Caenorhabditis Elegans as a Model for Human Synucleopathies
Doctoral dissertation
To be presented by permission of the Faculty of Natural and Environmental Sciences of the University of Kuopio for public examination in Auditorium L21,
Snellmania building, University of Kuopio, on Saturday 14th April 2007, at 1 p.m.
Department of Neurobiology A.I. Virtanen Institute for Molecular Sciences Department of Biosciences Faculty of Natural and Environmental Sciences University of Kuopio
SUVI VARTIAINEN
JOKA KUOPIO 2007
KUOPION YLIOPISTON JULKAISUJA G. - A.I. VIRTANEN -INSTITUUTTI 52 KUOPIO UNIVERSITY PUBLICATIONS G.
A.I. VIRTANEN INSTITUTE FOR MOLECULAR SCIENCES 52
Distributor: Kuopio University Library P.O. Box 1627
FI-70211 KUOPIO FINLAND
Tel. +358 17 163 430 Fax +358 17 163 410
http://www.uku.fi/kirjasto/julkaisutoiminta/julkmyyn.html
Series Editors: Research Director Olli Gröhn, Ph.D.
Department of Neurobiology
A.I. Virtanen Institute for Molecular Sciences Research Director Michael Courtney, Ph.D.
Department of Neurobiology
A.I. Virtanen Institute for Molecular Sciences
Author’s address: Finnish Research Unit on Mitochondrial Biogenesis and Disease Institute of Medical Technology
FI-33014 University of Tampere
Supervisors: Professor Garry Wong, Ph.D.
Department of Neurobiology
A. I. Virtanen Institute for Molecular Sciences Researcher Merja Lakso, Ph.D.
Department of Neurobiology
A. I. Virtanen Institute for Molecular Sciences
Reviewers: Ana Vaz Gomez, Ph.D.
Vascaia AB
Stockholm, Sweden
Docent Carina Holmberg-Still
Molecular and Cancer Biology Program Biomedicum Helsinki, University of Helsinki
Opponent: Research Director Christian Neri, Ph.D.
Laboratory of Genomic Biology Centre Paul Broca
Inserm, France
ISBN 978-951-27-0611-2 ISBN 978-951-27-0433-0 (PDF) ISSN 1458-7335
Kopijyvä Kuopio 2007 Finland
Vartiainen, Suvi. Caenorhabditis elegans as a model for human synucleopathies.
Kuopio University Publications G. – A.I. Virtanen Institute for Molecular Sciences 52.
2007. 94 p.
ISBN 978-951-27-0611-2 ISBN 978-951-27-0433-0 (PDF) ISSN 1458-7335
ABSTRACT
Neurodegenerative diseases are devastating conditions that impair human function and are becoming far more common due to the increasing proportion of senior citizens in society. Neurodegenerative diseases include synucleopathies, which have the common feature of α-synuclein rich protein aggregates. We used the nematode Caenorhabditis elegans to model human synucleopathies. The nematode research began forty years ago with definitions of its structure and development, and recently the research has expanded to its utilization as a model organism. This has been due to mostly new molecular biology techniques that have made it possible to examine the experimental animal on a whole new level. To investigate our nematode model of human synucleopathies, we have further used some of the newly developed techniques such as cDNA microarrays and RNA interference.
The worm model of human synucleopathies was succesful in uncovering some of the pathogenic hallmarks and deficits of synucleopathy found in Parkinson’s disease. The worm showed impaired motor functions as defined by a trashing assay and loss of dopaminergic neurons. In microarray analysis of α-synuclein induced gene expression changes, proteasomal and mitochondrial genes were up-regulated, whereas some of the histone genes were down-regulated. A study on aging profiles revealed that the α-synuclein transgenic worms have extended lifespan and further analyses suggest it to originate from dietary restriction. Finally, a large scale RNAi experiment was carried out on chromosome I. From the ca. 2300 genes screened, 20 had a significant impact on the movement of the α-synuclein worm.
In conclusion, we had three major findings with our worm model. Firstly, we demonstrated that a nonendogenous α-synuclein can induce pathogenic events comparable to the human disease. Secondly, the results gave support to the current hypothesis that energy metabolism and protein degradation are involved in α-synuclein toxicity. Thirdly, new information was gained concerning histone gene involvement in the pathogenesis. These results provide new information about the effects of α- synuclein over-expression including data at the gene expression and interference levels. The established α-synuclein over-expressing C. elegans worm model has already been provided to other researchers who have used it as a valuable model in their own original studies.
National Library of Medicine Classification: QY 58, QX 203, WL 359
Medical Subject Headings: Disease Models, Animal; Animals, Genetically Modified;
Invertebrates; Nematoda; Caenorhabditis elegans; Neurodegenerative Diseases;
Parkinson Disease; alpha-Synuclein; Lewy Bodies; Gene Expression; Gene Expression Regulation; Energy Metabolism; Proteins/metabolism; Histones;
Oligonucleotide Array Sequence Analysis; RNA Interference
To Ilkka, Lumi and the dogs
And to my mom and dad
ACKNOWLEDGEMENTS
These studies were carried out in the A. I. Virtanen Institute for Molecular Sciences, University of Kuopio, during the years 2002-2006.
I am most grateful to my supervisors Professor Garry Wong, Ph.D and Merja Lakso Ph.D. for showing me the world of functional genomics and giving me the opportunity to work with C. elegans. Garry’s enthusiasm and open mindness towards science has been an inspiration and Merja’s hard working has set an example how the research should be done. I also thank Garry for revising the language of this thesis.
Special thanks goes to the official reviewers of this thesis Ana Vaz Gomez Ph.D, and Docent Carina Holmberg-Still Ph.D. Their invaluable comments and suggestions improved the quality of this thesis remarkably.
The co-authors of the publications Garry Wong Ph.D, Merja Lakso Ph.D., Jouni Sirviö Ph.D., Anu-Maarit Moilanen Ph.D., James Thomas Ph.D., Richard Nass Ph.D., Randy Blakely Ph.D., Petri Pehkonen M. Sc. and Vuokko Aarnio M.Sc.
I wish to thank all lab members I had the possibility to work with. I have been very fortunate to have worked with such people as Markéta Marvanová Ph.D, Markus Storvik Ph.D, Suvi Asikainen M.Sc., Jussi Paananen M.Sc., Jani Kekäläinen M.Sc., Pekka Tiikkainen M.Sc., Matti Kankainen M.Sc., Kaja Reisner M.Sc., Outi Kontkanen Ph.D, Ms Anne Lehtelä and Martinj van Iersel M.Sc.
I also thank the whole personnel of A. I. Virtanen Institute. Practical issues seemed to run smoothly and help was always available also from other research groups. The atmosphere in the institute was one of openness and I hope it will stay like that in the future.
Thanks goes to my loved Ilkka for reviewing my language for the first time and for trying to understand what is this all about. To the genuine Finnish horse Toivo with whom my thoughs could not be anywhere else but at the present moment.
The Ministry of Education and Sigrid Juselius Foundation have financially supported this study.
