Suppression of Machado-Joseph disease pathogenesis by serotonergic signalling: the contribution of serotonin receptors

Universidade do Minho Escola de Medicina Joana Catarina Pereira de Sousa Suppression of Machado-Joseph disease pathogenesis by serotonergic signalling: the contribution of serotonin receptors Setembro de 2021 UMinho |2021 Joana Catarina Pereira de Sousa Suppression of Machado-Joseph disease pathogenesis by serotonergic signalling: the contribution of serotonin receptors Joana Catarina Pereira de Sousa Suppression of Machado-Joseph disease pathogenesis by serotonergic signalling: the contribution of serotonin receptors Tese de Doutoramento em Ciências da Saúde Trabalho efetuado sob a orientação da Doutora Andreia Cristiana Teixeira de Castro e da Doutora Luísa Alexandra Meireles Pinto Setembro de 2021 ii Direitos de Autor e Condições de Utilização do Trabalho por Terceiros Este é um trabalho académico que pode ser utilizado por terceiros desde que respeitadas as regras e boas práticas internacionalmente aceites, no que concerne aos direitos de autor e direitos conexos. Assim, o presente trabalho pode ser utilizado nos termos previstos na licença abaixo indicada. Caso o utilizador necessite de permissão para poder fazer um uso do trabalho em condições não previstas no licenciamento indicado, deverá contactar o autor, através do RepositóriUM da Universidade do Minho. Atribuição-NãoComercial-SemDerivações CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/ iii Agradecimentos / Acknowledgements À Andreia Castro por todo o apoio incondicional e as horas infindáveis de debate científico, apoio moral e amena cavaqueira sempre que era necessário. Ajudaste-me a crescer como cientista e jamais te poderei retribuir esse favor. • À Luisa Pinto. • À Patrícia Maciel, por todo o input científico e sorriso afável. • A toda a malta dos C. elegans e do “Maciel Lab ”, “ Five of a kind ” - Raquel, Rui, Célia, Sónia e Filipa – e aos doentes com MJD – com quem tive a sorte de contactar. • Aos meus grandes e eternos amigos Sun, Mariana, Juanita e Dani. • Pai e Mãe, há muito tempo atrás vocês tiveram uma escolha e eram livres de tomar qualquer um dos caminhos, sem medos nem críticas. Mas tiveram o altruísmo de me escolher a mim e proporcionar-me a vida que hoje tenho e que me faz extremamente feliz. Obrigado por tudo, esta é também uma vitória vossa. Estarei para sempre ao vosso lado! • Avô António e avó Ana, apesar de já não estarem por aqui comigo, vocês também foram a causa de tudo o que me tornei. Por isso não poderia completar esta aventura sem um pensamento caloroso e um enorme obrigado. Avó Ana, ensinaste-me a sorrir em todos os momentos da vida e essa foi, sem dúvida, a melhor de todas as lições que aprendi. Avô António, sem o teu apoio talvez o meu sonho em me tornar cientista tivesse ficado apenas por aí, obrigado por isso e por nunca teres escondido o orgulho que sentias por mim. • E como não poderia deixar de ser, um obrigado aos homens da minha vida! Toni por toda a tua paciência e por me teres sempre apoiado incondicionalmente. Nunca achaste que a minha ida para Braga fazer um doutoramento, mesmo com um filho pequeno, fosse má ideia. Estiveste lá, a fazer de pai e mãe, sempre que foi preciso. És grande! Obrigado por estares comigo. Ao meu pequeno Romeu (viste não te chamei bebé!), esta tese também é tua! Obrigado por teres sempre percebido quando a mãe não podia ir buscar-te à escola, nem levar-te aos treinos, e sempre que me atrasei para jantar. És um pequeno excepcional, nunca te esqueças disso. Por fim, ao pequeno Benjamin, tu que ainda nem viste o mundo e a mãe já puxa por ti. Desculpa as longas horas de escrita e as noites mal dormidas, mas foi por um bom motivo. Esta tese é, sem dúvida, tua também (não fosses tu parte de mim neste momento!). Mesmo antes de nasceres, obrigado por ajudares a mãe e não me roubares os neurónios todos. Finally, I would like to acknowledge all members of NERD research domain at Life and Health Sciences Research Institute (ICVS), School of medicine of the university of Minho and Bn’ML – Behavioral & Molecular Lab. 1 1 The work presented in this thesis was performed at the Life and Health Sciences Research Institute (ICVS) and at the Bn’ML – Behavioral & Molecular Lab, at the School of Medicine, University of Minho. Financial support was provided by grants from the Portuguese Foundation for Science and Technology (FCT): Doctoral Programme in Applied Health Sciences fellowship (PD/BDE/127834/2016) to J. Pereira-Sousa; and by FEDER, through the Competitiveness Internationalization Operational Programme (POCI) under the scope of the project POCI-01-0145-FEDER-031987 to A. Teixeira-Castro, supported by North Portugal Regional Operational Programme (NORTE 2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF) and by ICVS Scientific Microscopy Platform, member of the national infrastructure PPBI Portuguese Platform of Bioimaging (PPBI-POCI-01-0145FEDER022122); and by National funds, through the FCT - project UIDB/50026/2020 and UIDP/50026/2020. This work was also funded through the National Ataxia Foundation (NAF), USA. iv Statement of Integrity I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Code of Ethical Conduct of the University of Minho. “Supressão da patogénese da doença de Machado-Joseph através da modulação da sinalização serotoninérgica: contribuição dos receptores de serotonina” v Resumo A doença de Machado-Joseph (DMJ) é uma doença debilitante caracterizada pela perda progressiva do equilíbrio e da coordenação motora. A nível molecular, a DMJ está associada à expansão da repetição de Citosina-Adenina-Guanina (CAG) no gene da ataxina-3. Isto traduz-se num segmento expandido de poliglutaminas na proteína ataxina-3 (ATXN3), que é propenso à auto-associação, sendo a agregação desta proteína uma característica da doença. Apesar dos vários esforços efectuados, não existe cura para esta doença fatal. A modulação da sinalização serotoninérgica foi anteriormente descrita como potencial estratégia terapêutica para a DMJ. O tratamento com citalopram (CIT), um inibidor seletivo da recaptação de serotonina, suprimiu a patogénese da DMJ em vários modelos animais, sendo o efeito do CIT dependente do transportador de serotonina, o seu alvo molecular, com provável contribuição dos receptores serotoninérgicos 5-HT1A/SER-4 e 5-HT2/SER-1. No entanto, o papel de cada receptor serotoninérgico na supressão da proteotoxicidade da ATXN3 mutante não foi ainda elucidado. Aqui, com o objetivo de determinar a contribuição de cada receptor de serotonina (5-HTR) na supressão da proteotoxicidade da ATXN3 usámos um modelo de C. elegans que recapitula as principais características da DMJ, seguindo abordagens químico-genéticas, farmacológicas e bioquímicas, bem como técnicas de microscopia in vivo . Inicialmente, explorámos a contribuição do receptor 5-HT1A usando befiradol, um agonista potente e altamente específico destes receptores. A administração crónica e aguda deste composto melhorou a função motora dos mutantes, sendo o ortólogo do receptor 5-HT1A no nematode, o SER-4, necessário para a sua ação. De forma a aprofundar o conhecimento sobre o impacto do tratamento com befiradol na proteotoxicidade da ATXN3, optimizámos ensaios de retardação em filtro e fracionamento bioquímico para avaliar as espécies agregadas da ATXN3, demonstrando-se que ambas as modalidades de tratamento afectaram a solubilidade desta proteína. De realçar que, usando químicogenética e medições de atividade neuronal, foi-nos possível definir os autoe hetero-receptores 5HT1A/SER-4 como essenciais para o efeito terapêutico do befiradol e propor o receptor 5-HT1A como novo alvo terapêutico para a DMJ. Além disso, identificámos uma contribuição parcial dos receptores 5HT6/SER-5 e 5-HT7/SER-7 para a ação do CIT. Surpreendentemente, o antagonismo dos receptores 5HT2/SER-1, 5-HT6/SER-5 e 5-HT7/SER-7 mostrou-se benéfico para a função motora de animais mutantes, o que é consistente com dados de perfil de expressão de RNA e farmacoterapia multimodal obtidos no nosso laboratório. Embora promissores, são necessários mais estudos para reforçar estes achados e elucidar as vias específicas pelas quais cada 5-HTR individual afecta a proteotoxicidade da ATXN3. Palavras Chave: SCA3, serotonina, receptores serotoninérgicos, C. elegans , ISRS “Suppression of Machado-Joseph disease pathogenesis by serotonergic