggregation. However, we do not have conclusive evidence that B crystallin stably binds to mutant SOD1. In ALS, motor neurons are less capable of i nducing heat shock proteins (198). In mouse models of FALS, Hsp40, Hsp60. Hsp70, and Hs p90 remain unchanged throughout the disease course (196). Only the small heat shock proteins Hsp25 and B crystallin are upregulated in symptomatic SOD1 transgenic mice (196). Previous studies investigating the effects of heat shock proteins on ALS have shown little, if a ny, effect on disease course and progression (199201). However, these studies focus largely on Hsp27 and Hsp70, which are normally expressed in motor neurons; thus, overexpression of these proteins may not enhance the ability of the cell to combat accumulating mutant SOD1. Our findings that mutant SOD1 aggregat es in cells in which Hsp70 and Hsp40 are constitutively, highly expressed, suggests that Hsp70 and Hsp40 may be less capable of inhibiting SOD1 aggregation. However, it is also possible that Hs p70 and Hsp40 have little effect on SOD1 aggregation in cell culture due to the high levels of overexpression of the mutant protein in this model. Disease Threshold in SOD1 Transgenic Mice To determine whether B crystallin is capable of altering disease course and pathology in SOD1 transgenic mice, an asymptomatic tr ansgenic mouse, PrP.G37R, was studied. Heterozygous PrP.G37R mice do not develop AL S pathology and do not form SOD1 aggregates; however, homozygous PrP.G37R develop hindlimb paralysis and form SOD1 aggregates in the spinal cord (149). Thus, the acquisition of th e ALS phenotype in our mouse model is based on threshold levels of mutant SOD1. Using hetero zygous PrP.G37R mice, we asked if reducing or eliminating B crystallin could lower the threshold of mutant SOD1 expression that is required

PAGE 120

120 Figure 5-12. In Gn.G37R mice, B crystallin does not alter localization of SOD1 accumulation. Mice were perfused with 4% paraformaldehyde and spinal cord tissue was immersed in sucrose prior to cryostat sectioning ( 14 microns). Sections were stained with hSOD1 antiserum (A, D, G, J, M, P, S, V) and GFAP (B, H, N, T) or NeuN (E, K, Q, W). Sections were stained with seco ndary fluorescent antibodies: anti-rabbitAlexaFluor 568 (A, D, G, J, M, P, S, V) and anti-mouse-AlexaFluor 488 (B, E, H, K, N, Q, T, W). Ventral horn. 40x magnificat ion. The image shown is representative of 4 repetitions of the experiment.

PAGE 121

121 Figure 5-13. In Gn.L126Z mice, B crystallin does not alter localization of SOD1 accumulation. Mice were perfused with 4% paraformaldehyde and spinal cord tissue was immersed in sucrose prior to cryostat sectioning ( 14 microns). Sections were stained with hSOD1 antiserum (A, D, G, J, M, P) and GFAP (B, H, N) or NeuN (E, K, Q). Sections were stained with secondary fl uorescent antibodies: an ti-rabbit-AlexaFluor 568 (A, D, G, J, M, P) and anti-mouse-Ale xaFluor 488 (B, E, H, K, N, Q). Ventral horn. 40x magnification. The image shown is representative of 4 repetitions of the experiment.

PAGE 122

122 Figure 5-14. In PrP.G37R mice, B crystallin does not alter localization of SOD1 accumulation. Mice were perfused with 4% paraformaldehyde and spinal cord tissue was immersed in sucrose prior to cryostat sectioning ( 14 microns). Sections were stained with hSOD1 antiserum (A, D, G, J, M, P) and GFAP (B, H, N) or NeuN (E, K, Q). Sections were stained with secondary fl uorescent antibodies: an ti-rabbit-AlexaFluor 568 (A, D, G, J, M, P) and anti-mouse-Ale xaFluor 488 (B, E, H, K, N, Q). Ventral horn. 40x magnification. The image shown is representative of 4 repetitions of the experiment.

PAGE 123

123 Figure 5-15. B crystallin is upregulated in astrocytes of Gn.L126Z mice. Mice were perfused with 4% paraformaldehyde and spinal cord tissue was immersed in sucrose prior to cryostat sectioning (14 microns). Sections were stained with B crystallin. Sections were stained with secondary fluorescent antibodies: anti-rabbit-AlexaFluor 568. Ventral horn. 40x magnification. The image s hown is representative of 4 repetitions of the experiment. to induce disease. Reduction or elimination of B crystallin did not produce a phenotype or SOD1 aggregation in PrP.G37R mi ce (data not shown). These results suggest that the absence of B crystallin does not produce enough of a burde n to induce disease. However, we cannot disregard evidence in symptomatic SOD1 tr ansgenic mice (Gn.G37R and Gn.L126Z), which developed the ALS phenotype at si gnificantly earlier ages when B crystallin was reduced or

PAGE 124

124 eliminated. Collectively, these mice demonstrate that B crystallin is likely only a modest modifier of disease phenotype. Conclusions In cell culture, sm all heat shock proteins are robust modifiers of SOD1 aggregation. However, heat shock proteins show m odest effects on the ALS disease course in vivo Thus, it is possible that the induction of one small heat shock protein alone is not robust enough to modify disease. It may be possible th at heat shock proteins can work in concert to prevent protein misfolding and mutant SOD1 toxicity.

PAGE 125

125 CHAPTER 6 CONCLUSIONS SOD1-linked FALS is characterized by aggregat ion of the mutant SOD1 protein. To date, all FALS-lin ked SOD1 mutants studied in cell cu lture share a propensity to form aggregates, despite variable biophysical characteristics of i ndividual SOD1 mutants. All SOD1 transgenic mice that overexpress FALS-linked SOD1 mutant s develop hindlimb paralysis, motor neuron loss, and SOD1 aggregates in the brainstem and spinal cord. Wild-type (WT) SOD1 does not form SOD1 aggregates in cell culture, and WT SOD1 transgenic mice do not develop the ALS phenotype. Despite the apparent importance of SOD1 aggregates the mechanisms of mutant SOD1 aggregate formation and how these structures contribute to diseas e pathogenesis is poorly understood. In this study, we hypot hesized that SOD1 aggregates mediate ALS. We used cell culture and mouse models to st udy the factors that mediate SOD1 aggregate formation and to understand the role of SOD1 aggregat es in the ALS disease course. Composition of Mutant SOD1 Aggregates In these stud ies, we provide evidence that i nherent structural aspects of the SOD1 protein mediate aggregate formation. Previous work has suggested that SOD1 aggregates are predominantly stabilized by high-molecular-weigh t, disulfide cross-linke d species (23, 141). Through the manipulation of the four cysteine residues in the SOD1 protein, we found that cysteine 6 and cysteine 111 are important for mediating SOD1 aggregate formation and that these residues mediate aggregation by a mechan ism other than disulfide cross-linking (Chapter 2). However, we demonstrated that cysteine residues are not required fo r SOD1 aggregation to occur (Chapter 2). We provide evidence that eliminating cysteine 111 in the context of a highly aggregating FALS mutant significantly reduces aggregate formation (Chapter 2). Some groups suggest that