Kuopio, March 2007
Suvi Vartiainen
ABBREVIATIONS
5-HT serotonin
6-OHDA 6-hydroxy dopamine AcH acetylcholine
AD Alzheimers Disease ADE anterior deirid neurons
ADPKD autosomal dominant polycystic kidney disease ANOVA analysis of variance
AR-JP autosomal recessive juvenile parkinsonism ATP adenosine triphosphate
cDNA complementary DNA
CeDAT C. elegans dopamine transporter CEP cephalic neurons
CMA chaperone mediated autophagy COMT catechol-O-methyl carboxylase
DA dopamine
DBS deep brain stimulation DLB dementia with Lewy Bodies DNA deoxyribonucleic acid dsRNA double-stranded RNA EMS ethyl methanesulfate ER endoplasmic reticulum FuDR 5-fluorodeoxyuridine GABA γ-amino butyric acid GCI glial cytoplasmic inclusions GFP green fluorescent protein GNI glial nuclear inclusions Glu glutamate
GO gene ontology Gpi globus pallidus HD Huntington's disease Hsp heat shock protein
IPTG isopropyl ß-D-1-thiogalactopyranoside
LB Lewy body
LRR leucine-rich repeat
LRRK2 leucine rich repeat kinase 2 MAO-B monoamine oxidase B miRNA micro RNA
MM mismatch
MPTP 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine MSA multiple system atrophy
NAC non-AB-component of amyloid plaques
NACP precursor of non-AB-component of amyloid plaques PCR polymerase chain reaction
PD Parkinson's Disease PDE posterior deirid neurons PINK1 PTEN -induced kinase 1 PM perfect match
PolyQ polyglutamine
PTEN phosphatase and tensin homolog qRT-PCR quantitative reverse transcribed PCR RdDM RNA-directed DNA methylation RdRP RNA-directed RNA polymerase RISC RNA-induced silencing complex RNA ribonucleic acid
RNAi RNA interference RT room temperature RT-PCR reverse transcribed PCR SD standard deviation SEM standard error of the mean
SPET single photon emission tomography siRNA short interfering RNA
SNP single nucleotide polymorphism SNpc substantia nigra pars compacta SNpr substantia nigra pars reticulata STN subthalamic nucleus
TGS transcriptional gene silencing TH tyrosine hydroxylase
TMP trimethylpsoralen
UCHL1 ubiquitin carboxy-terminal hydrolase L1 UPS ubiquitin-proteasome system
UV ultra violet
VAChT vesicular acetylcholine transporter VMAT vesicular monoamine transporter VNC ventral nerve cord
WT wild type
LIST OF ORIGINAL PUBLICATIONS
This thesis is based on the following publications referred to by their corresponding Roman numerals (I-IV):
I Lakso, M., S. Vartiainen, Moilanen,A., Sirviö J., Thomas J., Nass R., Blakely R., Wong G. (2003). "Dopaminergic neuronal loss and motor deficits in Caenorhabditis elegans overexpressing human α-synuclein." Journal of Neurochemistry 86(1): 165- 172.
II Vartiainen, S., Pehkonen P, Lakso M, Nass R, Wong G. (2006). "Identification of gene expression changes in transgenic C. elegans overexpressing human alpha- synuclein." Neurobiology of Disease 22(3): 477-486.
III Vartiainen, S., Aarnio V, Lakso M, Wong G. (2006). "Increased lifespan in transgenic Caenorhabditis elegans overexpressing human alpha-synuclein."
Experimental Gerontology 41(9): 871-876.
IV Vartiainen, S., Lakso M., Wong G. “Identification of genes modifying movement in a C. elegans model of Parkinson's Disease by RNA interference” In preparation.
TABLE OF CONTENTS
1. INTRODUCTION --- 15
2. REVIEW OF THE LITERATURE--- 18
2.1. PROTEIN α-SYNUCLEIN --- 18
2.1.1. Proposed function of α-synuclein --- 18
2.1.2. Structure and membrane binding properties of α-synuclein --- 19
2.1.3. Aggregation and fibrillization --- 21
2.1.4. Degradation--- 24
2.2. SYNUCLEOPATHIES --- 26
2.2.1. Dementia with Lewy Bodies --- 26
2.2.2. Parkinson’s Disease--- 28
2.2.2.1. Models of Parkinson’s Disease--- 31
2.2.3. Multiple System Atrophy --- 33
2.3. CAENORHABDITIS ELEGANS --- 34
2.3.1. Basic worm biology --- 34
2.3.2. Nervous system of C. elegans--- 35
2.3.3. Mutagenesis --- 38
2.3.4. Transgenesis --- 39
2.3.5. DNA Arrays --- 40
2.3.6. RNA interference in C. elegans--- 42
2.3.7. High throughput screens --- 45
2.4. C. ELEGANS AS A DISEASE MODEL--- 45
2.4.1. C. elegans and genetic diseases --- 45
2.4.2. C. elegans and addiction--- 46
2.4.3. C. elegans and infectious diseases --- 47
2.4.4. C. elegans models of neurodegenerative diseases--- 47
3. AIMS OF THE STUDY --- 50
4. MATERIALS AND METHODS--- 51
4.1. MAINTENANCE AND STORAGE OF C. ELEGANS --- 51
4.2. TRANSGENIC ANIMALS --- 51
4.2.1. Production of transgene constructs --- 51
4.2.2. Production of transgenic animals --- 52
4.2.3. Transgene integration--- 52
4.2.4. C. elegans lines used --- 53
4.2.5. Crossing of worm lines --- 53
4.2.6. Synchronization of C. elegans cultures --- 54
4.3. METHODS FOR SCREENING α-SYNUCLEIN PHENOTYPE IN C. ELEGANS --- 54
4.3.1. Behavioural assays--- 54
4.3.2. Screening for dopaminergic neurons --- 55
4.3.3. Immunoassays--- 56
4.4. MICROARRAY ANALYSIS --- 57
4.4.1. Affymetrix technology --- 57
4.4.2. Sample preparation--- 57
4.4.3. Data generation and analysis--- 57
4.4.3.1. dChip 1.3 and Genespring 7.2--- 57
4.4.3.2. Data analysis--- 58
4.5. qRT-PCR --- 59
4.5.1. Primer design --- 59
4.5.2. RT-PCR protocols --- 59
4.6. AGING STUDY --- 60
4.6.1. Aging protocol--- 60
4.7. SCREENING THROUGH THE FIRST CHROMOSOME OF C. ELEGANS WITH RNA INTERFERENCE TECHNIQUE --- 61
4.7.1. Screening protocol--- 61
4.7.2. Data collection and quantification--- 61
4.8. STATISTICAL ANALYSIS --- 62
5. RESULTS--- 63
5.1. EFFECT OF EXTRACHROMOSOMAL α-SYNUCLEIN TRANSGENE EXPRESSION ON C. ELEGANS --- 63
5.2. GENE EXPRESSION CHANGES INDUCED BY α-SYNUCLEIN --- 64
5.3. DIFFERENT AGING PROFILES --- 65
5.4. CHROMOSOME I-WIDE RNA INTERFERENCE SCREEN ON α-SYNUCLEIN EXPRESSING C. ELEGANS --- 66
6. DISCUSSION --- 69
6.1. NEURODEGENERATION AND MOVEMENT IMPAIRMENT INDUCED BY EXTRACHROMOSOMAL EXPRESSION OF HUMAN α-SYNUCLEIN --- 69
6.2. TRANSGENE α-SYNUCLEIN INDUCED CHANGES IN GENE EXPRESSION ASSAYED WITH FULL GENOME MICROARRAY STUDY--- 71
6.3. AGING PROFILE OF α-SYNUCLEIN TRANSGENIC C. ELEGANS IN DIFFERENT GENETIC BACKGROUNDS--- 72
6.4. EFFECTS OF FIRST CHROMOSOME-WIDE SINGLE GENE KNOCKDOWN STUDY --- 73
6.5. POSSIBLE MECHANISMS FOR α-SYNUCLEIN INDUCED NEURODEGENERATION AND MOVEMENT IMPAIRMENT--- 74
6.6. UTILITY OF C. ELEGANS IN MODELLING HUMAN NEURODEGENERATIVE DISEASES --- 77
7. SUMMARY --- 79
8. REFERENCES --- 81
1. INTRODUCTION
The post-genomic era began with the publicly available full genome sequence data from different organisms. The human genome sequencing projects especially opened up new possibilities to genetically dissect human diseases and disorders.
However, a collection of new approaches and technologies were needed to examine and combine data from multiple sources. The term functional genomics was introduced in order to incorporate the massive genomic level data to gene and protein functions and interactions. Functional genomics uses sophisticated molecular biology methods.
The focus of the research is aimed at the function of a gene and gene interdependency. Methods such as DNA microarrays, gene silencing and transgenesis in combination with bioinformatics tools provide the backbone for the experiment design. Nevertheless, gene expression is not the finale of the biological event, but rather a starting point, thus proteomics tools are also utilized to dismantle the event to its core components.