signalling: the contribution of serotonin receptors” vi Abstract Machado-Joseph disease (MJD) is a debilitating disease mainly characterized by a progressive loss of balance and motor coordination. At the molecular level, this autosomal dominant disease is associated with an expansion of the Cytosine-Adenine-Guanine (CAG) repeat in the ataxin-3 gene. This translates into a pathologic polyglutamine tract in the ataxin-3 protein (ATXN3) that is prone to selfassociate, aggregation of ATXN3 being a hallmark of the disease. Although several efforts have been made, no cure is yet available for this fatal disorder. Serotonergic signalling modulation, by selective serotonin reuptake inhibitors (SSRIs), was previously described as a potential therapeutical strategy for MJD. Treatment with citalopram (CIT), a SSRI used in the clinics to treat depression, suppressed MJD pathogenesis in a disease-modifying manner, in several animal models. CIT’s effect was dependent on its molecular target, the serotonin transporter SERT, likely with the additional need of the serotonin receptors 5-HT1AR/SER-4 and 2-HT2R/SER-1. However, the exact role of each serotonin receptor (5-HTR) in the suppression of mutant ATXN3 proteotoxicity remains to be fully elucidated. Here, we aimed at determining the contribution of each of the 5-HTRs in the suppression of ATXN3 proteotoxicity using a C. elegans model that recapitulates major hallmarks of MJD, by employing chemical genetics, pharmacologicaland biochemicalapproaches, as well as in vivo dynamic imaging techniques. First, we explored the contribution of the 5-HT1AR using a highly specific and potent agonist of these receptors, befiradol. Chronic and acute administration of befiradol rescued the motor function of mutant animals, the 5-HT1AR orthologue in the nematode, SER-4, being required for its action. To further explore the impact of befiradol treatment in ATXN3 proteotoxicity, we optimised filter retardation and biochemical fractionation assays to assess mutant ATXN3 aggregation states, and we showed that both treatment modalities impacted ATXN3 solubility. Importantly, using chemical genetics and single neuron neuronal activity measurements, we defined 5-HT1AR/SER-4 autoand heteroreceptors’ function to be essential to befiradol therapeutic outcome, and the 5-HT1AR as a novel therapeutic target for MJD. Additionally, we found a partial contribution of 5-HT6R/SER-5 and 5-HT7R/SER-7 for CIT action and, surprisingly, antagonism of 5-HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7 showed to be beneficial for motor function of mutant animals, which is consistent with RNA-expression profiling and multimodal pharmacotherapy results obtained in our laboratory. Although promising, further studies are needed to support these findings and to elucidate the specific pathways by which each individual serotonin receptor impacts ATXN3 proteotoxicity. Keywords: SCA3, serotonin, serotonin receptors, C. elegans , SSRI vii Table of Contents Direitos de Autor e Condições de Utilização do Trabalho por Terceiros ................................................... ii Agradecimentos / Acknowledgements ................................................................................................... iii Statement of Integrity ............................................................................................................................. iv Resumo................................................................................................................................................... v Abstract.................................................................................................................................................. vi Table of Contents .................................................................................................................................. vii List of Abbreviations ............................................................................................................................... xi List of Figures ....................................................................................................................................... xiii List of Tables ......................................................................................................................................... xv Thesis Aims and Layout ........................................................................................................................ xvi Chapter 1. General Introduction ............................................................................................................. 1 1. Machado-Joseph disease – an introduction ........................................................................................... 2 1.1. A polyglutamine disorder............................................................................................................. 2 1.2. Disruption of protein homeostasis as a mechanism of MJD pathogenesis .................................... 3 1.3. MJD therapeutics ........................................................................................................................ 5 2. Serotonin modulation and the impact on neurodegenerative diseases .................................................... 6 3. Serotonin .............................................................................................................................................. 7 3.1. Serotonin synthesis and metabolism ........................................................................................... 7 3.2. Serotonin function in the brain .................................................................................................... 9 3.3. Serotonin receptors and signalling ............................................................................................... 9 4. Caenorhabditis elegans – a small nematode with gigantic possibilities ................................................. 14 4.1. C. elegans as a model system to study neurodegenerative diseases........................................... 15 4.2. The C. elegans nervous system ................................................................................................. 15 4.3. Serotonergic signalling in C. elegans ......................................................................................... 16 5. Relevance of the study ........................................................................................................................ 20 References.................................................................................................................................................... 21 Chapter 2. Filter retardation and biochemical fractionation assays for detecting and quantifying neuronal polyglutamine aggregates using C. elegans protein lysates: protocol development attempts ................ 36 Abstract ........................................................................................................................................................ 38 1. Introduction ........................................................................................................................................ 39 2. Material and methods ......................................................................................................................... 41 2.1. Caenorhabditis elegans strains and maintenance ...................................................................... 41 2.2. Drug preparation ...................................................................................................................... 41 xiv Figure 5. 5-HT1A autoand heteroreceptors’ function contribute to befiradol treatment efficacy of ATXN3 mutant animals. ................................................................................................................................... 79 Figure 6. Hypothetical model of befiradol action in a C. elegans model of ATXN3 proteotoxicity. ......... 80 Supplementary figure 1. Befiradol and tandospirone toxicity evaluated using the food clearance assay. ............................................................................................................................................................ 86 Supplementary figure 2. Total steady-state levels of ATXN3 protein. .................................................... 