PAGE 126

126 cysteine 111 is vulnerable to oxidative modification or aberrant copper binding (152-155), but it is also possible that cysteine 111 works in con cert with other residues in the SOD1 protein to promote misfolding. We provide evidence that a ggregation is enhanced in FALS mutants via residues in -strand 6 and -strand 7, which includes cyst eine 111 (Chapter 3). Because strands 6 and 7 contain structures that are vital for stability of the protein, including copper binding sites and portions of the dimer interface (38), and because we have yet to identify a single residue that is required for SOD1 aggregation, we propose that SOD1 mutations result in global misfolding of the protein, me diated by destabilization between -strand 6 and -strand 7. To define the composition of SOD1 aggregat es, we studied spinal cord tissues from symptomatic mutant SOD1 transgenic mice. We demonstrate that aggregates at disease endstage are predominantly composed of disulfide-reduced SOD1 protein (Chapter 4). It has been suggested that SOD1 must bind copper and zinc before the in tramolecular disulfide bond can form between cysteine 57 and cysteine 146. Thus immature forms of the protein (with respect to metal binding and disulfide bond formation) are mo re prone to aggregate. It is possible that the immature forms of the mutant protein are suff iciently degraded early in the disease; however, over time, these species accumulate and become pref erentially incorporated into aggregates that are prominent at disease endstage (Fig. 6-1). Whether these imma ture forms of the protein also form soluble oligomers remains unknown. Despite evidence that disulfide cross-linked species are prominent at disease endstage (23, 141), this work demonstrates that disulfide cross-linking is not required for aggregate stabilization. Aggr egates remain intact in the presence of high concentrations of reducing agents which dissociate the disulfide cr oss-links (Chapter 4). Thus, we suggest that aberrant disulfide bonding occu rs after the aggregates are stabilized by other bonding forces (Fig. 6-1). SOD1 aggregat es may become stabilized by extensive -strand

PAGE 127

127 stacking, as many SOD1 transgenic mice have Th ioflavin-S positive inclusions in affected tissues (104). Figure 6-1. Mutant SOD1 folding pathways throughout the ALS disease course. Proposed folding pathways of mutant SOD1. When mice are asymptomatic (A), misfolded SOD1 protein is sufficiently degraded. It is possible that toxic, soluble oligomers are present and impart toxicity. As mice appro ach disease onset (B), detergent-insoluble aggregates are present. These aggregates become more extensively disulfide crosslinked as the mice approach disease endstage (C). Arrow thickness reflects relative abundance. +, proposed SOD1 oligomers that are larger th an two monomers. The Role of SOD1 Aggregates in ALS Disease Course Understanding the role of SOD1 aggregati on in disease course is crucial to our understanding of the disease mech anism. In mouse models of ALS, we found that aggregation occurs late in the disease (Chapter 4). Low levels of SOD1 aggregates were detected just prior to overt paralysis, suggesting that these species are important for ac quisition of disease phenotype (paralysis) (Chapter 4). SOD1 aggregate progression was similar in mutant SOD1 transgenic mice that expressed mutants with a range of bi ophysical characteristics. Furthermore, we found that high-molecular-weight, disulfide cross-linke d SOD1 species formed concurrently with detergent-insoluble aggregates at disease ends tage, which provides further evidence that

PAGE 128

128 disulfide cross-linking is a sec ondary structural feature of aggregates (Chapter 4). Figure 6-2. Disease progression in SOD1 transgenic mice. During presymptomatic stages of the disease, portions of SOD1 are normally fold ed (blue circles) with a small fraction of SOD1 protein that does not have the in tramolecular disulfide bond (red circles). Pathology including gliosis and motor endplat e denervation are prominent prior to disease endstage. After dis ease onset, SOD1 aggregates (detergent-insoluble SOD1) are prominent in addition to disulfide cross-linked SOD1 aggregates. Pathology including muscle atrophy, reduced chaperone function, activation of small heat shock proteins, and paralysis occur at disease endstage. Soluble SOD1 oligomers may be toxic at disease stages; however, currently, we do not have techni ques to isolate these species. We propose that at disease onset, th e formation of SOD1 aggregates triggers enhanced misfolding of immature SOD1 monomer, which becomes preferentially incorporated into the aggregat e. This feed forward loop re sults in a rapid increase in aggregates and disease endstage. Despite the robust presence of aggregates at disease endstage in SOD1 transgenic mice, we found that abnormal pathology occurs prior to the formation of detergent-insoluble SOD1 aggregates, including denervation of the motor endplate, gliosis, and motor neuron loss (Chapter 2). Together, our studies suggest that early toxic events occur in SOD1 transgenic mice that are

PAGE 129

129 not directly associated with the presence of de tergent-insoluble SOD1 aggregates (Fig. 6-2). Yet, it remains possible that other forms of mutant SOD1 are toxic early in the disease which are distinct from detergent-insoluble SOD1 aggreg ates or which are precursors to the species detectable at disease ends tage (Figs. 6-1 and 6-2). Modifiers of SOD1 Aggregation and FALS Pathogenesis From our study in mouse models of ALS, SO D1 aggregation appears to coincide with disease endstage. Because all mice that develop hindlimb paralysis have detectable levels of mutant SOD1 aggregates in the brainstem and spinal cord, aggregation remains an important pathological feature of the diseas e. Several groups have descri bed inhibition of aggregation in cell culture with the ap plication of heat shock proteins (131); however, these effects have translated poorly in vivo (199-201). We chose an alternativ e approach: to study a small heat shock protein that is upregulated at disease endstage and not expr essed in affected tissues. In cell culture, we found that th e small heat shock protein B crystallin was a strong modifier of mutant SOD1 aggregation (Chapter 5). However, in vivo evidence suggests that B crystallin modifies disease only slightly, as we were unable to detect any ch ange in the localization or the abundance of SOD1 aggregates in the absence of B crystallin. This study demonstrates that B crystallin is not a critical factor in disease. However, it does not ru le out the possibility that overexpression of B crystallin in motor neurons in SOD1 transgenic mice may have some suppressive effect on aggreg ate and disease course. Insights into Therapeutics Currently, therapies are largely ineffective in treating or slowing disease in ALS patients. This study provides insight into the disease c ourse and possible new avenues to approach

PAGE 130

130 treatment. Our findings suggest that SOD1 aggr egation is associated with onset of disease (paralysis). Our findings that SOD1 aggregation occurs late in the disease course and that pathologic events occur prior to SOD1 aggregation, suggest that therapeutic approaches targeted to disruption of detergent-insoluble SOD1 aggregates will likely have effects on the rate of disease progression. Because aggregates are detected so la te in disease, it is possible that toxic soluble forms of the mutant protein cause early pathologic abnormalities and aggregates form as a consequence of oligomer-induced toxi city. It is possible that thes e forms that must be targeted early in the disease prior to extensive muscle denervation and motor neuron loss. The identification of two regions of the human mutant SOD1 protein that enhance aggregation in cell culture may provide new aven ues for drug targets. If small compounds are identified that can stabili ze the interactions between -strands 6 and 7, perhaps the protein can stably fold or the rate of aggregation can be sl owed. This avenue is a ttractive because it is not specific to individual SOD1 mutants. This appr oach is dependent on so me form of misfolded SOD1 protein imparting toxicity. Because our results suggest that SOD1 aggregation occurs at the onset of paralysis and rapidl y increases at disease endstage, drugs that stabilize the mutant SOD1 protein and slow the rate of aggregat ion may slow the progr ession of paralysis. Additionally, our findings that B crystallin has only slight e ffects on disease course when absent in SOD1 transgenic mice taken with previous studies that demonstrate negligible effects on disease course when Hsp70, Hsp25, or Hsf1 ar e overexpressed (199-201) provides evidence that modulating a single heat s hock protein or a single pathway is insufficient to alter disease course. Thus, SOD1 aggregation and ALS pathogenesis is likely a breakdown of multiple systems and will require therapies that have many targets.