Modeling human functions with different organisms has been a major starting point for biological research since its beginning. A ‘model organism’ is used when a distinct biological phenomenom is studied in a living being. If the study is succesfully accomplished, the discoveries made with one organism can be extrapolated into the biological operations of other organisms. There is a wide range of extensively used model organisms, and each have their own properties that are advantagous or disadvantageous. It is of crucial importance to select the best possible model for a given investigation. Small models, such as yeast, have similar cellular structure to multicellular organisms. The studies made with Saccharomyces cerevisiae and Schizosaccharomyces pombe revealed the existence of cell cycle proteins common to all eukaryotes (Hartwell et al., 1970; Lee and Nurse, 1987; Nurse, 1975); these studies were awarded the Nobel Prize in 2001. Small invertebrates, e.g. Drosophila melanogaster and Caenorhabditis elegans, have a fundamental advantage as research objects: they are multicellular organisms whose specialized cell types have counterparts in human. The maintenance of invertebrate models is fairly simple and they generally have a short life span, which accelerates the research. A wide spectrum of vertebrate models are used in research, ranging from zebra fish (Danio rerio) to non- human primates, the latter being the closest to human in the phylogenetic tree.
The nematode C. elegans was selected as a model organism in studies of developmental biology. For this purpose C. elegans had remarkable advantages:
transparency, small size, ease of maintenance etc. Highly consequential findings were made from the studies on cell fate during development, including the discovery of programmed cell death (apoptosis) inducing proteins (Ellis et al., 1991). This study, together with earlier data on cell lineage from fertilization to adult, led to the Nobel Prize in Physiology or Medicine in 2002. Another groundbreaking study on C. elegans was recently carried out by professors Andrew Fire and Craic C. Mello. They discovered that double stranded RNA was the molecule inducing RNA interference (Fire et al., 1998). This study was also awarded with the Nobel Prize in Physiology or Medicine (2006).
Alzheimer's Disease and Parkinson’s Disease are the two major neurodegenerative diseases of the modern world. The disease onset is age related and thus they will affect more people in the future as a greater proportion of the human population has the possibility to reach older age. The effects of neurodegeneration are devastating for the patient. As the disease progresses, patients will require constant medical and supportive care from society. Due to the progressive nature of the neurodegenerative diseases, and current treatments being able to provide only limited neuroprotection at best without any curative or preventive therapy available, they present an enormous task to researchers of many discliplines. Parkison’s Disease belongs to the family of synucleopathies, the common feature of which is the accumulation of α-synuclein rich protein aggregates. The mechanisms leading to pathogenic function of α-synuclein are unresolved, albeit active and long lasting research has been conducted. If the pathogenesis of Parkinson’s disease and other synucleopathies can be determined, it would provide a new path for development of novel therapeutics that could possibly even prevent the onset of the disease.
In this dissertation, we have studied the effects of transgenic human α-synuclein and its disease-causing mutant A53T in the model organism C. elegans. The study began with creation of an overexpression model for α-synuclein in C. elegans, which was followed by assessment of the effects of the overespression by diverse methods.
The model showed a similar phenotype to Parkinson’s Disease, demonstrating movement impairment and loss of dopaminergic neurons (Lakso et al., 2003). The levels of gene expression affected by α-synuclein overexpression were studied at the
whole genome level with DNA microarrays (Vartiainen et al., 2006b). Single gene silencing with RNA interference was carried out on chromosome I, to study the pathogenesis and whether our model could be used for such a screen. The effect of aging mutants on α-synuclein nematodes was also tested (Vartiainen et al., 2006a).
The results from this work support the use of model organisms such as C. elegans to study neurodegenerative and neuropathologic mechanisms and gain insight into human disease.
2. REVIEW OF THE LITERATURE
2.1. PROTEIN α-SYNUCLEIN
2.1.1. Proposed function of α-synuclein
Protein α-synuclein was first isolated from Torpedo californica (Maroteaux et al.
1988) where it was localized to presynaptic nerve terminals and to the nuclear envelope, hence the name synuclein. Interest towards α-synuclein in neurodegeneration was first raised when the 35 amino acid non-AB-component (NAC) of amyloid plaques found in Alzheimer patients was solubilized and the precursor of this peptide named NACP was first cloned and sequenced (Ueda et al., 1993). Two homologous peptides of 134 and 140 amino acids were isolated from human brain, the second of these being identical to the NACP (Jakes et al., 1994). The NACP was found to be homologous to the rat synuclein and named as α-synuclein and the 134 amino acid protein was named as β-synuclein (Jakes et al., 1994). The homology to the rat synuclein was substantiated by Campion D, (1995) and Iwai et al. (1995). When Spillantini et al., (1997) found with immunostaining that α-synuclein was the major component of Lewy bodies, intracellular protein aggregates found in Parkinson’s Disease and dementia with Lewy bodies, the research around α-synuclein was further accelerated.
The exact function of α-synuclein is not currently known. It is expressed abundantly throughout the brain. It’s localization to presynaptic nerve terminals has suggested that is has a role in synaptic transmission. There is evidence from avian researchers who found that the α-synuclein homolog synelfin RNA is upregulated during song learning in the zebra finch, Taeniopygia guttata (George et al., 1995).
Results from α-synuclein knockout studies (Cabin et al., 2002; Murphy et al., 2000) demonstrate decrease of the synaptic vesicles in the presynaptic pool. However, the amount of vesicles docked at the presynaptic plasma membrane – an indication of immediate availability for release to the synaptic cleft – was unchanged. In α-synuclein knockout mice studies, prolonged stimulation showed a great decrease in synaptic response, suggesting that vesicles are depleted faster in knockouts and supporting the idea that α-synuclein functions in the pathway of synaptic vesicle development (Cabin
et al., 2002). Also α-synuclein mRNA and protein levels have been shown to be elevated with cocaine abusers (Mash et al., 2003) and in cocaine treated rats (Brenz Verca et al., 2003).
Another function for α-synuclein is its proposed chaperone activity. α-Synuclein protein has sequence homology to 14-3-3-family proteins and it binds to similar ligands as 14-3-3 protein chaperones like protein kinase C and BAD (Ostrerova et al., 1999).
α-synuclein's ability to prevent aggregation of both thermally denatured alcohol dehydrogenase and chemically denatured insulin provides evidence of its chaperone activity (Souza et al., 2000).
2.1.2. Structure and membrane binding properties of α-synuclein
α-Synuclein is a small 19 kD protein consisting of 140 amino acids (Figure 1).
There are seven imperfect 11-amino-acid repeats (residues 1-87) which lie in the N- terminal region (residues 1-60). Most of the repeats contain the motif KTKEGV which is similar to the alpha-helix domain seen in membrane binding apolipoproteins. The central region (residues 61-95) harbors the highly hydrophobic NAC peptide which is found in Alzheimer plaques (Ueda et al., 1993). The hydrophilic C-terminal region (residues 96-140) is very acidic and the proposed Ca2+ binding motif lies there (Nielsen et al., 2001). Deletion of the C-terminus reduced the chaperone activities of α- synuclein (Kim et al., 2002). Phosphorylation of α-synuclein occurs on serine residues and most prominently on serine 129, phosphorylation is induced by casein kinase 1 and 2 (Okochi et al., 2000). The other two proteins of the family are β- and γ-synuclein which are homologous to each other. There is no direct link between β- or γ-synuclein and neurodegenerative diseases but some evidence is presented on their effects on each other. As an example, transgenic β-synuclein has been shown to decrease the amount of α-synuclein protein in mouse brain (Fan et al., 2006).
There are four disease-linked mutations found in α-synuclein protein. The first mutation A53T was found in Greek families with hereditary Parkinson’s Disease (Polymeropoulos et al., 1997), the second mutation A30P was found in a German family with susceptibility to Parkinson’s disease (Kruger et al., 1998). The third point mutation E46K in α-synuclein was recently found in a Spanish family suffering from parkinsonism and Lewy Body Dementia (Zarranz et al., 2004). A locus triplication of
wild type (WT) α-synuclein leading to autosomal dominant Parkinson’s disease was found by Singleton (2003). However, in a study where 190 patients with familial Parkinson’s disease were investigated for changes in gene dosage, no positives were found (Gispert et al., 2005). Later studies have found positive results from synuclein gene dublication and triplication in families with familial Parkinson’s Disease (Fuchs et al., 2007; Kenya Nishioka, 2006) showing that increase in gene dosage can lead to human parkinsonism.