87 Supplementary figure 3. Filter trap membranes. .................................................................................. 88 Supplementary figure 4. Whole animal mRNA levels. ........................................................................... 89 Supplementary figure 5. 17-DMAG and CCI779 actions on mutant ATXN3-mediated motor dysfunction decreases in the absence of SER-5.. .................................................................................................... 90 Supplementary figure 6. Befiradol and tandospirone actions on mutant ATXN3-mediated motor dysfunction are independent of dopaminergic receptors.. .................................................................... 91 Supplementary figure 7. Ca2+ dynamics in HSN neuron upon befiradol chronic and acute treatment. .. 92 Chapter 4. Addressing the role of serotonin receptors in the suppression of mutant ataxin-3 proteotoxicity Figure 1. Schematic representation of the impact of tool compounds on a receptor’s response. ....... 102 Figure 2. The effect of citalopram in the amelioration of motor dysfunction of AT3q130 animals is partially dependent on the G protein-coupled receptors SER-5 and SER-7 in C. elegans . ................................. 107 Figure 3. Chronic administration of antagonists, but not agonists of 5-HT2R/SER-1, 5-HT6R/SER-5 and 5HT7R/SER-7, ameliorates motor dysfunction of ATXN3 proteotoxic model. ......................................... 110 Figure 4. Chronic administration of 5-HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7 inverse agonists fail to impact motor dysfunction on ATXN3 mutant animals. ................................................................... 111 Figure 5. Acute administration of LY 266097, a 5-HT2R / SER-1 antagonist, ameliorated mutant ATXN3mediated motor dysfunction. .............................................................................................................. 112 Chapter 5. General discussion and conclusions Figure 1. Hypothetical model of the beneficial effect induced by the remodelling of serotonergic signalling in the suppression of mutant ATXN3 proteotoxicity. ........................................................................... 134 xv List of Tables Chapter 1. General introduction Table 1. Non-canonical pathways elicited by the different 5-HT receptors subtypes. ............................ 10 Chapter 2. Filter retardation and biochemical fractionation assays for detecting and quantifying neuronal polyglutamine aggregates using C. elegans protein lysates: protocol development attempts Table 1. C. elegans strains used in this study. ..................................................................................... 41 Supplementary table 1. Statistical report. ............................................................................................. 68 Chapter 3. Identification of the 5-HT1A serotonin receptor as a novel therapeutic target in a C. elegans model of Machado-Joseph disease Supplementary table 1. Strains used in the study. ............................................................................... 93 Supplementary table 2. List of primers used in the generation of double mutant strains. .................... 94 Supplementary table 3. List of primers used in reverse-transcriptase quantitative PCR. ....................... 95 Supplementary table 4. Statistical report. ............................................................................................. 96 Chapter 4. Addressing the role of serotonin receptors in the suppression of mutant ataxin-3 proteotoxicity Table 1. C. elegans strains used in this study. ................................................................................... 103 Table 2. Tool compounds used in this study and mammalian target receptors. ................................. 104 Table 3. Most efficient concentration determined in dose-response curves, of each tool compound, used in the experiments summarized in figure 3. ....................................................................................... 110 Supplementary table 1. Statistical report. ........................................................................................... 125 xvi Thesis Aims and Layout The main aim of this thesis was to address the contribution of each of the serotonin receptors (5HTRs) in the suppression of mutant ATXN3 proteotoxicity in C. elegans . For that, we have established the following specific aims: - To implement biochemical techniques to analyse mutant ATXN3 aggregation states in C. elegans ; - To determine the role of 5-HT1A autoand hetero-receptors in the suppression of mutant ATXN3 aggregation and proteotoxicity; - To investigate the potential contribution of other 5-HTRs in citalopram-mediated suppression of Machado-Joseph disease (MJD) pathogenesis; - To establish potential novel therapeutic targets for MJD. The present dissertation is organized in 5 chapters. Chapter 1 is the general introduction, the chapters comprising experimental work are presented in Chapter 2 to 4 (in the format of research articles) and Chapter 5 is the general discussion of the work. The research work presented in Chapter 3 was published in Neurobiology of Disease journal. Before addressing the aims of this work, we provide an overview of the literature about MJD, describing the pathological features of the disease, the main pathogenic mechanism(s) and the potential therapeutic strategies discovered in the last decades, focusing mainly on the modulation of the serotonergic neurotransmission. Moreover, we described the serotonergic system in mammals, giving special attention to serotonin receptors, and we established a parallel with the organization of the serotonergic system of the nematode C. elegans – Chapter 1. To address the first aim of this thesis and assess the impact of serotonergic system modulation in MJD, we used two major readouts and hallmarks of the disease: mutant ATXN3-mediated motor dysfunction and protein aggregation, using a C. elegans model of ATXN3 proteotoxicity. We optimised two biochemical techniques for the assessment of ATXN3 aggregation states, the filter retardation and biochemical fractionation assays. These two techniques explore size and solubility of the aggregates. By introducing internal loading controls, these methods became semi-quantitative, allowing for statistical analysis. We therefore successfully established and improved the laboratory protocol for assessment of ATXN3 aggregation in C. elegans . These two methods can complement the information given by in vivo confocal dynamic imaging, used to assess aggregation in this model in previous studies – Chapter 2. xvii Since, previously, Teixeira-Castro and colleagues reported a dependency on the nematode serotonin receptor SER-4, the orthologue of the 5-HT1A mammalian receptor, for the effect of CIT on mutant ATXN3-induced motor dysfunction, we first explored the contribution of this receptor to the suppression of MJD pathogenesis using a highly potent and selective 5-HT1A receptor agonist, befiradol – Chapter 3. Next, we determined the dependency on other serotonin receptors for CIT’s beneficial effect on ATXN3-induced motor impairment and we started to explore the contribution of the relevant receptors to the suppression of MJD pathogenesis using tool compounds with different intrinsic binding properties to each receptor – Chapter 4. We concluded that a complex remodelling of serotonin receptors activity may underlie CIT’s action in MJD. Finally, we integrate the results and describe their relevance in the field, as well as describe future perspectives – Chapter 5. 1 Chapter 1. General Introduction Chapter 1. 