PAGE 131

131 Future Directions Protein m isfolding and aggregation are hallmark pathologic features in many neurodegenerative diseases. Conclusive eviden ce is lacking regarding which forms of the misfolded proteins are toxic and the mechanism of this toxicity. In ALS, SOD1 aggregation is still poorly understood, and new t echniques must be pursued to identify alternative forms of misfolded SOD1. Attempts have been made to isolate SOD1 oligomers by exploiting hydrophobic residues that are enhanced in the misfolded protein (168). These methods, however, can only identify a subset of mutant s, and because we are searching for a common mechanism of toxicity, new techniques for isol ating oligomeric forms of the protein will be important. To better study SOD1 aggregation and SOD1 toxicity, we are looking into new model systems to screen SOD1 mutants. While SOD1 transgenic mice are excellent models of FALS disease course, developing and characterizing th ese models is time consuming and costly. Because all mutant SOD1 transgenic mice devel op motor neuron loss, SOD1 aggregates in the brainstem and spinal cord, and hi ndlimb paralysis, developing a new mutant SOD1 transgenic mouse requires a mutant or paradigm that could provide new insights into th e disease. Using the most interesting cysteine mutati ons, we are currently establishing C. elegans that express SOD1 mutants under the contro l of muscle and neuronal promoters in collaboration with the laboratory of Dr. Richard Morimoto (202). Using this wo rm model, we will screen these mutants for effects on SOD1 aggregation and toxicity (measured by change in motility). This model is appealing because it allows for more rapid screening of a large variety of mutants. However, there are also several caveats: the model utilizes YFP fused to SOD1, which likely alters folding patterns of the protein and the mutant SOD1 worms that have been characterized develop

PAGE 132

132 aggregates with low toxicity ( 202). Nevertheless, new model sy stems may provide insight into fundamental questions of S OD1 misfolding and toxicity. Until a SOD1 mutant is identified that causes the disease without the formation of SOD1 aggregates in the affected tissues, we cannot be sure that aggregates are not participating in important toxic events in the ALS disease cour se. To further examine the contribution of aggregates to disease, we can create transg enic SOD1 mice that conditionally overexpress mutant SOD1 protein. Using a tTA system, expression of the mutant SO D1 transgene can be turned off using doxycycline at time points thro ughout the disease course. This system will require high expression levels of the mutant protein (as dose eff ects determine the acquisition of disease phentoype in this model (133, 149)) and extensive char acterization of the disease course when the transgene remains active is critical to ensure that any changes in disease progression are real effects. This conditional model could ad dress several unanswered questions in the field: 1) at what rate do aggregate form in vivo ; 2) does the rate of aggreg ation depend on the stage of disease; 3) do aggregates precursors contribu te to toxicity; and 4) are SOD1 aggregates responsible for disease phenotype. Conclusions These studies provide new in sights into how SOD1 aggregates for m and how these structures evolve in ALS dis ease course. While much remains unknown as to which structures are toxic, our findings suggest that SOD1 a ggregates are composed of globally misfolded, immature protein. Based on these studies, it appear s that SOD1 aggregation is associated with a single phase of the disease: disease duration after onset. Together, these findings will allow for new avenues in therapeutic design and research into disease mechanisms.

PAGE 133

133 LIST OF REFERENCES 1. Mulder DW, Kurland LT, Offord KP, et al. (1986) Familial adult motor neuron disease: amyotrophic lateral sclerosis. Neurology 36:511-517. 2. Bendotti C & Carri MT (2004) Lessons from models of SOD1-linked familial ALS. Trends Mol Med 10:393-400. 3. Bensimon G, Lacomblez L, & Meininger V ( 1994) A controlled trial of riluzole in amyotrophic lateral sclerosis. ALS/Riluzole Study Group. N Engl J Med 330:585-591. 4. Rosen DR, Siddique T, Patterson D et al. (1993) Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic la teral sclerosis. Nature 362:5962. 5. Hand CK, Khoris J, Salachas F et al. (2002) A novel locus for familial amyotrophic lateral sclerosis, on chromosome 18q. Am J Hum Genet 70:251-256. 6. Vance C, Rogelj B, Hortobagyi T et al. (2009) Mutations in FU S, an RNA processing protein, cause familial amyotrophic lateral sclerosis type 6. Science 323:1208-1211. 7. Abalkhail H, Mitchell J, Habgood J et al. (2003) A new familial amyotrophic lateral sclerosis locus on chromosome 16q12.1-16q12.2. Am J Hum Genet 73:383-389. 8. Sapp PC, Hosler BA, McKenna-Yasek D et al. (2003) Identification of two novel loci for dominantly inherited familial amyotrophic lateral sclerosis. Am J Hum Genet 73:397403. 9. Sreedharan J, Blair IP, Tripathi VB et al. (2008) TDP-43 mutations in familial and sporadic amyotrophic lateral sclerosis. Science 319:1668-1672. 10. Hosler BA, Siddique T, Sapp PC et al. (2000) Linkage of familial amyotrophic lateral sclerosis with frontotemporal de mentia to chromosome 9q21-q22. Jama 284:1664-1669. 11. Hutton M, Lendon CL, Rizzu P et al. (1998) Association of missense and 5'-splice-site mutations in tau with the inherited dementia FTDP-17. Nature 393:702-705. 12. Hirokawa N (1994) Microtubule organizat ion and dynamics dependent on microtubuleassociated proteins. Curr Opin Cell Biol 6:74-81. 13. Puls I, Jonnakuty C, LaMonte BH et al. (2003) Mutant dynactin in motor neuron disease. Nat Genet 33:455-456. 14. Waterman-Storer CM, Karki S, & Holzbaur EL (1995) The p150Glued component of the dynactin complex binds to both microtubules and the actin-related protein centractin (Arp-1). Proc Natl Acad Sci U S A 92:1634-1638.

PAGE 134

134 15. Nishimura AL, Mitne-Neto M, Silva HC et al. (2004) A mutation in the vesicletrafficking protein VAPB causes late-onset spinal musc ular atrophy and amyotrophic lateral sclerosis. Am J Hum Genet 75:822-831. 16. Nishimura AL, Mitne-Neto M, Silva HC et al. (2004) A novel locus for late onset amyotrophic lateral sclerosis/moto r neurone disease variant at 20q13. J Med Genet 41:315-320. 17. Foster LJ, Weir ML, Lim DY, et al. (2000) A functional role for VAP-33 in insulinstimulated GLUT4 traffic. Traffic 1:512-521. 18. Chance PF, Rabin BA, Ryan SG et al. (1998) Linkage of the gene for an autosomal dominant form of juvenile amyotrophic lateral sclerosis to chromosome 9q34. Am J Hum Genet 62:633-640. 19. Chen YZ, Bennett CL, Huynh HM et al. (2004) DNA/RNA helicase gene mutations in a form of juvenile amyotrophic lateral sclerosis (ALS4). Am J Hum Genet 74:1128-1135. 20. Kunita R, Otomo A, Mizumura H et al. (2004) Homo-oligomeri zation of ALS2 through its unique carboxyl-terminal regions is esse ntial for the ALS2-associated Rab5 guanine nucleotide exchange activity and its re gulatory function on endosome trafficking. J Biol Chem 279:38626-38635. 21. Otomo A, Hadano S, Okada T et al. (2003) ALS2, a novel guanine nucleotide exchange factor for the small GTPase Rab5, is implicated in endosomal dynamics. Hum Mol Genet 12:1671-1687. 22. Hentati A, Ouahchi K, Pericak-Vance MA et al. (1998) Linkage of a commoner form of recessive amyotrophic lateral sclerosi s to chromosome 15q15-q22 markers. Neurogenetics 2:55-60. 23. Wang J, Xu G, & Borchelt DR (2006) Ma pping superoxide dismut ase 1 domains of nonnative interaction: roles of intraand in termolecular disulfide bonding in aggregation. J Neurochem 96:1277-1288. 24. Borchelt DR, Lee MK, Slunt HS et al. (1994) Superoxide dismutase 1 with mutations linked to familial amyotrophic lateral sc lerosis possesses significant activity. Proc Natl Acad Sci U S A 91:8292-8296. 25. Potter SZ & Valentine JS (2003) The pe rplexing role of copper-zinc superoxide dismutase in amyotrophic lateral sc lerosis (Lou Gehrig's disease). J Biol Inorg Chem 8:373-380. 26. Valentine JS & Hart PJ (2003) Misfolded CuZnSOD and amyotrophic lateral sc lerosis. Proc Natl Acad Sci U S A 100:3617-3622.