N C
Acidic tail Imperfect KTKEGV repeats
Mutations Phosphorylation sites
A30P E46K A53T
S87 Y125 S129
Aggregation
61-95 hydrophobic NAC, promotes
aggregation
120-140 hydrophilic decreases aggregation
Figure 1. The structure of α-synuclein protein showing the sites of known pathological mutations A30P, E46K and A53T. The KTKEGV repeats on residues 1-87 assume α- helix structure when in contact with phospholipids (Chandra et al., 2003).
Phosphorylation sites are shown. S129 is phosphorylated also in α-synuclein observed in Lewy Bodies (Anderson et al., 2006). The figure is adapted from (Cookson, 2005).
α-Synuclein has two locations in the nerve terminal: it is found both in cytosol and in the structures of the synaptosome (Kahle et al., 2000; Maroteaux and Scheller, 1991). In its cytosolic free-form state the protein has been considered as natively
unfolded and in conformational studies it has proven resistant to various stress factors, such as high or low pH, heat treatment and different salt concentrations (Weinreb et al., 1996). Neither A53T nor A30P mutation was seen to affect this random coil conformation of WT protein (Conway et al., 1998; Li et al., 2001). More recent data, nonetheless, showed that monomeric α-synuclein populates a wide range of dynamically changing conformations (Bertoncini et al., 2005b) and that familial mutants A30P and A53T cause distortions toward the WT conformation (Bertoncini et al., 2005a). The data reporting α-synuclein to have a continuum of conformations instead of being natively unfolded is still very new and needs more confirmation.
The α-helix structures in the N-terminal of α-synuclein define its membrane binding properties, while the 40 amino acids in the C-terminus do not associate with membranes (Eliezer et al., 2001). α-Synuclein has been shown to bind to phospholipid monolayers surrounding triglycerides (Cole et al., 2002) and to small synthetic unilamellar phospholipids (Davidson et al., 1998). α-Synuclein conformation has been seen to change from unfolded to α-helix structure when the protein is incubated with small unilamellar vesicles that contain phosphatidylserine. This indicates that phospholipids induce protein folding of α-synuclein (Chandra et al., 2003; Davidson et al., 1998). The mutation A30P has been shown to avoid vesicle binding (Cole et al., 2002; Fortin et al., 2004; Jensen et al., 1998) and thus remain mainly cytosolic (Cole et al., 2002). Unlike A30P, the most recent mutation E46K has been shown to increase phospholipid binding (Choi et al., 2004).
2.1.3. Aggregation and fibrillization
In in vitro studies the natively unfolded protein α-synuclein and its A53T and A30P mutants have both been shown to aggregate and to further fibrillize in the absence of other Lewy body associated molecules. Moreover, the formed fibrils resembled those seen in Lewy bodies (Conway et al., 1998). It is unknown how the aggregation, also referred to as oligomerization, and further fibrillization happens in vivo or what the consequences of this phenomenon encompass. Numerous studies have been performed in order to understand the perplexity of α-synuclein pathogenesis. One of the problems has been to determine whether it is the cytosolic or the membrane bound form that forms the aggregates. On the basis of the current data, the membrane bound α-synuclein is considered to be the cause of aggregation. A
recent study described how in vitro incubation of α-synuclein with brain fractions resulted in high molecular weight aggregations of α-synuclein. The same study also showed that aggregates were found in membrane fractions but not in cytosolic fractions. Nevertheless, increased cytosolic concentration of α-synuclein accelerated the rate of aggregation (Lee et al., 2002).
The aggregation of α-synuclein is dependent on its hydrophobic NAC region (Bodles et al., 2001; Giasson et al., 2001), and it is restrained by the acidic C-terminus (Hoyer et al., 2004; Murray et al., 2003; Serpell et al., 2000). β-Sheet structure is found in amyloid fibrils (Sunde et al., 1997) sharing similarity with fibrillized α-synuclein.
During fibril formation the conformation of α-synuclein monomer changes from random coil to β-sheet structure in in vitro conditions when no membranes are present (Serpell et al., 2000). To date no study of β-sheet formation in α-synuclein aggregates in a membrane rich environment has been performed.
Mutations A53T, A30P and E46K have different effects on aggregation and fibrillization. The A53T and E46K mutants increased the rate of filament assembly, whereas WT and A30P mutant did not (Choi et al., 2004; Conway et al., 1998).
Another study showed that also the A30P mutation is capable of increasing the rate of aggregation although less aggressively than A53T (Li et al., 2001). Increased α- synuclein levels in gene locus triplication patients boosts protein aggregation in brain (Miller et al., 2004).
Figure 2. The proposed model for α-synuclein aggregation and Lewy Body formation.
The pathogenic cascade is not fully understood, but it is probable that the membrane bound α-synuclein originates the process and the cytosolic monomers are recruited to form oligomers which then further fibrillize to and form Lewy Bodies. Formation of annular protofibrils is well documented and there is evidence that oxidative ligation of dopamine is selective for protofibrils thus causing their accumulation (Conway et al., 2001). The figure is reproduced from (Cookson, 2005).
The pathway from α-synuclein monomer to fibrillar aggregates has not been completely revealed (Figure 2). At the moment, the most widely accepted hypothesis states that the membrane bound aggregate acts as a nuclei which then seeds the aggregation of cytosolic monomers (Lee et al., 2002). The next step is claimed to be the most crucial one for α-synuclein pathology. Already in the first fibrillization studies, small spherical α-synuclein aggregates were detected before fibril formation (Conway et al., 1998) (Figure 3). The pre-fibril structures of aggregated -synuclein are referred
to as protofibrils or as oligomers. The protofibrils seem to have potential to form either β-sheet fibril structure or spherical pore-like species. Permeabilization of membranes caused by these spherical α-synuclein compositions was studied by Volles and Lansbury (2002) and reported that protofibrils from A30P and A53T α-synuclein had higher vesicle permeabilizing activity than WT α-synuclein protofibrils. They also showed that high concentration of monomeric α-synculein except A30P monomer can permeabilize unstable vesicles. Another study established a cellular model in which fragmentation of Golgi Apparatus was found to correlate with the presence of prefibrillar α-synuclein species (Gosavi et al., 2002). Observations of human post mortem brain in which nerve cells containing Lewy Bodies appeared healthier than neighboring neurons also point towards oligomers being toxic instead of fibrils (Goldberg and Lansbury Jr, 2000).
Figure 3. Spherical pore-like α-synuclein protofibril that could have properties for membrane permeabilization. The size of the pore is about 3-5 nm. The picture is reproduced from Goldberg and Lansbury (2000).
2.1.4. Degradation
There are three pathways for protein degradation that α-synuclein has been shown to utilize. The most studied of these in the context of α-synuclein is the ubiquitin-proteasome pathway. This is most likely due to the fact that mutations in two enzymes on the pathway have been characterized to cause juvenile parkinsonism.
Mutations have been found in the parkin gene in a Japanese family (Kitada et al., 1998). German siblings with Parkinson´s Disease had a mutation in the gene ubiquitin carboxy-terminal hydrolase L1 (UCHL1) (Betarbet et al., 2005; Leroy et al., 1998) but
more supporting data is needed about the neuropathologic effects of the Ile93Met UCHL1 mutation.
The ubiquitin-proteasome system (UPS) is considered a method for breaking down dysfunctional and short lived proteins. UPS proteins are tagged with ubiquitin and when four or more ubiquitin moieties are attached, the 19S proteasome recognizes the degradable object. The protein parkin is one of the E3 ubiquitin ligases functioning as a last step in ubiquitin tagging. Before the active period of E3, two other enzymes, namely E1 ubiquitin activator and E2 ubiquitin conjugating enzyme are needed. It is of special interest that a single E1 enzyme is capable of transferring ubiquitins to several E2 enzymes, each of which can activate various E3 enzymes. In studies with α-synuclein, it has been shown that both WT and A30P deteriorate proteasome function in a PC12 cell line (Tanaka et al., 2001) and in yeast (Chen et al., 2005).