2 1. Machado-Joseph disease – an introduction 1.1. A polyglutamine disorder Machado-Joseph disease (MJD), also known as Spinocerebellar ataxia (SCA) type 3 [1], is an inherited neurodegenerative disease with autosomal dominant transmission caused by a genetic alteration in the ataxin-3 gene ( ATXN3 ) [2]. Clinically, it is characterized by i) progressive cerebellar ataxia, detectable as motor incoordination that affects balance, gait, and speech; ii) peripheral neuropathy; iii) oculomotor abnormalities; iv) facial and lingual fasciculations; v) extrapyramidal signs, including dystonia, rigidity and/or bradykinesia; vi) weight loss and vii) sleep disorders [3; 4; 5; 6]. Among all dominant ataxias worldwide, MJD is the most prevalent, presenting in some countries, like Portugal and Brazil, 58% to 92% of relative frequency within this group of diseases [3; 7; 8]. MJD is one of the polyglutamine (polyQ) disorders, characterized by expansion of a tandem repetition of the Cytosine-Adenine-Guanine triplet (CAG) that encodes the amino acid glutamine (Q) [9]. Within the human proteome polyQ tracts are found in 66 proteins, mainly associated with two biological processes: i) DNA dependent regulation of transcription and ii) neurogenesis [10]. PolyQ disorders include SCA types 1, 2, 6, 7 and 17 and MJD, Huntington's disease, dentatorubral pallidoluysian atrophy (DRPLA) and spinal bulbar muscular atrophy X-linked type 1 (SMAX1/SBMA) [9; 11]. The nine proteins that cause polyQ disorders differ from the other polyQ tract-containing proteins by displaying a higher genomic instability, which causes the stretching of the tandem CAG repeats to a pathological size [12]. The ATXN3 gene was mapped on chromosome 14q32.1 (initially designated as MJD1 [13]), and shown to possess a range of CAG units encoding for 10 to 44 glutamines in normal subjects, whereas in MJD patients it presents a polyQ tract containing 61 to 87 residues [14; 15; 16]. Intermediate polyQ lengths are not common but associated with incomplete phenotype penetrance [15]. Furthermore, a negative correlation between expansion size and age at disease onset was found [14]. At the neuropathological level, the expanded polyQ tract makes ataxin-3 protein (ATXN3) prone to aggregation, a hallmark of MJD [17]. ATXN3 is constituted by a N-terminus Josephin domain, two or three ubiquitin-interacting motifs (UIMs) and the polyQ tract, closest to the C-terminus [18; 19]. The Josephin domain makes ATXN3 prone to self-associate, inducing the formation of oligomers. Oligomers of ATXN3 with normal polyQ tracts are SDS-soluble whereas oligomers of expanded ATXN3, with Chapter 1. 3 pathological polyQ tracts, undergo a second step in the aggregation process, resulting in the formation of SDS-insoluble amyloid-like aggregates [20; 21; 22]. Post-mortem , MJD patients’ brain typically present atrophy of the cerebellum and brainstem, with a remarkable neuronal depletion at the cerebellar dentate nuclei, pallidum, substantia nigra, thalamus, pontine nuclei, red nuclei, various cranial nerve nuclei, anterior horn cells of the spinal cord, and Clarke’s columns [23; 24; 25; 26; 27]. Also, volumetric analyses using MRI have demonstrated atrophy of the cerebellum, basal ganglia, brainstem and pallidum and suggested damage of the cerebellar– thalamocortical and basal ganglia–thalamocortical pathways [28; 29]. The ATXN3 aggregated species are often found in affected areas of MJD patients’ brain, as the pons, substantia nigra, globus pallidus, dorsal medulla and dentate nucleus [17; 30]. Nonetheless, aggregates can also be found in brain areas that are typically spared by the disease [30; 31; 32]. In contrast, brain regions in which mutant ATXN3 inclusions are never detected can also undergo cellular death (e.g. the thalamus) [33]. These observations underlie the controversial role of aggregation in MJD, as some researchers consider that aggregation is a cause of the pathologic process, whereas others see it as inert or even as a cellular protective mechanism. One possible explanation for these events could be that in some brain regions the appearance of mutant ATXN3 inclusions precedes toxicity and neuronal cell death, whereas in others, the bigger aggregates are never formed as toxicity overrides aggregation. That could be the case of thalamus, where although with degeneration, the ATXN3 aggregates were not found [33]. In addition, as aggregation can be a controlled-process by quality-control mechanisms of the cell, these large aggregates can function in a way to avoid toxicity and maintained it subthreshold, preventing cell death. This can explain the ATXN3 inclusions in spared brain areas. 1.2. Disruption of protein homeostasis as a mechanism of MJD pathogenesis The exact mechanism of MJD pathogenesis is still under debate, however it is generally accepted that a gain-of-function of toxic expanded ATXN3 contributes greatly to MJD pathology. Also, it has been proposed that disruption of the protein homeostasis can be an initiating factor of pathogenesis (reviewed in [34] and [35]). Briefly, the expansion of the polyQ tract leads to the accumulation of the abnormal ATXN3 protein in amorphous aggregates or amyloid fibres. These changes in ATXN3 may lead to a partial loss of its deubiquitylase activity, impact aggresome formation, autophagy, the endoplasmic reticulum–associated protein degradation (ERAD) and the degradation of the proteins by the proteasome, all of which cellular Chapter 1. 4 functions ATXN3 is thought to contribute for. This imbalance will impact protein homeostasis, with important consequences for cellular physiology, affecting the nucleus, mitochondria, endoplasmic reticulum and cellular communication [34]. Protein homeostasis or proteostasis is ensured by a protein quality-control (PQC) network composed of molecular chaperones, proteolytic systems and their regulators [36]. To ensure the correct proteostasis in the cell, this network i) assists the folding of newly synthetized proteins; ii) ensures properly folded proteins are generated at the correct time, cellular location, and with expected stoichiometry to allow assembly of oligomeric protein complexes and iii) prevents protein misfolding and aggregation. In case of aggregate formation, the PQC ensures the refolding or the removal of the aggregated proteins by orchestrating folding, disaggregation, and degradation, by autophagy or proteasome-mediated processes. All this surveillance, performed by the PQC network, avoids the accumulation of dysfunctional proteins and of protein aggregates. However, some insults, such as those that accumulate with aging and genetic mutations that underlie disease, can imbalance these cellular maintenance mechanisms, impairing proteostasis, and a toxic threshold can be achieved [37; 38; 39]. Therefore, any mutations within the genome which increase aggregation propensity, constitute an endogenous stress to neuronal cells and challenges their proteostasis. The deubiquitinating enzyme activity of ATXN3 [40] consists in the removal of ubiquitin from polyubiquitinated proteins, thus modulating proteasomal and/or autophagic degradation system ([18; 41; 42; 43; 44; 45] reviewed in [45]). By editing polyubiquitin chains, ATXN3 can facilitate the degradation of some proteins or protect others from degradation [46; 47; 48; 49]. ATXN3 also was reported to be involved in the formation of aggresome [47; 50], a structure to which misfolded proteins are transported when the chaperone mediated-refolding and the ubiquitin proteasome systems are overloaded [51]. These aggresomes facilitate the clearance of misfolded proteins by autophagy. Furthermore, ATXN3 is involved in ERAD, being described that expanded ATXN3 binds excessively to VCP, a key protein responsible for extracting ERAD substrates from the endoplasmic reticulum (ER), decreasing ER retrotranslocation and degradation of ERAD substrates [52; 53]. Due to direct and indirect impact on cellular function, as explained above, ATXN3 aggregation causes an imbalance in proteostasis and a cellular burden leading to cell death and degeneration. Therefore, several compounds that modulate pathways of cellular maintenance have been tested as a potential therapy for MJD. Chapter 1. 