PAGE 135

135 27. Hayward LJ, Rodriguez JA, Kim JW et al. (2002) Decreased meta llation and activity in subsets of mutant superoxide dismutases associated with fami lial amyotrophic lateral sclerosis. J Biol Chem 277:15923-15931. 28. Nishida CR, Gralla EB, & Valentine JS (1994) Characterization of three yeast copperzinc superoxide dismutase mutants analogous to those coded for in familial amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A 91:9906-9910. 29. Ratovitski T, Corson LB, Strain J et al. (1999) Variation in th e biochemical/biophysical properties of mutant superoxide dismutase 1 enzymes and the rate of disease progression in familial amyotrophic lateral sclerosis kindreds. Hum Mol Genet 8:1451-1460. 30. Wiedau-Pazos M, Goto JJ, Rabizadeh S et al. (1996) Altered reactivity of superoxide dismutase in familial amyotrophic lateral sclerosis. Science 271:515-518. 31. Tiwari A & Hayward LJ (2003) Familial amyotrophic lateral sclerosis mutants of copper/zinc superoxide dismutase ar e susceptible to disulfide reduction. J Biol Chem 278:5984-5992. 32. Reaume AG, Elliott JL, Hoffman EK et al. (1996) Motor neurons in Cu/Zn superoxide dismutase-deficient mice develop normally but exhibit enhanced cel l death after axonal injury. Nat Genet 13:43-47. 33. Beckman G, Lundgren E, & Tarnvik A (1973) Superoxide dismutase isozymes in different human tissues, their genetic control and intracellular localization. Hum Hered 23:338-345. 34. Crapo JD, Oury T, Rabouille C et al. (1992) Copper,zinc superoxide dismutase is primarily a cytosolic protein in human cells. Proc Natl Acad Sci U S A 89:10405-10409. 35. Keller GA, Warner TG, Steimer KS et al. (1991) Cu,Zn superoxide dismutase is a peroxisomal enzyme in human fi broblasts and hepatoma cells. Proc Natl Acad Sci U S A 88:7381-7385. 36. Jaarsma D, Rognoni F, van Duijn W et al. (2001) CuZn superoxide dismutase (SOD1) accumulates in vacuolated mitochondria in transgenic mice expressing amyotrophic lateral sclerosis-linked SOD1 mutations. Acta Neuropathol (Berl) 102:293-305. 37. Potter SZ, Zhu H, Shaw BF et al. (2007) Binding of a single zinc ion to one subunit of copper-zinc superoxide dismut ase apoprotein substantially influences the structure and stability of the entire homodimeric protein. J Am Chem Soc 129:4575-4583. 38. Parge HE, Hallewell RA, & Tainer JA (1992) Atomic structures of wild-type and thermostable mutant recombinant hu man Cu,Zn superoxide dismutase. Proc Natl Acad Sci U S A 89:6109-6113.

PAGE 136

136 39. Lyons TJ, Gralla EB, & Valentine JS ( 1999) Biological chemis try of copper-zinc superoxide dismutase and its link to amyotrophic lateral sclerosis. Met Ions Biol Syst 36:125-177. 40. Nakanishi T, Kishikawa M, Miyazaki A et al. (1998) Simple and defined method to detect the SOD-1 mutants from patients with familial amyotrophic lateral sclerosis by mass spectrometry. J Neurosci Methods 81:41-44. 41. Nakano R, Sato S, Inuzuka T et al. (1994) A novel mutation in Cu/Zn superoxide dismutase gene in Japanese familial amyotrophic lateral sclerosis. Biochem Biophys Res Commun 200:695-703. 42. Rosen DR, Bowling AC, Patterson D et al. (1994) A frequent al a 4 to val superoxide dismutase-1 mutation is associated with a rapidly progressive familial amyotrophic lateral sclerosis. Hum Mol Genet 3:981-987. 43. Wang J, Slunt H, Gonzales V et al. (2003) Copper-binding-site-null SOD1 causes ALS in transgenic mice: aggregates of nonnative SOD1 delineate a common feature. Hum Mol Genet 12:2753-2764. 44. Rodriguez JA, Valentine JS, Eggers DK et al. (2002) Familial amyotrophic lateral sclerosis-associated mutations decrease the thermal stability of distinctly metallated species of human copper/zinc superoxide dismutase. J Biol Chem 277:15932-15937. 45. Morita M, Aoki M, Abe K et al. (1996) A novel two-base mutation in the Cu/Zn superoxide dismutase gene associated with familial amyotrophic lateral sclerosis in Japan. Neurosci Lett 205:79-82. 46. Karch CM & Borchelt DR (2008) A limited role for disulfide cross-linking in the aggregation of mutant SOD1 linked to familial amyotrophic lateral sclerosis. J Biol Chem 283:13528-13537. 47. Kohno S, Takahashi Y, Miyajima H et al. (1999) A novel mutati on (Cys6Gly) in the Cu/Zn superoxide dismutase gene associated with rapidly progressive familial amyotrophic lateral sclerosis. Neurosci Lett 276:135-137. 48. Hirano M, Fujii J, Nagai Y et al. (1994) A new variant Cu /Zn superoxide dismutase (Val7-->Glu) deduced from ly mphocyte mRNA sequences from Japanese patients with familial amyotrophic la teral sclerosis. Biochem Biophys Res Commun 204:572-577. 49. Valentine JS, Doucette PA, & Zittin Potter S (2005) Copper-zinc superoxide dismutase and amyotrophic lateral sclerosis. Annu Rev Biochem 74:563-593. 50. Siddique T & Deng HX (1996) Genetics of amyotrophic late ral sclerosis. Hum Mol Genet 5 Spec No:1465-1470.

PAGE 137

137 51. Andersen PM, Sims KB, Xin WW et al. (2003) Sixteen novel mu tations in the Cu/Zn superoxide dismutase gene in amyotrophic la teral sclerosis: a decade of discoveries, defects and disputes. Amyotroph Lateral Scler Other Motor Neuron Disord 4:62-73. 52. Kim NH, Kim HJ, Kim M, et al. (2003) A novel SOD1 gene mu tation in a Korean family with amyotrophic lateral sclerosis. J Neurol Sci 206:65-69. 53. Penco S, Schenone A, Bordo D et al. (1999) A SOD1 gene mutation in a patient with slowly progressing familial ALS. Neurology 53:404-406. 54. Andersen PM, Nilsson P, Keranen ML et al. (1997) Phenotypic heterogeneity in motor neuron disease patients with CuZn-superoxi de dismutase mutations in Scandinavia. Brain 120 ( Pt 10):1723-1737. 55. Deng HX, Tainer JA, Mitsumoto H et al. (1995) Two novel SOD1 mutations in patients with familial amyotrophic lateral sclerosis. Hum Mol Genet 4:1113-1116. 56. Kawamata J, Shimohama S, Takano S et al. (1997) Novel G16S (GGC-AGC) mutation in the SOD-1 gene in a patient with appare ntly sporadic young-ons et amyotrophic lateral sclerosis. Hum Mutat 9:356-358. 57. Jones CT, Swingler RJ, & Brock DJ (1994) Identification of a novel SOD1 mutation in an apparently sporadic amyotrophic lateral sclerosis patient a nd the detection of Ile113Thr in three others. Hum Mol Genet 3:649-650. 58. Shi SG, Li LS, Chen KN et al. (2004) [Identificati on of the mutation of SOD1 gene in a familial amyotrophic la teral sclerosis]. Zhonghua Yi Xue Yi Chuan Xue Za Zhi 21:149152. 59. Boukaftane Y, Khoris J, Moulard B et al. (1998) Identification of six novel SOD1 gene mutations in familial amyot rophic lateral sclerosis. Can J Neurol Sci 25:192-196. 60. Gellera C, Castellotti B, Riggio MC et al. (2001) Superoxide dism utase gene mutations in Italian patients with familial and sporadic amyotrophic late ral sclerosis: identification of three novel missense mutations. Neuromuscul Disord 11:404-410. 61. Aoki M, Ogasawara M, Matsubara Y et al. (1993) Mild ALS in Japan associated with novel SOD mutation. Nat Genet 5:323-324. 62. Enayat ZE, Orrell RW, Claus A et al. (1995) Two novel mutations in the gene for copper zinc superoxide dismutase in UK famili es with amyotrophic lateral sclerosis. Hum Mol Genet 4:1239-1240. 63. Garcia-Redondo A, Bustos F, Juan YSB et al. (2002) Molecular analysis of the superoxide dismutase 1 gene in Spanish pati ents with sporadic or familial amyotrophic lateral sclerosis. Mus cle Nerve 26:274-278.