Figure 4. The ubiquitin-proteasome system which is used for degradation of proteins.
Ubiquitin (Ub) molecule is attached to the substrate (S) in an ATP requiring process with enzymes E1 (ubiquitin activating enzyme), E2 (ubiquitin conjugating enzyme) and E3 (ubiquitin ligase). When a minimum of four ubiguitin moieties are attached to the substrate it can enter to the 26S proteasome to undergo degradation in an ATP dependent process. The figure is reproduced from (Betarbet et al., 2005).
Macroautophagy is considered a bulk degradation of cytosolic proteins and organelles. In macroautophagy, also referred to as autophagy, cytoplasmic membranes form an autophagosome which then fuses with lysosome, the lytic enzymes of which degrade the contents of the autophagosome. Proteins containing aggregation prone polyglutamine [poly(Q)] and polyalanine [poly(A)] expansions were degradated by macroautophagy (Ravikumar et al., 2002), and later the same was proved to happen with α-synuclein (Webb et al., 2003). Even the proteasome inhibition induced α-synuclein inclusion bodies can be dismantled into small ubiquitinated aggregates by macroautophagy after 48 h incubation (Rideout et al., 2004). The other form of autophagy that has been connected with α-synuclein degradation is chaperone mediated autophagy (CMA). CMA differs from other forms of autophagy as no vesicles are needed. Instead soluble cytosolic proteins are recognized from the KFERQ motif by heat shock 70kDa protein (hsc70). The substrate protein-hsc70-cochaperone complex binds to lysosome associated membrane protein 2a (lamp2a) at the lysosome membrane. Transmembrane protein lamp2a unfolds the substrate protein and transfers it into the lysosomal lumen (Majeski and Dice, 2004). α-Synuclein has the motif VKKDQ which is related to the CMA motif and it was shown to be degraded by CMA (Cuervo et al., 2004). Although the mutations A30P and A53T do not lie within the CMA motif, the mutated proteins bind the lamp2a receptor more tightly than WT α- synuclein (Cuervo et al., 2004). Thus, the mutant proteins are not translocated into the lysosomal lumen. Instead of being degradated, they impair the dismantlement of other substrates of CMA (Cuervo et al., 2004).
2.2. SYNUCLEOPATHIES
2.2.1. Dementia with Lewy Bodies
Lewy bodies, named after their discoverer Frederich Lewy, are intracellular inclusion bodies that are rich in protein. The Lewy bodies (LB) have been seen to contain ubiquitin and neurofilaments in addition to α-synuclein (Spillantini et al., 1997).
The classical LBs are spherical and have a pale halo consisting of radially oriented filaments around them. Less uniform are the Lewy neurites that are cyto- or axoplasmic elongated inclusions that were also first defined by Lewy in 1912. In
Dementia with Lewy Bodies (DLB), the inclusion bodies are located at the neocortex, the limbic cortex, sub cortical nuclei, and the brainstem (Gräber and Müller, 2003).
Figure 5. Lewy Bodies in substantia nigra from patients with Parkinson’s Disease. In picture A there is a neuron with four Lewy Bodies three of which are positive for both ubiquitin and α-synuclein antibody (scale bar 30 um). Picture B shows three Lewy Bodies within one neuron all being positive for both ubiquitin and a-synuclein antibodies. The pale halo is immunoreactive for ubiquitin (scale bar 10 um). The figure is reproduced from (Spillantini et al., 1998).
The definition for DLB is still evolving with the main feature being cognitive impairment, but that may not be the key symptom in the beginning of the disease. As the disease progresses, the memory impairment becomes persistent. Other features include visual hallucinations and falls. When compared to patients with Alzheimer’s Disease (AD), the DLB patients have more difficulties in tests requiring attention and visual perception (Calderon et al., 2001).
Although DLB is a common form of dementia, accounting for approximately 20%
of dementia cases, it is still largely misunderstood and often diagnosed as AD (Weiner, 1999). In order to help diagnostics of DLB, a consensus criteria has been established.
Together with cognitive impairment, two of the following features must be met for diagnosis of DLB (McKeith et al., 1996):
a) Fluctuating cognition with pronounced variations in attention and alertness.
b) Recurrent visual hallucinations which are typically well formed and detailed.
c) Spontaneous motor features of parkinsonism.
In addition to cognitional testing, brain imaging techniques have been used to study both AD and DLB brain. A magnetic resonance imaging (MRI) study (Burton et al., 2002) found better preserved grey matter in medial temporal lobe structures in the brains of DLB patients compared to AD. In the study by Walker et al. (1999), an in vivo single photon emission tomography (SPET) was suggested to distinguish DLB patients from AD cases.
There is no curative treatment for DLB, however cholinesterase inhibitors have been reported to have a positive impact on patients (McKeith, 2002). Although DLB patients often show psychotic symptoms, treatment with neuroleptics must be done with great caution since it can result in severe side effects like irreversible parkinsonim and further impairment of the level of consciousness (McKeith, 2002).
2.2.2. Parkinson’s Disease
James Parkinson described the disease in 1817. Since then it has been characterized as the second most common neurodegenerative disease after Alzheimers Disease (AD), affecting approximately 1% of the 50 year old population (Polymeropoulos et al., 1996). The main symptoms of Parkinson’s disease (PD) are resting tremor, muscle rigidity and bradykinesia (slowness of movement). Other characteristic features are postural abnormalities, dysautonomia, dystonic cramps, mask-like facial expression, dementia and festinating gait where very small and accelerated steps are taken during walking (Goedert, 2001). The symptoms are apparent when as many as 80 % of dopaminergic neurons in substantia nigra pars compacta (SNpc) have lost their function (Nussbaum and Polymeropoulos, 1997). The histological hallmarks of this disease are LBs located particularly at neurons in substantia nigra.
Substantia nigra is one of the basal ganglia which belong to the extrapyramidal circuit and it is divided into two compartments: pars reticulata (SNpr) and pars compacta (SNpc), both of which have different signalling pathways. In a simplified model, SNpc receives inhibitory signals from striatum, the neurotransmitter being γ- amino butyric acid (GABA). SNpc signals back to the medium spiny neurons in striatum with dopamine (DA). Medium spiny neurons have large dendritic fields which allow them to collect information also from cerebral cortex, large and medium spiny
neurons, thalamus and small interneurons. The receptors in medium spiny neurons define the action of DA. In humans, five types of G-protein coupled DA receptors have been characterized (named D1-D5) and divided into two types D1 (receptors D1 and D2) and D2 (receptors D3, D4, D5). The symptoms in PD result from depletion of dopaminergic signalling which normally has a modifying effect towards muscle contraction.
Cerebral cortex
Striatum
Thalamus
GPe
STn
SNpc Gpi, SNpr
Figure 6. Simplified model of the neuronal circuitry in the basal ganglia. The neuronal input is received from the cerebral cortex and passed through the striatum. Globus pallidus interna (GPi), SNpr and SNpc receive signal from striatum through a direct or an indirect pathway, latter of which utilizes Globus pallidus externa (Gpe) and subthalamic nucleus (STn). The figure is adapted from (Nestler et al., 2001).
The mechanisms of how the neurodegeneration is targeted to dopaminergic neurons in PD have raised many questions. Oxidative stress that could arise either from DA metabolism or impaired mitochondrial function, has been seen to increase α- synuclein levels in vitro in rotenone treated cell cultures (Sherer et al., 2002). α- Synuclein also reacts with the oxidized form of DA, DA-quinone, and forms covalent DA-α-synuclein adducts which in turn have a stabilizing effect in α-synuclein protofibrils (Conway et al., 2001). On the other hand, α-synuclein has been suggested to have a
role in sustaining DA homeostasis. In that model, impairment of α-synuclein function as a tyrosine hydroxylase regulator as well as in DA packing to vesicles in conjunction with vesicular monoamine transporter (VMAT) would lead to increased levels of cytosolic dopamine and thus increased oxidative stress and α-synuclein aggregation (Perez and Hastings, 2004).