5 1.3. MJD therapeutics Currently, there is no effective disease-modifying treatment for this devastating disease. However, several treatments are being investigated and involve i) the direct targeting of ATXN3 gene expression, ii) reducing ATXN3 aggregation, iii) increasing degradation, iv) reducing the nuclear import of ATXN3, v) inhibiting the generation of toxic fragments of the mutant protein, regulating vi) abnormal protein-protein interactions, vii) mitochondrial function, energy availability and oxidative stress or viii) correcting abnormal neuronal firing and neurotransmission (reviewed in [34; 54; 55]). As referred above, one of the possible treatments explored for MJD relates with neurotransmission modulators. In an unbiased drug screen, Teixeira-Castro and colleagues found that citalopram (CIT), a selective serotonin reuptake inhibitor (SSRI), successfully suppressed MJD pathogenesis in animal models [56]. CIT chronic treatment was able to ameliorate motor symptoms of MJD C. elegans and mouse models and decrease ATXN3 aggregation, the effect being dependent on the serotonin transporter (SERT) [56; 57]. The blocking of SERT action resulting from SSRI treatment, likely increases serotonin availability at the synaptic cleft triggering the remodelling of the serotonin receptors [58; 59]. Interestingly, in an ATXN3-proteotoxicity C. elegans model, CIT’s action was dependent not only on MOD-5, the SERT orthologue in the nematode, but also on SER-4 and SER-1, the orthologs of 5-HT1A and 5-HT2 receptors, respectively [56]. Although the use of fluoxetine, a less specific SSRI than CIT [58], failed to improve motor function of MJD patients in a small and short term clinical trial [60], other studies with serotonin 5-HT1A receptor partial agonists showed efficacy in ameliorating cerebellar ataxia symptomatology [61; 62]. In these open label clinical trials, two 5-HT1AR partial agonists, buspirone and tandospirone, were used. Tandospirone four-week treatment improved the ataxia rating scale (ARS) and total length travelled (TLT) scores in MJD and SCA6 patients [63]. Furthermore, in another study, MJD patients showed improved gait, posture, kinetic and total ARS [64] but after the withdrawal of the drug the symptomatology returned, showing a non-disease-modifying but rather symptomatic effect of tandospirone treatment in MJD pathogenesis [64]. Treatment with the other 5-HT1AR partial agonist, buspirone, in patients with cerebellar cortical atrophy and olivopontocerebellar atrophy, improved clinical and self-assessment test scores [65], however the effect showed to be only partial and not major [66]. In summary, the serotonergic system is promising and should be further investigated as a potential therapeutic target for MJD. However, further studies are needed to better understand the mode of action of these drugs in the context of this and other neurodegenerative disorders (NDs). Chapter 1. 6 2. Serotonin modulation and the impact on neurodegenerative diseases Serotonergic system modulation was also shown to be beneficial to other NDs [67; 68; 69; 70; 71]. The common ground among these diseases may be encountered at the pathological level. Despite distinct protein abnormalities are associated with each ND, the formation of insoluble protein aggregates constitutes a common feature, these diseases being collectively often defined as “proteinopathies” [72]. Although none of the known NDs have serotonergic system dysfunction as a sole pathological basis, some reports showed serotonergic degeneration or found evidence of altered serotonin levels as a prognostic factor. Azmitia and Nixon reported a serotonergic axon dystrophy in brains of Parkinson's disease (PD), frontal lobe dementia (FLD) and Diffuse Lewy-Body dementia (DLBD) patients, however the study presented a low sample size and incomplete clinical characterization, as highlighted by the authors [73]. Altered levels of serotonin were also found in the spinal cords of amyotrophic lateral sclerosis (ALS) patients [74], with the serotonin levels in platelets being correlated with survival of these patients [75]. On the other hand, a study in multiple sclerosis (MS) showed that patients treated with fluoxetine show a tendency towards a reduction in the formation of new lesions [71]. Despite these studies pointing towards an involvement of serotonin in these diseases, they are few and the proof of evidence is tenuous. However, a large body of evidence of serotonergic involvement in disease progression and potential for therapy emerged for PD and Alzheimer’s disease (AD). In PD, serotonergic denervation was found in all stages of disease development, albeit failing to correlate with disease progression or patient’s disability [76]. However, the involvement of serotonergic system in motor symptoms, as tremors, and non-motor symptoms in PD, as depression, weight loss, fatigue, and visual hallucinations was reported [77; 78; 79]. The most well studied contribution of serotonergic system in PD is in L-Dopa induced dyskinesia (LID). LID is a movement disorder often seen in PD patients due to the prolonged treatment with L-DOPA [80]. Serotonergic neurons located probably in the striatum, have all the enzymatic machinery needed to convert L-DOPA into dopamine. Therefore, in presence of L-DOPA, serotonergic neurons will store and release dopamine as a “false neurotransmitter”, potentiating the LID phenotype; they are unable to regulate dopamine release due to the lack of regulatory feedback for this neurotransmitter [79]. In agreement, the administration of serotonergic autoreceptor agonists improves LID by the suppression of serotonin release, and consequently, of the “false neurotransmitter” release [81; 82; 83; 84]. Serotonin also seems to play a role in AD. Decreased levels of serotonin was described in the patients’ brains [85; 86] with serotonergic denervation occurring in the raphe nucleus [87; 88; 89; 90; Chapter 1. 13 Although some functions can be attributed to 5-HTRs, as described above, the mechanism by which each receptor acts is mostly unknown. However, the probable answer in the most cases is by the modulation of other neurotransmitter systems. 5-HT1AR modulates several neurotransmitters as dopamine, acetylcholine and is a regulator of the serotonergic system itself (Figure 2). 5-HT1BR also mediates the inhibition of GABA, cholinergic and glutamatergic neurotransmission. The 5-HT2R family plays a role in the modulation of GABA, glutamatergic, and dopaminergic transmission (Figure 2). Lastly, it was shown that 5-HT6R impacts the extracellular levels of acetylcholine, glutamate, GABA, dopamine and norepinephrine (reviewed in [112]). Therefore, serotonin action in the brain can be difficult to predict, due to the innumerous ways it impacts brain function (by projecting to several areas, the high number of receptors, the multiple signalling pathways – canonical and non-canonical – and by the modulation of other neurotransmitters). Furthermore, it is established that selective agonists of 5-HTRs are prone to bias agonism, which means that the signalling cascade activated by the engaging of drug-receptor is compound specific, and that different agonists can be more prone to activate one cellular response than another [158; 159]. This property can be extremely useful for the design of drugs with less side effects or more specific to treat specific diseases. One more layer of complexity comes from the concepts of constitutive activity (reviewed in [160]) and homoand heterodimerization of these receptors (reviewed in [161]). The constitutive activity corresponds to the ability of the receptors to signal, and activate a specific signalling cascade, even in the absence of a ligand. Most of the studies that point to spontaneous activity of 5-HTRs were performed in vitro , however some compelling in vivo evidence was found for at least the 5-HT2CR (reviewed in [160]). Homoand heterodimerization of 5-HTRs were also described, and 5-HTRs may dimerize not only with other serotonin receptor types but also with glutamatergic and dopaminergic receptors. The most well described heterodimers are 5-HT2AR-5-HT2CR, 5-HT1AR-5-HT7R, 5-HT2AR-D2, 5-HT2AR-mGlu2. These associations modulate the signalling pathways of the monomeric receptors. As an example, 5-HT1AR-5HT7R heterodimerization increases 5-HT1AR internalization and decreases Gi canonical signalling but increases 5-HT1AR-mediated ERK signalling (reviewed in [161]). The complexity of signalling modulation by these receptors makes the serotonergic system extremely robust and versatile. Chapter 1. 