PAGE 138

138 64. Shaw CE, Enayat ZE, Chioza BA et al. (1998) Mutations in all five exons of SOD-1 may cause ALS. Ann Neurol 43:390-394. 65. Stewart HG, Mackenzie IR, Eisen A et al. (2006) Clinicopathologi cal phenotype of ALS with a novel G72C SOD1 gene mutation mimicking a myopathy. Muscle Nerve 33:701706. 66. Segovia-Silvestre T, Andreu AL, Vives-Bauza C et al. (2002) A novel exon 3 mutation (D76V) in the SOD1 gene associated with slowly progressive ALS. Amyotroph Lateral Scler Other Motor Neuron Disord 3:69-74. 67. Alexander MD, Traynor BJ, Miller N, et al. (2002) "True" sporadic ALS associated with a novel SOD-1 mutation. Ann Neurol 52:680-683. 68. Andersen PM (2006) Amyotrophic lateral scle rosis associated with mutations in the CuZn superoxide dismutase gene. Curr Neurol Neurosci Rep 6:37-46. 69. Beck M, Sendtner M, & Toyka KV (2007) N ovel SOD1 N86K mutation is associated with a severe phenotype in familial ALS. Muscle Nerve 36:111-114. 70. Hayward C, Brock DJ, Minns RA et al. (1998) Homozygosity for Asn86Ser mutation in the CuZn-superoxide dismutase gene produces a severe clinical phe notype in a juvenile onset case of familial amyotrophic lateral sclerosis. J Med Genet 35:174. 71. Rezania K, Yan J, Dellefave L et al. (2003) A rare Cu/Zn superoxide dismutase mutation causing familial amyotrophic lateral sclerosis with variable age of onset, incomplete penetrance and a sensory neuropathy. Amyotroph Lateral Scler Other Motor Neuron Disord 4:162-166. 72. Andersen PM, Nilsson P, Ala-Hurula V, et al. (1995) Amyotrophic lateral sclerosis associated with homozygosity for an Asp90Ala mutation in CuZn-superoxide dismutase. Nat Genet 10:61-66. 73. Chou CM, Huang CJ, Shih CM et al. (2005) Identification of three mutations in the Cu,Zn-superoxide dismutase (Cu,Zn-SOD) gene with familial amyotrophic lateral sclerosis: transduction of human Cu,ZnSOD into PC12 cells by HIV-1 TAT protein basic domain. Ann N Y Acad Sci 1042:303-313. 74. Esteban J, Rosen DR, Bowling AC et al. (1994) Identification of two novel mutations and a new polymorphism in the gene for Cu/Z n superoxide dismutase in patients with amyotrophic lateral sclerosis. Hum Mol Genet 3:997-998. 75. Elshafey A, Lanyon WG, & Connor JM (1994) Identification of a new missense point mutation in exon 4 of the Cu/Zn superoxide dismutase (SOD-1) gene in a family with amyotrophic lateral sclerosis. Hum Mol Genet 3:363-364.

PAGE 139

139 76. Hosler BA, Nicholson GA, Sapp PC, et al. (1996) Three novel mutations and two variants in the gene for Cu/Zn superoxide dismutase in familial amyotrophic lateral sclerosis. Neuromuscul Disord 6:361-366. 77. Hand CK, Mayeux-Portas V, Khoris J, et al. (2001) Compound he terozygous D90A and D96N SOD1 mutations in a recessive amyotrophic lateral sclerosis family. Ann Neurol 49:267-271. 78. Yulug IG, Katsanis N, de Belleroche J et al. (1995) An improved protocol for the analysis of SOD1 gene mutati ons, and a new mutation in exon 4. Hum Mol Genet 4:1101-1104. 79. Sato T, Yamamoto Y, Nakanishi T et al. (2004) Identification of two novel mutations in the Cu/Zn superoxide dismutase gene with familial amyotrophic lateral sclerosis: mass spectrometric and genomic analyses. J Neurol Sci 218:79-83. 80. Jones CT, Shaw PJ, Chari G et al. (1994) Identification of a novel exon 4 SOD1 mutation in a sporadic amyotrophi c lateral sclerosis patient. Mol Cell Probes 8:329-330. 81. Tan CF, Piao YS, Hayashi S et al. (2004) Familial amyotrophic lateral sclerosis with bulbar onset and a novel Asp101Tyr Cu/Zn superoxide dismutase gene mutation. Acta Neuropathol 108:332-336. 82. Ikeda M, Abe K, Aoki M et al. (1995) Variable clinical symptoms in familial amyotrophic lateral sclerosis with a novel point mutation in the Cu/Zn superoxide dismutase gene. Neurology 45:2038-2042. 83. Orrell RW, Habgood JJ, Gardiner I et al. (1997) Clinical and func tional investigation of 10 missense mutations and a novel frameshift in sertion mutation of the gene for copperzinc superoxide dismutase in UK famili es with amyotrophic lateral sclerosis. Neurology 48:746-751. 84. Shibata N, Kawaguchi M, Uchida K et al. (2007) Protein-bound crotonaldehyde accumulates in the spinal cord of superoxide dismutase-1 mutation-associated familial amyotrophic lateral sclerosis and its transgenic mouse model. Neuropathology 27:49-61. 85. Kostrzewa M, Burck-Lehmann U, & Muller U (1994) Autosomal dominant amyotrophic lateral sclerosis: a novel mutation in the Cu/Zn superoxide dismutase-1 gene. Hum Mol Genet 3:2261-2262. 86. Jackson M, Al-Chalabi A, Enayat ZE et al. (1997) Copper/zinc superoxide dismutase 1 and sporadic amyotrophic lateral sclerosis: an alysis of 155 cases a nd identification of a novel insertion mutation. Ann Neurol 42:803-807. 87. Pramatarova A, Goto J, Nanba E et al. (1994) A two basepair de letion in the SOD 1 gene causes familial amyotrophic lateral sclerosis. Hum Mol Genet 3:2061-2062.

PAGE 140

140 88. Takehisa Y, Ujike H, Ishizu H, et al. (2001) Familial amyotrophic lateral sclerosis with a novel Leu126Ser mutation in the copper/zinc superoxide dismutase gene showing mild clinical features and lewy body-like hyaline inclusions. Arch Neurol 58:736-740. 89. Pramatarova A, Figlewicz DA, Krizus A et al. (1995) Identification of new mutations in the Cu/Zn superoxide dismutase gene of patients with familial amyotrophic lateral sclerosis. Am J Hum Genet 56:592-596. 90. Watanabe M, Aoki M, Abe K et al. (1997) A novel missense point mutation (S134N) of the Cu/Zn superoxide dismutase gene in a patient with familial motor neuron disease. Hum Mutat 9:69-71. 91. Nogales-Gadea G, Garcia-Arumi E, Andreu AL et al. (2004) A novel exon 5 mutation (N139H) in the SOD1 gene in a Spanish family associated with incomplete penetrance. J Neurol Sci 219:1-6. 92. Naini A, Musumeci O, Hayes L et al. (2002) Identification of a novel mutation in Cu/Zn superoxide dismutase gene associated w ith familial amyotrophic lateral sclerosis. J Neurol Sci 198:17-19. 93. Deng HX, Hentati A, Tainer JA et al. (1993) Amyotrophic lateral sclerosis and structural defects in Cu,Zn superoxide dismutase. Science 261:1047-1051. 94. Sapp PC, Rosen DR, Hosler BA, et al. (1995) Identification of three novel mutations in the gene for Cu/Zn superoxide dismutase in patients with familial amyotrophic lateral sclerosis. Neuromuscul Disord 5:353-357. 95. Kostrzewa M, Damian MS, & Muller U (1996) Superoxide dismutase 1: identification of a novel mutation in a case of fam ilial amyotrophic lateral sclerosis. Hum Genet 98:48-50. 96. Valentine JS (2002) Do oxidatively modified proteins cause ALS? Free Radic Biol Med 33:1314-1320. 97. Yim HS, Kang JH, Chock PB et al. (1997) A familial amyotrophic lateral sclerosisassociated A4V Cu, Zn-superoxide dismut ase mutant has a lower Km for hydrogen peroxide. Correlation between clinical severity and the Km value. J Biol Chem 272:88618863. 98. Yim MB, Chock PB, & Stadtman ER (1990) Copper, zinc superoxide dismutase catalyzes hydroxyl radical production from hydrogen peroxide. Proc Natl Acad Sci U S A 87:5006-5010. 99. Beckman JS, Carson M, Smith CD et al. (1993) ALS, SOD and peroxynitrite. Nature 364:584. 100. Crow JP, Ye YZ, Strong M et al. (1997) Superoxide dismut ase catalyzes nitration of tyrosines by peroxynitrite in the rod a nd head domains of neurofilament-L. J Neurochem 69:1945-1953.