In addition to the α-synuclein gene, also other mutations affecting the likelihood of PD have been found. The mutated proteins work mainly in the pathway of protein degradation. Mutations in protein parkin were found in patients with autosomal recessive juvenile parkinsonism (AR-JP) (Kitada et al., 1998). Later this protein was described as E3 ubiquitin-protein ligase (Shimura et al., 2000). Ubiquitin carboxy- terminal hydroxylase L1 (UCH-L1) was found to carry a missense mutation I93M in a German family with an autosomal dominant form of PD (Leroy et al., 1998). Another mutation in UCH-L1 S18Y found in an Asian population was however seen to decrease the risk for PD (Maraganore et al., 1999; Satoh and Kuroda, 2001). A third gene, mutations of which were found responsible for autosomal dominant PD, is leucine rich repeat kinase 2 (LRRK2 or dardarin) (Khan et al., 2005).
Mutations linked to recessive autosomal parkinsonism are harboured within the oncogene DJ-1. The mutation L166P found by Bonifati et al. (2003) was later seen to disrupt the structure needed for formation of functional DJ-1 (Tao and Tong, 2003).
The function of DJ-1 is not completely understood. It has been seen to positively regulate androgen receptor (Takahashi et al., 2001), and when down-regulated by RNA interference it promotes oxidative stress, stress to endoplasmic reticulum (ER) and proteasome inhibition in vitro (Yokota et al., 2003). In cancer studies, DJ-1 was suggested to have a role as a negative regulator of phosphatase and tensin homolog (PTEN) (Kim et al., 2005). Recessive autosomal parkinsonism is linked also to PTEN – induced kinase1 (PINK1) gene whose predicted function is mitochondrial serine threonine kinase. The PINK1 mutation G309D and truncation mutation W437OPA were found by Valente et al. (2004a; 2004b). Since then, several other mutations in the gene have been found (Hatano et al., 2004). The PINK1 mutations are not considered to cause early onset parkinsonism but rather to influence the pathological pathway of non-mendelian PD (Healy et al., 2004; Rogaeva et al., 2004).
There is no curative treatment for PD, but the symptoms are substantially relieved by Levodopa that is transformed to DA in dopaminergic and also in
nondopaminergic neurons. Levodopa is administered together with catechol-O-methyl carboxylase (COMT) inhibitors that minimize the peripheral Levodopa catabolism thus ensuring higher Levodopa levels in brain and eliminating the need to raise the dosage.
Monoamine oxidase B (MAO-B) inhibitors have both neuroprotective and DA degradation inhibiting properties which is one rationale for their use in both monotherapy and in combination with Levodopa (Thobois et al., 2005). Patients treated with Levodopa may develop motor fluctuations and dyskinesias as a side effect, furthermore, the efficacy of Levodopa wears out after 5-10 years of use (Mercuri and Bernardi, 2005; Thobois et al., 2005). Dopamine agonists, such as bromocriptine and apomorphine, react with post-synaptic dopaminergic type D2 receptors. Dopamine agonists are used as a monotherapy, especially in the early stages of PD, when the level of clinical therapy achieved is comparable to that of Levodopa (Thobois et al., 2005). Postponing Levodopa treatment has been reported to delay motor complications in patients treated with DA agonists (Montastruc et al., 1994).
Amantadine and anticholinergics have also been tested for PD treatment. Amantadine decreased PD dyskinesias but the duration of remedial effect was fairly short.
Anticholinergic treatment was shown to induce some improvement in motor activity of PD patients (Thobois et al., 2005). There are also surgical lesion therapies available that are used to relieve some of the symptoms of PD, such as bradykinesia, tremor and rigidity (Thobois et al., 2005). In the lesion of subthalamic nucleus (STN), the dosage of medicinal treatment could be decreased (Patel et al., 2003). Continuous electrical stimulation of brain in high frequency with an implanted electrode is called deep brain stimulation (DBS). In PD, DBS can be administered to thalamus, Gpi and STN. Studies have revealed the latter to be more efficient, leading to approximately 40% and 50 % improvement in motor score and the improvement considered permanent during the time of investigation (Walter and Vitek, 2004).
2.2.2.1. Models of Parkinson’s Disease
PD models can be divided into three categories; ageing of an intact organism, toxin induced model, and genetic manipulation model. Combinations of these three are also commonly used. The organisms used vary from cell free in vitro cultures to non- human primates, such as macaques and baboons. The most commonly used toxins are 6-hydroxy dopamine (6-OHDA) and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine
(MPTP), both of which cause specific neurodegeneration of dopaminergic neurons.
Treatment with 6-OHDA is considered as an acute model of neurotoxicity and thus fails to cover the progressive nature of PD pathology (Betarbet et al., 2002), however, 6- OHDA is considered as a very potent method for testing neuroprotective compounds (Betarbet et al., 2005; Schober, 2004). MPTP crosses the blood-brain-barrier (BBB) and it has been used to study the pathogenesis of PD, especially in monkeys (Betarbet et al., 2002). Environmental toxins like rotenone (insecticide) and paraquat (herbicide) also provide a model for PD pathogenesis studies. Chronic exposure to rotenone in rats (Betarbet et al., 2000) and in vitro (Testa et al., 2005) has led to damage to tyrosine hydroxylase (TH) positive neurons. All the toxins mentioned above have been proposed to have mitochondrial actions especially in the mitochondrial complex I, and subsequent oxidative stress is considered one mechanisms for the toxic effects (Betarbet et al., 2002).
Non-mitochondrial toxins are also available: inhibition of proteasome function via administration of lactacystin or carbobenzoxy-L-leucul-L-leucyl-L-leucinal (MG132) or some other proteasome inhibitor(s) has resulted in formation of α-synuclein positive inclusion bodies both in vitro (Rideout et al., 2001) and in vivo (McNaught et al., 2004;
Miwa et al., 2005). Studies on PD patients have demonstrated decreased proteasomal function in substantia nigra (McNaught and Jenner, 2001), which further highlights the role of proteasome in PD progression. In addition, α-synuclein itself has been reported to cause impairment to proteasome function (Chen et al., 2005; Lindersson et al., 2004).
Genetic disease models can be knockdowns or knockouts, transgenic over- expression models or models expressing mutant proteins. In fruit fly Drosophila melanogaster expression of α-synuclein WT or A53T covered the essential characteristics of PD (Feany and Bender, 2000) and knockout of parkin gene by null mutant resulted in degeneration of indirect flight muscles (Pesah et al., 2004). In cell culture of human neuroblastoma cells, knockdown of DJ-1 gene by short interfering RNAs induced accelerated cell death in hydrogen peroxide, MPP+ or 6-OHDA treated cells (Taira et al., 2004). The nature of PD is an adult onset progressive neurodegeneration and therefore many of the disease models are linked with ageing studies that compare symptoms of the model to the normal counterpart.
2.2.3. Multiple System Atrophy
Multiple system atrophy (MSA) is an adult onset neurodegenerative disease, however, the median age at onset is 56 years (Klockgether et al., 1998) and under 30 years of age is considered an exclusion criteria (Gilman et al., 1999). The symptoms include parkinsonian movement impairment and dysfunction of the autonomic nervous system. The disease is called MSA-P if parkinsonian features predominate and MSA-C if the main symptoms are cerebellar. MSA progresses rapidly and the mean survival time for patients is 9 years after diagnosis (Klockgether et al., 1998). The majority of MSA patients develop symptoms of PD (Gilman et al., 1999) and it is important to distinguish these two diseases. The main symptoms for MSA diagnosis are as follows and the diagnosis is considered possible if one of the criteria is met together with poor response to Levodopa treatment (Gilman et al., 1999):
Orthostatic hypotension and/or urinary features like incontinence Parkinsonian symptoms
Cerebellar features like different ataxias Dysfunction of corticospinal tract
The typical pathological finding in post mortem MSA patients are glial cytoplasmic inclusions (GCI) that mainly consist of α-synuclein (Gai et al., 2003). The GCIs together with glial nuclear inclusions (GNI) are primarily found in oligodendrocytes located at basal ganglia, cerebellum and autonomic nuclei (Christine and Aminoff, 2004). α-Synuclein mRNA has no apparent expression in oligodendrocytes but it has been proposed that low levels of expression together with decreased metabolism could induce aggregate formation in MSA (Wenning and Jellinger, 2005). The other possibility is that glial α-synuclein is of neuronal origin (Wenning and Jellinger, 2005).