14 4. Caenorhabditis elegans – a small nematode with gigantic possibilities The complexity of the serotonergic system in mammals increases the level of difficulty in determining the contribution of a specific receptor to a given biological process, not being easy also to predict the mode of action of some drugs in vivo . Therefore, simpler model systems can allow a deeper understanding of drug action in vivo and in a pathological situation. In this regard, Caenorhabditis elegans ( C. elegans ), a small nematode, has long proved to be a potent platform for the study of NDs and a suitable model for drug testing in a high-throughput manner. C. elegans, as a model organism, was first described by Sydney Brenner. In 1974, Brenner appreciated that to study the functionality of the nervous system “ what was needed was an experimental organism which was suitable for genetical study and in which one could determine the complete structure of the nervous system”. “Drosophila,” – he wrote – “with about 105 neurons, is much too large, and, looking for a simpler organism, my choice eventually settled on the small nematode, Caenorhabditis elegans ” [162]. Indeed, the C. elegans nervous system in adulthood is composed of 302 neurons with a completely known lineage, structure and connectivity [163; 164; 165]. Despite this apparent simplicity, a single sensory neuron can express up to 14 types of neurotransmitter receptors and 10 neuropeptides [166]. In 1998, the complete sequence of the C. elegans genome was revealed [167]; importantly, 83% of this nematode genome presented a homologous sequence in humans, corresponding to 15344 sequences [168], half of these sequences (7943) being orthologs of human genes [169]. C. elegans is a transparent, free-living nematode with about 1 mm in length, with a short life cycle of about 3.5 days from hatching to adulthood (at 20ºC). Being a self-fertilizing hermaphrodite, it produces a large number of genetically identical progeny, being for that reason a good choice for high-throughput screening [170]. It is also easily maintained in laboratory conditions, on a diet of Escherichia coli bacteria. The transparency of this animal allows the study of several cellular processes in vivo , by tagging proteins with fluorescence tags. Furthermore, it is extremely amenable to genetic manipulation. RNA interference (RNAi) [171] and clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 technology [172] are well established in the C. elegans field, allowing easy genetic manipulation. The sum of all these features makes C. elegans an ideal model organism to study human diseases, and in particular NDs with protein aggregation as molecular basis. Chapter 1. 15 4.1. C. elegans as a model system to study neurodegenerative diseases Nowadays several models of NDs exist in C. elegans , namely of polyQ diseases, including HD, and MJD, as well as AD, Tauopathies, ALS, PD and prion diseases. All the models generated for each disease can, however, vary in specific tissue-driven transgene expression, length of transgene (full protein or truncated forms), mutation and, in the case of polyQ diseases, the expansion size also often varies (reviewed in [173; 174]). For MJD, a C. elegans ATXN3 proteotoxicity model (AT3q130) was generated by Teixeira-Castro and colleagues [175], that addresses construct, face and predicted validity as a model of disease. This strain is a transgenic model in which the human ATXN3 gene containing an expansion of 130 glutamines is expressed in the nervous system of the animals, having therefore the same genetic cause of the human disease (construct validity). This transgenic insertion in the C. elegans genome led to the development of phenotypes that are analogous to the human disease (face validity), namely motor defects and aggregation of the mutant ATXN3 protein. Furthermore, the pathogenicity of mutant ATXN3 correlates positively with the size of the polyQ expansion, animal age, degree of motor dysfunction and aggregation. Moreover, improvement of these phenotypes with specific drug treatments [56; 175; 176; 177] showed to be effective in vertebrate models of the disease [56; 57; 178; 179], and in human patients [180] (predictive validity). Despite the simplicity of this model organism, the conservation of basic cellular mechanisms, including those related to nervous system function, associated to a validated MJD model, makes it a powerful platform for drug testing and target validation. 4.2. The C. elegans nervous system C. elegans do not have a proper “brain”, their nervous system being organized in ganglia, located in the head and tail, and the ventral nerve cord, that would correspond to the mammalian spinal cord. Its 302 neurons are divided in 118 distinct classes according to their topology and synaptic connections being possible to split them into four functional categories: motor neurons, sensory neurons, interneurons and polymodal neurons (reviewed in [163; 181]). One third of the total number of neurons in the nematode are motor neurons. These cells make synaptic contacts with muscles and therefore control the locomotor behaviour repertoire, namely crawling, swimming and motility of alimentary and reproductive systems. Motor neurons encompass 8 Chapter 1. 16 classes (AS, DA, DB, DD, VA, VB, VC and VD). A and B-types (DA, DB, VA and VB) are cholinergic and excitatory, D-type (DD and VD) neurons produce γ-aminobutyric acid (GABA), an inhibitory neurotransmitter, and C-type (VC) neurons synthetize several neurotransmitters, such as acetylcholine and serotonin. Sensory neurons are divided into sensillar neurons, that compose the 24 sensillar organs in the nematode, and several isolated sensory neurons. This category of neurons is responsible for the perception of the surrounding environment. Therefore, sensory neurons, as sensory specialized cells, can perceive specific cues of the surrounding, allowing mechanosensation, nociception, chemosensation, thermosensation, as well as light and oxygen sensation. Interneurons constitute another group of neurons in C. elegans . They function as information processors, receiving input from one or more neurons, integrating the information, and transmitting it to the effector neuron. Lastly, polymodal neurons can perform more than one function, combining motor and sensory functions, interneuron and sensory or interneuron and motor functions. As an example, the NSM neuron is simultaneously a neurosecretory and motor neuron, with a sensory function in proprioception. All these neuronal classes, as in mammals, communicate by chemical or electrical synapses, or neuromuscular junctions. Hence, C. elegans synapses can be explored and described by the type of neurotransmitter released by the pre-synaptic neuron. Chemical neurotransmission is accomplished by glutamatergic [182], cholinergic [183], and GABAergic [184] signalling, neuropeptides [185] and by monoaminergic [186] signalling that comprises octopamine, tyramine, dopamine and serotonin. 4.3. Serotonergic signalling in C. elegans Despite its compact nervous system, C. elegans is capable of performing both basic and relatively complex behaviours. Locomotion, foraging, feeding, and defecation are the basic behaviours for animal survival, but they can also discriminate and avoid or prefer some chemicals, odorants, temperatures, and food sources (reviewed in [187]). Complex behaviours can be accomplished by this nematode, such as social feeding behaviour [188], being most of these modulated by associative learning and memory [189]. In C. elegans , serotonin-dependent pathways have been described in the modulation of basic behaviours as egg-laying, feeding, locomotion, and olfactory learning [190; 191; 192; 193; 194; 195]. Chapter 1. 17 Egg-laying is a well-studied process that involves alteration in the locomotion pattern of the C. elegans and is divided in two phases, the active state, where the eggs are laid in clusters, and the inactive state, in which eggs are retained. The transition between the two states was shown to be modulated by 5-HT [196; 197] and more recently by the co-released neuropeptide NLP-3, in an apparently redundant manner [191]. Feeding, as for all living animals, is an indispensable process for C. elegans survival. Feeding is accomplished by two processes that take place in the pharynx: pharyngeal pumping and isthmus peristalsis. The action of the serotonergic system in the pharynx was shown to be crucial for both processes, probably through the modulation of the cholinergic system [198; 199; 200]. Locomotion behaviour is not only modulated by 5-HT in the context of egg-laying but also in the context of responsiveness to food. It was shown that the so-called “enhanced slowing response”, a behaviour in which food-deprived animals slow dramatically their locomotion when they encounter a food source, is dependent on serotonergic signalling. In this manner, the “enhanced slowing response” allows starved animals not to leave a favourable environment [193; 201]. Furthermore, 5-HT is essential for the learning and discrimination of good and dangerous food sources [189; 194], being involved in the induction of dwelling states of the nematode [202]. Therefore, 5-HT impacts several dimensions of feeding behaviour. Besides neurochemical action, 5-HT plays a role in neuro-humoral signalling in C. elegans . Upon a stress cue, that can be environmental or intrinsic, cellular responses to suppress deregulation of protein homeostasis (namely proteostasis) are triggered and communicated through the different tissues (cell non-autonomous communication), in a process that requires 5-HT release from neurons [203; 204; 205]. It was described that 5-HT can induce the heat-shock response [206] and the unfolded protein response of the mitochondria (UPRmt) [207; 208] in distal tissues, being protective against protein aggregation (Figure 3). Chapter 1. 