PAGE 141

141 101. Estevez AG, Crow JP, Sampson JB et al. (1999) Induction of n itric oxide-dependent apoptosis in motor neurons by zinc-deficient superoxide dismutase. Science 286:24982500. 102. Goto JJ, Zhu H, Sanchez RJ et al. (2000) Loss of in vitro meta l ion binding specificity in mutant copper-zinc superoxide dismutases associated with fam ilial amyotrophic lateral sclerosis. J Biol Chem 275:1007-1014. 103. Strange RW, Antonyuk S, Hough MA et al. (2003) The structure of holo and metaldeficient wild-type human Cu, Zn superoxide dismutase and its relevance to familial amyotrophic lateral sclerosis. J Mol Biol 328:877-891. 104. Wang J, Xu G, Gonzales V et al. (2002) Fibrillar incl usions and motor neuron degeneration in transgenic mice expressing superoxide dismutase 1 with a disrupted copper-binding site. Neurobiol Dis 10:128-138. 105. Cleveland DW & Rothstein JD (2001) From Charcot to Lou Gehrig: deciphering selective motor neuron death in ALS. Nat Rev Neurosci 2:806-819. 106. Khare SD, Caplow M, & Dokholyan NV (2004) The rate and equilibrium constants for a multistep reaction sequence for the aggregation of superoxide dismutase in amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A 101:15094-15099. 107. Khare SD, Wilcox KC, Gong P et al. (2005) Sequence and structural determinants of Cu, Zn superoxide dismutase aggregation. Proteins 61:617-632. 108. DiDonato M, Craig L, Huff ME et al. (2003) ALS mutants of human superoxide dismutase form fibrous aggregat es via framework destabilization. J Mol Biol 332:601615. 109. Stathopulos PB, Rumfel dt JA, Karbassi F et al. (2006) Calorimetric analysis of thermodynamic stability and aggregation for apo and holo amyotrophi c lateral sclerosisassociated Gly-93 mutants of superoxide dismutase. J Biol Chem 281:6184-6193. 110. Leinweber B, Barofsky E, Barofsky DF, et al. (2004) Aggregation of ALS mutant superoxide dismutase expressed in Escherichia coli. Free Radic Biol Med 36:911-918. 111. Durham HD, Roy J, Dong L et al. (1997) Aggregation of mu tant Cu/Zn superoxide dismutase proteins in a culture model of ALS. J Neuropathol Exp Neurol 56:523-530. 112. Matsumoto S, Kusaka H, Ito H et al. (1996) Golgi apparatus and intraneuronal inclusions of anterior horn cells in amyotrophic lateral sclerosis: an immunohistochemical study. Acta Neuropathol (Berl) 91:603-607. 113. Bruijn LI, Becher MW, Lee MK et al. (1997) ALS-linked SOD1 mutant G85R mediates damage to astrocytes and promotes rapidl y progressive disease with SOD1-containing inclusions. Neuron 18:327-338.

PAGE 142

142 114. Wang J, Xu G, & Borchelt DR (2002) Hi gh molecular weight complexes of mutant superoxide dismutase 1: age-depend ent and tissue-specific accumulation. Neurobiol Dis 9:139-148. 115. Wang J, Xu G, Li H et al. (2005) Somatodendritic accumulation of misfolded SOD1L126Z in motor neurons mediates dege neration: alphaB-crystallin modulates aggregation. Hum Mol Genet 14:2335-2347. 116. Okado-Matsumoto A & Fridovich I (2002) Amyotrophic lateral sclerosis: a proposed mechanism. Proc Natl Acad Sci U S A 99:9010-9014. 117. Johnston JA, Dalton MJ, Gurney ME, et al. (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 97:12571-12576. 118. Boillee S, Vande Velde C, & Cleveland DW (2 006) ALS: a disease of motor neurons and their nonneuronal neighbors. Neuron 52:39-59. 119. Gurney ME, Pu H, Chiu AY et al. (1994) Motor neuron degeneration in mice that express a human Cu,Zn supe roxide dismutase mutation. Science 264:1772-1775. 120. Wong PC & Borchelt DR (1995) Motor neuron disease caused by mutations in superoxide dismutase 1. Curr Opin Neurol 8:294-301. 121. Ripps ME, Huntley GW, Hof PR, et al. (1995) Transgenic mice expressing an altered murine superoxide dismutas e gene provide an animal model of amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A 92:689-693. 122. Jonsson PA, Graffmo KS, Brannstrom T, et al. (2006) Motor neuron disease in mice expressing the wild type-like D90A mutant superoxide dismutase-1. J Neuropathol Exp Neurol 65:1126-1136. 123. Jonsson PA, Ernhill K, Andersen PM et al. (2004) Minute quant ities of misfolded mutant superoxide dismutase-1 cau se amyotrophic lateral sclerosis. Brain 127:73-88. 124. Gurney ME (1997) Transgenic animal models of familial amyotrophic lateral sclerosis. J Neurol 244 Suppl 2:S15-20. 125. Watanabe M, Dykes-Hoberg M, Culotta VC et al. (2001) Histological evidence of protein aggregation in mutant SOD1 transgen ic mice and in amyotrophic lateral sclerosis neural tissues. Neurobiol Dis 8:933-941. 126. Bruijn LI, Houseweart MK, Kato S et al. (1998) Aggregation and motor neuron toxicity of an ALS-linked SOD1 mutant i ndependent from wild-type SOD1. Science 281:18511854.