The therapeutic options for MSA are not promising, although the parkinsonian symptoms can be relieved with Levodopa and other PD drugs and dysautonomic features can also be treated. For cerebellar symptoms, no effective treatment is available (Wenning et al., 2005).
2.3. CAENORHABDITIS ELEGANS 2.3.1. Basic worm biology
Active worm research was originated by Dr Sydney Brenner and co-workers in the early 1960s. The reference strain, named N2, was dug up from soil in Bristol. The transparency, small size (approx. 1 mm long) and ease of maintenance made the nematode Caenorhabditis elegans a good target for cellular analysis completed by Sulston and Horvitz (1977) and Sulston et al. (1983). For C. elegans, the change from free living nematode into model organism was not radical. In nature, the nematode Caenorhabditis elegans lives in soil and eats bacteria. Laboratory conditions alter their life style very little. The most common laboratory habitat for a worm is bacteria seeded on agar in a petri dish. Another option, especially suitable for growing large quantities of worms or for metabolic observations, is to grow them in a liquid culture with bacteria or nutrient solution. The most favorable temperatures for worms are the same as in nature. The most commonly used are +20°C or room temperature (RT). When kept at +15°C the lifespan of the worms lengthens and at +25°C it shortens. C. elegans has two sexes: hermaphrodite and male. Under self-fertilized reproduction the hermaphrodites produce ~300 eggs and when fertilized by males the number of eggs rises to ~1000. The C. elegans hermaphrodites have two X chromosomes (designated XX) and the males are hemizygous carrying only one X chromosome (designated XO).
While males can arise from self reproduction of hermaphrodites, mating with males increases the proportion of male progeny.
The lifespan of a worm is approximately three weeks and adulthood is reached in three days. The development of a worm from fertilization to adult is fully characterized and the cell lineage for every cell is known. During their development the worms undergo four larval stages, named L1, L2, L3 and L4, each of which includes molting. C. elegans also has an option to go into the dauer stage if living conditions turn unfavourable, the term dauer coming from german and meaning endurance or duration. The worms can turn into dauers after the L2 stage and can stay in this developmental arrest stage having minimal metabolism for months waiting for the conditions to change.
Figure 7. A young C. elegans hermaphrodite.
C. elegans was the first multicellular animal to have its genome sequenced (The C.elegans Sequencing Consortium, 1998) The completion of the C. elegans genome sequence without gaps was reached in 2002. The C. elegans genome consists of 100 Mb that encode 23977 genes, 92.9 % of these have molecular information, 33.7 % are confirmed genes, and 46.3 % have partial confirmation meaning that some but not all exon bases are covered with transcript evidence (wormbase version 170, March 2007, www.wormbase.org). The genes are organized into five autosomes and one sex chromosome termed X.
Forty years of worm research have yielded numerous methods for molecular biology, genetic studies, neurobiology and most recently for bioinformatics studies.
Worms are easy to handle and population studies are simple to execute. In addition to transparency, short life span and other advantageous features, the worms are complete organisms with nervous- and reproductive system, muscles and digestion, thus providing an opportunity to study genetic and molecular functions and interactions in an easily manageable system.
2.3.2. Nervous system of C. elegans
Similar to the cell lineage, the nervous system, comprised of 302 neurons in hermaphrodites, has also been characterized and the wiring pattern for each neuron is known (Figure 8). Most of the cell bodies of neurons are located at the nerve ring
which lies in the worm cephalic region around the pharynx bulbs (White et al., 1986).
Neurons can be roughly divided into: sensory neurons that collect sensory data from the environment; motor neurons that innervate muscles; and interneurons that mediate neuronal input between neurons. Nematodes use dopamine (DA), γ-aminobutyric acid (GABA), acetylcholine (AcH), glutamate (Glu), serotonin (5-HT) and neuropeptides for signal transduction. The signal transduction machinery is conserved and both worms and mammals have similar neurotransmitter transporters and receptors, only the voltage activated sodium channel gene being absent in worm (Bargmann, 1998).
However, gap junctions and olfactory receptors are present in worm although the encoding gene families are not related to the corresponding genes in vertebrates (Bargmann, 1998).
There are eight dopaminergic neurons in C. elegans hermaphrodites: four cephalic neurons (CEP), two anterior deirid neurons (ADE) and two posterior deirid neurons (PDE) (White et al., 1986). CEPs, ADEs and PDEs all have ciliated dendritic endings (White et al., 1986). Ablation studies have shown that worms need dopaminergic neurons for sensing food mechanistically. When on food, the worms respond by slowing the rate of movement, but when all dopaminergic neurons are ablated this action fails (Sawin et al., 2000). Males have six additional dopaminergic neurons located in the tail which play a part in mating behavior (Liu and Sternberg, 1995).
There are 26 GABAergic neurons in C. elegans. These neurons are further divided into six classes: DD, VD, RME, AVL, DVB and RIS (Mclntire et al., 1993; White et al., 1986). The DD and VD are D-motor neurons that mediate inhibitory GABA transmission to dorsal and ventral body wall muscles and thereby causing relaxation of the respective muscles. The sinusoidal movement of a worm is evoked by contracting muscles on one side of the body and relaxing them on the opposite side. When the GABAergic neurons were ablated, worms displayed movement impairment and hyper contracted when touched (Mclntire et al., 1993). The AVL and DVB neurons provide excitatory stimulus on enteric muscles during defecation. The RME neurons innervate head muscles and their ablation affects foraging behaviour. RIS is an interneuron that does not have any phenotype if ablated (Mclntire et al., 1993). The worm GABAergic system differs from vertebrates, for it uses mainly neuromuscular junctions instead of
neuronal synapses; also GABAergic neurons are proportionally fewer in worm (Schuske et al., 2004).
Cholinergic neurons in C. elegans are mostly excitatory. The number of cholinergic neurons is approximately 100 and they can be found in pharynx, ventral nerve cord (VNC), and from neurons innervating vulval muscles (Altun and Hall, 2002- 2006). The expression pattern of choline transporter Pcho-1::GFP showed expression in nerve ring and motor neurons on VNC (Okuda et al., 2000). In the body wall muscles, there are one type of GABA receptor and two types of acetylcholine receptors - levamisole and nicotine sensitive - all of which located at the neuromuscular synapse (Richmond and Jorgensen, 1999). Although cholinergic motor neurons are located at the VNC, they innervate both dorsal and ventral body wall muscles. The pharyngeal motor neurons M1, M2L, M2R, M4 and M5 are presumably cholinergic (Altun and Hall, 2002-2006) and evidence from synapse receptor composition between MC neurons and pharyngeal muscle demonstrate that MC neurons are cholinergic (McKay et al., 2004). The vulval muscle innervating neurons VC4, VC5, HSNL and HSNR express vesicular acetylcholine transporter (VAChT) and VMAT (Duerr et al., 2001). Also other VC neurons VC1-3 and VC6 are cholinergic. The acetylcholine released from VC neurons is considered inhibitory but the HSN neurons are excitatory. A complete egg laying defect induced by laser ablation of HSN neurons can be overcome by serotonin released from VC4 and VC5 neurons (Altun and Hall, 2002-2006).
Figure 8. C. elegans hermaphrodite with green fluorescent protein (GFP) expression in its dopaminergic and cholinergic neurons.