18 Figure 3. Representation of C. elegans anatomy and behaviours controlled by the serotonergic system. The serotonergic system is involved in the regulation of behaviours represented in the yellow boxes. In the feeding process, upon encountering a loan of bacteria, starved C. elegans animals decrease drastically their exploratory movement, by the activation of the “enhanced slowing response” [193; 201]. Also, by the perception of the food in the environment, animals increase pharyngeal pumping [198; 199; 200] and dwelling behaviour, allowing a locally exploration of the food-rich environment increasing the turns and slowing movement [202]. In egg-laying, 5-HT released from the HSN neuron stimulates vulval muscle excitability, triggering the transition from an inactive to an active state, thus promoting egg release [196; 197]. Also, 5-HT is involved in defecation, since exogenous serotonin inhibits expulsion muscle contraction. Upon stress cues, distal tissue communication is ensured by 5-HT release, which causes the activation, in the target tissue, of stress responses and quality control mechanisms, such as the heat-shock response and the mitochondrial unfolded protein response (UPRmt) [203; 204; 205; 206]. The main anatomical features of the nematode C. elegans are represented. The digestive tract of the animal is constituted by the pharynx, intestine and anus. The reproductive system, in the hermaphrodite, comprises the gonads, spermatheca and vulval muscles. In the adult hermaphrodite, 5-HT is present in six types of neurons NSM; HSN; VC4/5; ADF; RIH and AIM (Figure 4). NSM, HSN and ADF being the only neurons that synthesize 5-HT, that is reuptaken by RIH, AIM and possibly by VC4 and VC5 [201; 209; 210]. The NSM is a polymodal neuron located in the pharynx (pharyngeal NeuroSecretory, Motor, sensory neuron). It projects directly to five types of cells in the pharynx; three of them are neurons and the other two are muscle cells. ADF is the only serotonergic sensory neuron in the hermaphrodite animal and may couple food sensing with serotonergic neurotransmission. AIM and RIH are both interneurons with pharyngeal localization, RIH being also a motor neuron. Both interneurons participate in the integration of external cues and modulation of C. elegans behaviour, namely on feeding response, locomotion, swimming and mating. HSN and VC4/5 are motor neurons located near the vulva, and they participate in egg-laying behaviour [211]. Similarly to the mammalian system, C. elegans serotonergic neurons project to several others that express different neurotransmitters, mainly glutamatergic (30%) and cholinergic (38.3%) neurons (Figure 4), some behaviours being accomplished by serotonergic modulation of these systems [191; 197; 212; 213]. Chapter 1. 19 Figure 4. Serotonergic neurons in C. elegans hermaphrodite. Serotonergic neurons (cell bodies) present a pharyngeal and vulval localization (Black boxes). The serotonergic neurons NSM, ADF, VC4/5, AIM, RIH and HSN (yellow boxes) project directly to several other neuronal types that synthesize GABA (green boxes), acetylcholine (red boxes), glutamate (blue boxes) or other (brown boxes) neurotransmitter as dopamine, 5-HT, neuropeptides or PDF. The neurons that do not possess classification in terms of neurotransmitter expression were omitted. From the total direct synapses established by serotonergic neurons, 38.3% are with cholinergic neurons and 30.0% with glutamatergic neurons. Just a small percentage of post-synaptic neurons are GABAergic, corresponding to 6.7%. All other classes of target neurons, that express other neurotransmitters, correspond to 15%, 10% of the synapses being established with neurons without neurotransmitter classification. Information of the neurotransmitters synthesized by the neurons was collected from [211]. Direct synapses established by serotonergic neurons collected from [214]. Until this date, four metabotropic 5-HTRs have been identified in C. elegans : SER-1, SER-4, SER5 and SER-7 with orthology to the mammalian 5-HT2R, 5-HT1AR, 5-HT6R and 5-HT7R, respectively [43; 169; 215; 216; 217; 218; 219; 220], to which adds a nematode specific serotonin-gated chloride channel Chapter 1. 20 MOD-1 [221]. These receptors are expressed in several neuronal and muscle cells in the head, vulva and tail of the animal. Overall, the synaptic localization of C. elegans serotonin receptors resembles that of its mammalian orthologs, with SER-4 being expressed in serotonergic neurons, as autoreceptors, and in nonserotonergic neurons, as heteroreceptors, as its orthologue 5-HT1AR (Figure 1). SER-1, SER-5, SER-7 and MOD-1 are heteroreceptors expressed in non-serotonergic neurons, as its orthologs 5-HT2R, 5-HT6R and 5-HT7R (Figure 1) [193; 215; 222]. G-proteins associated with C. elegans 5-HTRs are also conserved. SER-1 and SER-7 receptors couple to Gq (EGL-30) and Gs (GSA-1), respectively, and the SER-4 receptor couples to Go (GOA-1) [223]. Interestingly, several of these 5-HTRs synergistically or antagonistically modulate behavioural responses. SER-1, SER-5 and SER-7 are involved in feeding behaviour and increased pharyngeal pumping rate [192; 212; 224]. In another dimension SER-4 and MOD-1 are essential to the “enhanced slowing response” [193]. Regarding egg-laying behaviour, SER-1 is the master receptor for the stimulation of egg-laying [225; 226], however it was reported that SER-7 is also involved in this stimulation [224], and in contrast MOD1 inhibits egg-laying behaviour [225]. The current model is one where egg-laying can be regulated by all 5-HTRs depending on cell expression localization, MOD-1 and SER-4 having a suppressive effect, and SER-1, SER-7 (and possibly SER-5) an enhancing effect on egg-laying [227]. Antagonistic modulation of two 5-HTRs was also reported for lifespan, with SER-1 inhibition increasing survival of the animals, in opposition to SER-4 inhibition that decreases short to mid-lifespan, measured between day 0 and day 15 post-hatching [228]. 5. Relevance of the study Bearing in mind that i) no treatment is still available for MJD and the nefarious impact of the disease on patients’ quality of life, once ii) serotonergic signalling modulation constitutes a promising therapy for this condition, the main aim of this thesis was to address the contribution of each of the 5HTRs to the suppression of mutant ATXN3 proteotoxicity. For that we used the nematode C. elegans that presented a simpler although fully functional serotonergic system, with mammalian defined orthology, and took advantage of a C. elegans model of ATXN3 proteotoxicity (the AT3q130 transgenic strain), that mimics key pathologic hallmarks of MJD. The determination of the role of 5-HT1A autoand heteroreceptors in the suppression of ATXN3 proteotoxicity and the investigation of the contribution of other 5-HTRs on CIT-mediated action in MJD Chapter 1. 