PAGE 143

143 127. Cheroni C, Peviani M, Cascio P et al. (2005) Accumulation of human SOD1 and ubiquitinated deposits in the spinal cord of SOD1G93A mice during motor neuron disease progression correlates w ith a decrease of proteasome. Neurobiol Dis 18:509-522. 128. Basso M, Massignan T, Samengo G et al. (2006) Insoluble mutant SOD1 is partly oligoubiquitinated in amyotr ophic lateral sclerosis mice. J Biol Chem 281:33325-33335. 129. Seilhean D, Takahashi J, El Hachimi KH et al. (2004) Amyotrophic late ral sclerosis with neuronal intranuclear protein inclusions. Acta Neuropathol (Berl) 108:81-87. 130. Stieber A, Gonatas JO, & Gonatas NK (2000) Aggregates of muta nt protein appear progressively in dendrites, in periaxonal pr ocesses of oligodendrocytes, and in neuronal and astrocytic perikarya of mice expre ssing the SOD1(G93A) mutation of familial amyotrophic lateral sclerosis. J Neurol Sci 177:114-123. 131. Shinder GA, Lacourse MC, Minotti S et al. (2001) Mutant Cu/Zn-superoxide dismutase proteins have altered solubility and interact with heat shock/stress pr oteins in models of amyotrophic lateral sclerosis. J Biol Chem 276:12791-12796. 132. Watanabe Y, Yasui K, Nakano T et al. (2005) Mouse motor neuron disease caused by truncated SOD1 with or without C-terminal modification. Brain Res Mol Brain Res 135:12-20. 133. Jaarsma D, Teuling E, Haasdijk ED et al. (2008) Neuron-specific expression of mutant superoxide dismutase is sufficient to induce amyotrophic lateral scle rosis in transgenic mice. J Neurosci 28:2075-2088. 134. Deng HX, Shi Y, Furukawa Y et al. (2006) Conversion to the amyotrophic lateral sclerosis phenotype is associated with in termolecular linked inso luble aggregates of SOD1 in mitochondria. Proc Natl Acad Sci U S A 103:7142-7147. 135. Boillee S, Yamanaka K, Lobsiger CS et al. (2006) Onset and progr ession in inherited ALS determined by motor neurons and microglia. Science 312:1389-1392. 136. Miller TM, Smith RA, & Cleveland DW (2006) Amyotrophic lateral sclerosis and gene therapy. Nat Clin Pract Neurol 2:462-463. 137. Lobsiger CS, Boillee S, & Cleveland DW (2007) Inaugural Article: Toxicity from different SOD1 mutants dysregulates th e complement system and the neuronal regenerative response in ALS motor neurons. Proc Natl Acad Sci U S A 104:7319-7326. 138. Clement AM, Nguyen MD, Roberts EA et al. (2003) Wild-type nonneuronal cells extend survival of SOD1 mutant motor neurons in ALS mice. Science 302:113-117. 139. Dobrowolny G, Aucello M, Rizzuto E et al. (2008) Skeletal muscle is a primary target of SOD1G93A-mediated toxicity. Cell Metab 8:425-436.

PAGE 144

144 140. Prudencio M, Durazo A, Whitelegge JP et al. (2009) Modulation of mutant superoxide dismutase 1 aggregation by co-e xpression of wild-type enzyme. J Neurochem 108:10091018. 141. Furukawa Y, Fu R, Deng HX et al. (2006) Disulfide cross-li nked protein represents a significant fraction of ALS-associated Cu, Zn-s uperoxide dismutase aggregates in spinal cords of model mice. Proc Natl Acad Sci U S A 103:7148-7153. 142. Jonsson PA, Graffmo KS, Andersen PM et al. (2006) Disulphide-reduced superoxide dismutase-1 in CNS of transgenic amyotrophic lateral sclerosis models. Brain 129:451464. 143. Banci L, Bertini I, Durazo A et al. (2007) Metal-free superoxide dismutase forms soluble oligomers under physiological conditions: A possible general mechanism for familial ALS. Proc Natl Acad Sci U S A 104:11263-11267. 144. Cozzolino M, Amori I, Pesaresi MG et al. (2008) Cysteine 111 Affects Aggregation and Cytotoxicity of Mutant Cu,Zn-superoxi de Dismutase Associated with Familial Amyotrophic Lateral Sclerosis. J Biol Chem 283:866-874. 145. Niwa J, Yamada S, Ishigaki S et al. (2007) Disulfide bond mediates aggregation, toxicity, and ubiquitylation of familial amyotr ophic lateral sclerosislinked mutant SOD1. J Biol Chem 282:28087-28095. 146. Mizushima S & Nagata S (1990) pEF-BOS, a powerful mammalia n expression vector. Nucleic Acids Res 18:5322. 147. Wong PC, Pardo CA, Borchelt DR et al. (1995) An adverse property of a familial ALSlinked SOD1 mutation causes motor neur on disease characterized by vacuolar degeneration of mitochondria. Neuron 14:1105-1116. 148. Laemmli UK (1970) Cleavage of structural pr oteins during the assembly of the head of bacteriophage T4. Nature 227:680-685. 149. Wang J, Xu G, Slunt HH, et al. (2005) Coincident threshol ds of mutant protein for paralytic disease and protein aggregation cau sed by restrictively expressed superoxide dismutase cDNA. Neurobiol Dis 20:943-952. 150. Murphy RM (2002) Peptide aggregation in neurodegenerative disease. Annu Rev Biomed Eng 4:155-174. 151. Shaw BF, Lelie HL, Durazo A et al. (2008) Detergent insolubl e aggregates associated with amyotrophic lateral sclerosis in tran sgenic mice contain primarily full-length, unmodified superoxi de dismutase-1. J Biol Chem 152. Watanabe S, Nagano S, Duce J et al. (2007) Increased affinity for copper mediated by cysteine 111 in forms of mutant superoxi de dismutase 1 linked to amyotrophic lateral sclerosis. Free Radic Biol Med 42:1534-1542.

PAGE 145

145 153. Schinina ME, Carlini P, Polticelli F et al. (1996) Amino acid seque nce of chicken Cu, Zn-containing superoxide dismutase and identif ication of glutathionyl adducts at exposed cysteine residues. Eur J Biochem 237:433-439. 154. Ogawa Y, Kosaka H, Nakanishi T et al. (1997) Stability of mutant superoxide dismutase-1 associated with familial amyotr ophic lateral sclerosis determines the manner of copper release and induction of thioredoxin in erythrocytes. Biochem Biophys Res Commun 241:251-257. 155. Fujiwara N, Nakano M, Kato S et al. (2007) Oxidative modification to cysteine sulfonic acid of Cys111 in human copperzinc superoxide dismutase. J Biol Chem 282:3593335944. 156. Krishnan U, Son M, Rajendran B et al. (2006) Novel mutations that enhance or repress the aggregation potential of SOD1. Mol Cell Biochem 287:201-211. 157. Son M, Puttaparthi K, Kawamata H et al. (2007) Overexpression of CCS in G93ASOD1 mice leads to accelerated neurologica l deficits with severe mitochondrial pathology. Proc Natl Acad Sci U S A 104:6072-6077. 158. Stromer T & Serpell LC (2005) Structure and morphology of the Alzheimer's amyloid fibril. Microsc Res Tech 67:210-217. 159. Guo JT, Wetzel R, & Xu Y (2004) Molecular modeling of the core of Abeta amyloid fibrils. Proteins 57:357-364. 160. Morimoto A, Irie K, Murakami K et al. (2002) Aggregation and neurotoxicity of mutant amyloid beta (A beta) peptides with proline replacement: importance of turn formation at positions 22 and 23. Biochem Biophys Res Commun 295:306-311. 161. Sawaya MR, Sambashivan S, Nelson R et al. (2007) Atomic structures of amyloid crossbeta spines reveal varied steric zippers. Nature 447:453-457. 162. Thakur AK & Wetzel R (2002) Mutational analysis of the structural organization of polyglutamine aggregates. Proc Natl Acad Sci U S A 99:17014-17019. 163. Elam JS, Taylor AB, Strange R et al. (2003) Amyloid-like filaments and water-filled nanotubes formed by SOD1 mutant proteins linked to familial ALS. Nat Struct Biol 10:461-467. 164. Shaw BF, Durazo A, Nersissian AM et al. (2006) Local unfolding in a destabilized, pathogenic variant of superoxi de dismutase 1 observed with H/D exchange and mass spectrometry. J Biol Chem 281:18167-18176. 165. Khare SD, Ding F, & Dokholyan NV (2003) Fold ing of Cu, Zn superoxide dismutase and familial amyotrophic la teral sclerosis. J Mol Biol 334:515-525.