The expression of EAT-4, a homolog of mammalian brain-specific sodium- dependent inorganic phosphate cotransporter I (BNPI), is necessary for glutamatergic neurotransmission in C. elegans (Lee et al., 1999). In addition to the previously known glutamatergic neurons M3, ASH, IL1V, OLQV, and PVD, new neurons were found with eat-4::GFP and eat-4::lacZ expression patterns. These new eat-4 expressing neurons ADA, ALM, ASK, AUA, AVJ,AVM, FLP, IL1, LUA, OLL, OLQ, PLM, PVD, and PVR constitute a group of 34 possible glutamatergic neurons (Lee et al., 1999). GFP expression patterns have been used to investigate the localization of serotonergic and neuropeptide mediated signal transmission. Tryptophan hydroxylase (TPH-1) is needed for serotonin biosynthesis, and the Ptph-1::GFP fusion protein was seen to be expressed in the following neurons: ADF, AIM, CP, HSN, NSM and RIH (Sze et al., 2000). In the study by Nathoo et al. (2001) 32 neuropeptide-like protein genes (nlp-1 – nlp-32) were found in the C. elegans genome. In addition to these, also 22 FMRFamide (Phe-Met-Arg-Phe-NH2)-like neuropeptide (FaRPs) genes, termed flp-1 through flp-22, were found by Li et al. (1999). The expression pattern of these neuropeptides has been studied with GFP (Li et al., 1999; Nathoo et al., 2001).
2.3.3. Mutagenesis
C. elegans are mutated in vivo with methods selected for the purpose at hand.
The most common is the ethyl methanesulfate (EMS) that causes G/C -> A/T transitions. These transitions can yield functionally defective gene products that have identifiable phenotypes. EMS mutations are small and rather difficult to recognize with molecular biology tools, but the high potency of EMS mutagenesis makes it a useful tool for forward genetics research.
When genome rearrangements are under investigation, gamma, X-ray or UV irradiation treatment is a method of choice. Chemical treatment with trimethylpsoralen (TMP) combined with UV has been seen to cause a great number deletions at size ranges of 100 bp to 15 kbp, which can lead to a total loss-of-function phenotype in a single gene without inducing large deficiencies or genome rearrangements (Yandell et al., 1994).
Transposons are genetic elements that can change their location within the genome of a single cell thus causing spontaneous mutations. In C. elegans six transposons have been described and named Tc1-Tc6. These transposons are found
in high frequency in so called mutator strains which can be used to introduce spontaneous mutations to other C. elegans strains where they can be investigated more easily.
2.3.4. Transgenesis
Production of transgenic C. elegans is relatively easy and the transgenes can be targeted to selected tissues, or more precisely to cells, with endogenous C. elegans promoter sequences. The worms are injected with bacterial plasmids, phages, cosmids or yeast artificial chromosomes (Epstein and Shakes, 1995) carrying the transgene to both of the two gonad arms. The plasmids are taken into the developing egg nucleus where they form an independently replicating extrachromosomal array. A study by Mello et al. (1991) demonstrated that a concentration of DNA molecules over 100 µg/ml did not increase the transgenic population; furthermore, when two distinct DNA molecules were co-injected, the other DNA plasmid could be diluted to 1:100 and the extrachromosomal array including both transgenes was still formed. Co-injection of plasmids and phages was not effective for expressing two transgenes; Mello and co- workers (1991) postulated that regions of homology between injectected DNA molecules might be important for driving the assembly of the extrachromosomal array.
For analysis of transgenic progeny, co-injection is essential since it gives the possibility to inject the marker gene and the gene of interest in separate plasmids. Commonly used marker genes are GFP, (Lorenz et al., 1991) which can be seen with UV light, and rol-6, a dominant mutant causing easily identifiable roller phenotype (Kramer et al., 1990). The extrachromosomal arrays were heritable in a non-Mendelian manner if concentration and size of DNA molecules was sufficient (Mello et al., 1991).
In order to produce a stable transgenic worm line with Mendelian heritability, extrachromosomal arrays need to be integrated into the genome. Already in 1991, Mello et al. found that the integration was present and it could be promoted by simultaneus injection of single stranded oligonucleotide. A more straightforward method for achieving integrated lines is by gamma or X-ray irradiation (Riddle et al., 1997). The method dismantles genomic DNA, and as a part of its reconstruction the previously extrachromosomal transgene is taken into the genome. The irradiation
causes numerous genetic mutations, which need to be removed by outcrossing the integrated lines with worms with a normal genetic background.
2.3.5. DNA Arrays
The most convenient way to study genome-wide gene expression in an organism such as C. elegans are DNA arrays, also referred to as DNA chips or (DNA) microarrays. Gene expression changes can result from a given treatment, transgene, mutation, sex, etc. For higher organisms multiple single nucleotide polymorphism (SNP) arrays are available for genotyping needs. DNA arrays utilize a technology in which gene specific DNA probes (lengths varying from 25-70 bp) are immobilized into solid support. After the RNA is isolated from the investigated subject, it is enzymatically reverse transcribed into more stable cDNA. This cDNA contains the current gene- expression information and when hybridized into the microarray, the cDNA fragments find their respective DNA probes from the array hybridizing with them. Fluorescent labels are used to achieve signal from the hybridized probes, which in turn report which RNAs have been present in the the subject at the time of RNA isolation. There are several full genome DNA arrays available for C. elegans and most companies provide user specified custom arrays.
Figure 9. Affymetrix flowchart for labeling and hybridization of one Affymetrix GeneChip using total RNA as a starting material. The figure is reproduced from Affymetrix (www.affymetrix.com).
The amount of RNA needed for one microarray is high, therefore the question of starting material is of principal importance when making gene expression studies with C. elegans. The size of the nematode itself is far too diminutive for a single animal to provide an adequate quantity of isolated RNA. A couple of solutions have been developed to overcome this problem. The first and the most straightforward method is
to grow a large amount of worms and isolate the RNA with prevailing methods. This method requires synchronization of worms to the same developmental stage and age in order to achieve the most reliable data. The second option is to isolate, grow and sort the cells needed for analysis (Christensen et al., 2002; Colosimo et al., 2004). This method is very valuable if a certain tissue/cell type is under investigation, since RNA from other tissues does not dilute the signal of the target tissue. The third method utilizes powerful RNA amplification via reverse transcription, thus requiring only a very small amount of starting material; for example even the four cell C.elegans embryo is sufficient (Iscove et al., 2002; Robertson et al., 2004).
2.3.6. RNA interference in C. elegans
The research of RNA interference (RNAi, also referred to as post-transcriptional gene silencing or PTGS) started in 1990 when attempts to increase flavonoids in petunia plants with introduction of chalcone synthase (CHS) or dihydroflavonol-4- reductase (DFR) mRNA showed surprising results. The flavonoid pigmentation as well as the CHS or DFR mRNA levels were reduced significantly in many of the transgenic plants (Napoli et al., 1990; van der Krol et al., 1990). The next achievement in RNAi technology was the discovery of double-stranded RNA (dsRNA) as the most potent agent to cause the phenotype (Fire et al., 1998). In the same study, it was also discovered that a very small quantity of dsRNA was needed to produce the phenotype (Fire et al., 1998). Previously, all RNAi experiments in C. elegans were carried out by injecting the interfering substance into the body of the worm but subsequent experiments showed that also soaking in dsRNA liquid (Tabara et al., 1998) or feeding dsRNA-producing bacteria (Timmons and Fire, 1998) were effective methods to induce RNA mediated gene silencing.
The first genome-wide RNAi experiment was carried out by Kamath et al. (2003) in a study where a library of 19 000 genes was transformed into the Timmons and Fire (1998) feeding vector. Furthermore, this library was offered to other C. elegans researchers to use in their research. However, it has been observed that neurons in C.
elegans show particular resistance against RNAi that is independet of the dsRNA delivery method (Timmons et al., 2001). A study on RNA-directed RNA polymerase (RdRP) genes revealed that animals with loss of function of rrf-3 gene were more sensitive against RNAi and that deletion mutation in rrf-1 gene resulted in RNAi