PAGE 146

146 166. Sasaki S, Nagai M, Aoki M et al. (2007) Motor neuron diseas e in transgenic mice with an H46R mutant SOD1 gene. J Neuropathol Exp Neurol 66:517-524. 167. Fujita Y & Okamoto K (2005) Golgi apparatus of the motor neurons in patients with amyotrophic lateral sclerosis and in mice m odels of amyotrophic lateral sclerosis. Neuropathology 25:388-394. 168. Zetterstrom P, Stewart HG, Bergemalm D et al. (2007) Soluble misfolded subfractions of mutant superoxide dismutase-1s are enrich ed in spinal cords throughout life in murine ALS models. Proc Natl Acad Sci U S A 104:14157-14162. 169. Furukawa Y, Kaneko K, Yamanaka K, et al. (2008) Complete loss of post-translational modifications triggers fibrillar aggregation of SOD1 in the familial form of amyotrophic lateral sclerosis. J Biol Chem 283:24167-24176. 170. Furukawa Y, Torres AS, & O'Halloran TV (2004) Oxygen-induced maturation of SOD1: a key role for disulfide formation by the copper chaperone CCS. Embo J 23:2872-2881. 171. Wang J, Caruano-Yzermans A, Rodriguez A et al. (2007) Disease-associated mutations at copper ligand histidine resi dues of superoxide dismutas e 1 diminish the binding of copper and compromise dimer stability. J Biol Chem 282:345-352. 172. Lindberg MJ, Normark J, Holmgren A et al. (2004) Folding of human superoxide dismutase: disulfide reduction prevents dimerization and produces marginally stable monomers. Proc Natl Acad Sci U S A 101:15893-15898. 173. Hart PJ (2006) Pathogenic superoxide dismutase structur e, folding, aggregation and turnover. Curr Opin Chem Biol 10:131-138. 174. Yim MB, Kang JH, Yim HS et al. (1996) A gain-of-function of an amyotrophic lateral sclerosis-associated Cu,Zn-superoxide dismut ase mutant: An enhancement of free radical formation due to a decrease in Km for hydrogen peroxide. Proc Natl Acad Sci U S A 93:5709-5714. 175. Stathopulos PB, Rumf eldt JA, Scholz GA et al. (2003) Cu/Zn superoxide dismutase mutants associated with amyotrophic latera l sclerosis show enhanced formation of aggregates in vitro. Proc Natl Acad Sci U S A 100:7021-7026. 176. Fischer LR, Culver DG, Tennant P et al. (2004) Amyotrophic lateral sclerosis is a distal axonopathy: evidence in mice and man. Exp Neurol 185:232-240. 177. Kennel PF, Finiels F, Revah F et al. (1996) Neuromuscular function impairment is not caused by motor neurone loss in FALS mice: an electromyographic study. Neuroreport 7:1427-1431. 178. Hegedus J, Putman CT, & Gordon T (2007) Ti me course of preferential motor unit loss in the SOD1 G93A mouse model of amyotrophic lateral sclerosis. Neurobiol Dis 28:154164.

PAGE 147

147 179. Borchelt DR, Wong PC, Becher MW et al. (1998) Axonal transport of mutant superoxide dismutase 1 and focal axonal abnormalities in the proximal axons of transgenic mice. Neurobiol Dis 5:27-35. 180. Turner BJ & Talbot K (2008) Transgenics, t oxicity and therapeutics in rodent models of mutant SOD1-mediated familial ALS. Prog Neurobiol 85:94-134. 181. Kasperaviciute D, Weale ME, Shianna KV et al. (2007) Large-scale pathways-based association study in amyotr ophic lateral sclerosis. Brain 182. Kabashi E & Durham HD (2006) Failure of pr otein quality control in amyotrophic lateral sclerosis. Biochim Biophys Acta 1762:1038-1050. 183. Devlin GL, Carver JA, & Bottomley SP ( 2003) The selective inhibition of serpin aggregation by the molecular chaperone, alpha-crystallin, indicates a nucleationdependent specificity. J Biol Chem 278:48644-48650. 184. Sathish HA, Stein RA, Yang G et al. (2003) Mechanism of chap erone function in small heat-shock proteins. Fluorescence studies of the conformations of T4 lysozyme bound to alphaB-crystallin. J Biol Chem 278:44214-44221. 185. Tanksale A, Ghatge M, & Deshpande V (2002) Alpha-crystallin binds to the aggregationprone molten-globule state of alkaline protease : implications for pr eventing irreversible thermal denaturation. Protein Sci 11:1720-1728. 186. Kumar MS, Kapoor M, Sinha S et al. (2005) Insights into hydrophobicity and the chaperone-like function of alphaAand alpha B-crystallins: an isothermal titration calorimetric study. J Biol Chem 280:21726-21730. 187. Horwitz J (2003) Alpha-crystallin. Exp Eye Res 76:145-153. 188. Narberhaus F (2002) Alpha-crystallin-t ype heat shock proteins: socializing minichaperones in the context of a multichaperone network. Microbiol Mol Biol Rev 66:64-93; table of contents. 189. Kumar LV & Rao CM (2000) Domain sw apping in human alpha A and alpha B crystallins affects oligomerization and enhances chaperone-like activity. J Biol Chem 275:22009-22013. 190. Dubin RA, Wawrousek EF, & Piatigorsky J ( 1989) Expression of the murine alpha Bcrystallin gene is not restricted to the lens. Mol Cell Biol 9:1083-1091. 191. Iwaki T, Kume-Iwaki A, & Goldman JE (1990) Cellular distribution of alpha B-crystallin in non-lenticular tissues. J Histochem Cytochem 38:31-39. 192. Brady JP, Garland DL, Green DE et al. (2001) AlphaB-crystallin in lens development and muscle integrity: a gene knockout approach. Invest Ophthalmol Vis Sci 42:29242934.

PAGE 148

148 193. Suzuki A, Sugiyama Y, Hayashi Y et al. (1998) MKBP, a novel member of the small heat shock protein family, binds and activ ates the myotonic dyst rophy protein kinase. J Cell Biol 140:1113-1124. 194. Shama KM, Suzuki A, Harada K et al. (1999) Transient up-re gulation of myotonic dystrophy protein kinase-binding protein, MKBP, and HSP27 in the neonatal myocardium. Cell Struct Funct 24:1-4. 195. Jakob U, Gaestel M, Engel K et al. (1993) Small heat shock proteins are molecular chaperones. J Biol Chem 268:1517-1520. 196. Wang J, Martin E, Gonzales V et al. (2007) Differential regula tion of small heat shock proteins in transgenic mouse mode ls of neurodegenerative diseases. Neurobiol Aging 197. Sun Y & MacRae TH (2005) Characterization of novel sequence motif s within Nand Cterminal extensions of p26, a small heat shock protein from Artemia franciscana. Febs J 272:5230-5243. 198. Batulan Z, Shinder GA, Minotti S et al. (2003) High threshold for induction of the stress response in motor neurons is associat ed with failure to activate HSF1. J Neurosci 23:5789-5798. 199. Liu J, Shinobu LA, Ward CM, et al. (2005) Elevation of th e Hsp70 chaperone does not effect toxicity in mouse models of familial amyotrophic lateral sclerosis. J Neurochem 93:875-882. 200. Krishnan J, Vannuvel K, Andries M et al. (2008) Over-expression of Hsp27 does not influence disease in the mutant SOD1(G 93A) mouse model of amyotrophic lateral sclerosis. J Neurochem 106:2170-2183. 201. Sharp PS, Akbar MT, Bouri S et al. (2008) Protective effects of heat shock protein 27 in a model of ALS occur in the earl y stages of disease progression. Neurobiol Dis 30:42-55. 202. Gidalevitz T, Krupinski T, Garcia S et al. (2009) Destabilizing Protein Polymorphisms in the Genetic Background Direct Phenotypi c Expression of Mutant SOD1 Toxicity. PLoS Genet 5:e1000399.

PAGE 149

149 BIOGRAPHICAL SKETCH Celeste Karch was born in South Field, Mi chigan, in 1983. Afte r graduating from Northville High School in 2001, she attended Kalam azoo College for her undergraduate studies. She graduated with a Bachelor of Arts degree in biology in 2005. She began her graduate studies in the Interdisciplinary Program for Biomedical Research at the University of Florida in 2005. She joined the laboratory of Dr. David Borchelt in 2006.