21 pathogenesis may help to deepen the understanding of how serotonergic signalling impacts MJD and potentially reveal new therapeutical targets for this disease. References [1] G. Haberhausen, M.S. Damian, F. Leweke, and U. Muller, Spinocerebellar ataxia, type 3 (SCA3) is genetically identical to Machado-Joseph disease (MJD). J Neurol Sci 132 (1995) 71-5. doi: 10.1016/0022-510x(95)90927-i. [2] L. Lima, and P. Coutinho, Clinical criteria for diagnosis of Machado-Joseph disease: report of a nonAzorena Portuguese family. Neurology 30 (1980) 319-22. doi: 10.1212/wnl.30.3.319. [3] N. Mendonca, M.C. Franca, Jr., A.F. Goncalves, and C. Januario, Clinical Features of Machado-Joseph Disease. Adv Exp Med Biol 1049 (2018) 255-273. doi: 10.1007/978-3-319-71779-1_13. [4] H. Paulson, Machado-Joseph disease/spinocerebellar ataxia type 3. Handb Clin Neurol 103 (2012) 437-49. doi: 10.1016/B978-0-444-51892-7.00027-9. [5] J. Sequeiros, and P. Coutinho, Epidemiology and clinical aspects of Machado-Joseph disease. Adv Neurol 61 (1993) 139-53. http://www.ncbi.nlm.nih.gov/pubmed/8421964 [6] L. Sudarsky, and P. Coutinho, Machado-Joseph disease. Clin Neurosci 3 (1995) 17-22. http://www.ncbi.nlm.nih.gov/pubmed/7614089 [7] L. Schols, G. Amoiridis, J.T. Epplen, M. Langkafel, H. Przuntek, and O. Riess, Relations between genotype and phenotype in German patients with the Machado-Joseph disease mutation. J Neurol Neurosurg Psychiatry 61 (1996) 466-70. doi: 10.1136/jnnp.61.5.466. [8] I. Silveira, C. Miranda, L. Guimaraes, M.C. Moreira, I. Alonso, P. Mendonca, A. Ferro, J. Pinto-Basto, J. Coelho, F. Ferreirinha, J. Poirier, E. Parreira, J. Vale, C. Januario, C. Barbot, A. Tuna, J. Barros, R. Koide, S. Tsuji, S.E. Holmes, R.L. Margolis, L. Jardim, M. Pandolfo, P. Coutinho, and J. Sequeiros, Trinucleotide repeats in 202 families with ataxia: a small expanded (CAG)n allele at the SCA17 locus. Arch Neurol 59 (2002) 623-9. doi: 10.1001/archneur.59.4.623. [9] H.T. Orr, and H.Y. Zoghbi, Trinucleotide repeat disorders. Annu Rev Neurosci 30 (2007) 575-621. doi: 10.1146/annurev.neuro.29.051605.113042. [10] S.L. Butland, R.S. Devon, Y. Huang, C.L. Mead, A.M. Meynert, S.J. Neal, S.S. Lee, A. Wilkinson, G.S. Yang, M.M. Yuen, M.R. Hayden, R.A. Holt, B.R. Leavitt, and B.F. Ouellette, CAG-encoded polyglutamine length polymorphism in the human genome. BMC Genomics 8 (2007) 126. doi: 10.1186/1471-2164-8-126. [11] A.R. La Spada, and J.P. Taylor, Polyglutamines placed into context. Neuron 38 (2003) 681-4. doi: 10.1016/s0896-6273(03)00328-3. [12] Y. Takiyama, S. Igarashi, E.A. Rogaeva, K. Endo, E.I. Rogaev, H. Tanaka, R. Sherrington, K. Sanpei, Y. Liang, M. Saito, and et al., Evidence for inter-generational instability in the CAG repeat in the MJD1 gene and for conserved haplotypes at flanking markers amongst Japanese and Caucasian subjects with Machado-Joseph disease. Hum Mol Genet 4 (1995) 1137-46. doi: 10.1093/hmg/4.7.1137. [13] H.L. Paulson, S.S. Das, P.B. Crino, M.K. Perez, S.C. Patel, D. Gotsdiner, K.H. Fischbeck, and R.N. Pittman, Machado-Joseph disease gene product is a cytoplasmic protein widely expressed in brain. Ann Neurol 41 (1997) 453-62. doi: 10.1002/ana.410410408. [14] P. Maciel, C. Gaspar, A.L. DeStefano, I. Silveira, P. Coutinho, J. Radvany, D.M. Dawson, L. Sudarsky, J. Guimaraes, J.E. Loureiro, and et al., Correlation between CAG repeat length and clinical features in Machado-Joseph disease. Am J Hum Genet 57 (1995) 54-61. http://www.ncbi.nlm.nih.gov/pubmed/7611296 Chapter 1. 22 [15] Y. Kawaguchi, T. Okamoto, M. Taniwaki, M. Aizawa, M. Inoue, S. Katayama, H. Kawakami, S. Nakamura, M. Nishimura, I. Akiguchi, and et al., CAG expansions in a novel gene for MachadoJoseph disease at chromosome 14q32.1. Nat Genet 8 (1994) 221-8. doi: 10.1038/ng1194221. [16] H. Maruyama, S. Nakamura, Z. Matsuyama, T. Sakai, M. Doyu, G. Sobue, M. Seto, M. Tsujihata, T. Oh-i, T. Nishio, and et al., Molecular features of the CAG repeats and clinical manifestation of Machado-Joseph disease. Hum Mol Genet 4 (1995) 807-12. doi: 10.1093/hmg/4.5.807. [17] H.L. Paulson, M.K. Perez, Y. Trottier, J.Q. Trojanowski, S.H. Subramony, S.S. Das, P. Vig, J.L. Mandel, K.H. Fischbeck, and R.N. Pittman, Intranuclear inclusions of expanded polyglutamine protein in spinocerebellar ataxia type 3. Neuron 19 (1997) 333-44. doi: 10.1016/s08966273(00)80943-5. [18] Y. Chai, S.S. Berke, R.E. Cohen, and H.L. Paulson, Poly-ubiquitin binding by the polyglutamine disease protein ataxin-3 links its normal function to protein surveillance pathways. J Biol Chem 279 (2004) 3605-11. doi: 10.1074/jbc.M310939200. [19] M. Albrecht, D. Hoffmann, B.O. Evert, I. Schmitt, U. Wullner, and T. Lengauer, Structural modeling of ataxin-3 reveals distant homology to adaptins. Proteins 50 (2003) 355-70. doi: 10.1002/prot.10280. [20] A.E. Bevivino, and P.J. Loll, An expanded glutamine repeat destabilizes native ataxin-3 structure and mediates formation of parallel beta -fibrils. Proc Natl Acad Sci U S A 98 (2001) 11955-60. doi: 10.1073/pnas.211305198. [21] A.M. Ellisdon, B. Thomas, and S.P. Bottomley, The two-stage pathway of ataxin-3 fibrillogenesis involves a polyglutamine-independent step. J Biol Chem 281 (2006) 16888-96. doi: 10.1074/jbc.M601470200. [22] L. Gales, L. Cortes, C. Almeida, C.V. Melo, M.C. Costa, P. Maciel, D.T. Clarke, A.M. Damas, and S. Macedo-Ribeiro, Towards a structural understanding of the fibrillization pathway in MachadoJoseph's disease: trapping early oligomers of non-expanded ataxin-3. J Mol Biol 353 (2005) 64254. doi: 10.1016/j.jmb.2005.08.061. [23] C. Bettencourt, and M. Lima, Machado-Joseph Disease: from first descriptions to new perspectives. Orphanet J Rare Dis 6 (2011) 35. doi: 10.1186/1750-1172-6-35. [24] A. Durr, G. Stevanin, G. Cancel, C. Duyckaerts, N. Abbas, O. Didierjean, H. Chneiweiss, A. Benomar, O. Lyon-Caen, J. Julien, M. Serdaru, C. Penet, Y. Agid, and A. Brice, Spinocerebellar ataxia 3 and Machado-Joseph disease: clinical, molecular, and neuropathological features. Ann Neurol 39 (1996) 490-9. doi: 10.1002/ana.410390411. [25] J.D. Fratkin, and P.J. Vig, Neuropathology of degenerative ataxias. Handb Clin Neurol 103 (2012) 111-25. doi: 10.1016/B978-0-444-51892-7.00005-X. [26] U. Rub, E.R. Brunt, R.A. de Vos, D. Del Turco, K. Del Tredici, K. Gierga, C. Schultz, E. Ghebremedhin, K. Burk, G. Auburger, and H. Braak, Degeneration of the central vestibular system in spinocerebellar ataxia type 3 (SCA3) patients and its possible clinical significance. Neuropathol Appl Neurobiol 30 (2004) 402-14. doi: 10.1111/j.1365-2990.2004.00554.x. [27] Y. Takiyama, S. Oyanagi, S. Kawashima, H. Sakamoto, K. Saito, M. Yoshida, S. Tsuji, Y. Mizuno, and M. Nishizawa, A clinical and pathologic study of a large Japanese family with Machado-Joseph disease tightly linked to the DNA markers on chromosome 14q. Neurology 44 (1994) 1302-8. doi: 10.1212/wnl.44.7.1302. [28] W.O. Arruda, A.T. Meira, S.E. Ono, A. de Carvalho Neto, L. Betting, S. Raskin, C.H.F. Camargo, and H.A.G. Teive, Volumetric MRI Changes in Spinocerebellar Ataxia (SCA3 and SCA10) Patients. Cerebellum 19 (2020) 536-543. doi: 10.1007/s12311-020-01137-3. Chapter 1. 29 [123] B. Pouzet, M. Didriksen, and J. Arnt, Effects of the 5-HT(7) receptor antagonist SB-258741 in animal models for schizophrenia. Pharmacol Biochem Behav 71 (2002) 655-65. doi: 10.1016/s0091-3057(01)00744-4. [124] A. Wesolowska, A. Nikiforuk, K. Stachowicz, and E. Tatarczynska, Effect of the selective 5-HT7 receptor antagonist SB 269970 in animal models of anxiety and depression. Neuropharmacology 51 (2006) 578-86. doi: 10.1016/j.neuropharm.2006.04.017. [125] P.B. Hedlund, L. Kelly, C. Mazur, T. Lovenberg, J.G. Sutcliffe, and P. Bonaventure, 8-OH-DPAT acts on both 5-HT1A and 5-HT7 receptors to induce hypothermia in rodents. Eur J Pharmacol 487 (2004) 125-32. doi: 10.1016/j.ejphar.2004.01.031. [126] S. Hjorth, Hypothermia in the rat induced by the potent serotoninergic agent 8-OH-DPAT. J Neural Transm 61 (1985) 131-5. doi: 10.1007/BF01253058. [127] C.T. Dourish, P.H. Hutson, G.A. Kennett, and G. Curzon, 8-OH-DPAT-induced hyperphagia: its neural basis and possible therapeutic relevance. Appetite 7 Suppl (1986) 127-40. doi: 10.1016/s0195-6663(86)80058-7. [128] P. Mazzola-Pomietto, C.S. Aulakh, K.M. Wozniak, and D.L. Murphy, Evidence that mchlorophenylpiperazine-induced hyperthermia in rats is mediated by stimulation of 5-HT2C receptors. Psychopharmacology (Berl) 123 (1996) 333-9. doi: 10.1007/BF02246643. [129] A. Hayashi, M. Suzuki, M. Sasamata, and K. Miyata, Thermogenic effect of YM348, a novel 5-HT2Creceptor agonist, in rats. J Pharm Pharmacol 56 (2004) 1551-6. doi: 10.1211/0022357044841. [130] D.D. Lam, M.J. Przydzial, S.H. Ridley, G.S. Yeo, J.J. Rochford, S. O'Rahilly, and L.K. Heisler, Serotonin 5-HT2C receptor agonist promotes hypophagia via downstream activation of melanocortin 4 receptors. Endocrinology 149 (2008) 1323-8. doi: 10.1210/en.2007-1321. [131] V.S. Naumenko, E.M. Kondaurova, and N.K. Popova, On the role of brain 5-HT7 receptor in the mechanism of hypothermia: comparison with hypothermia mediated via 5-HT1A and 5-HT3 receptor. Neuropharmacology 61 (2011) 1360-5. doi: 10.1016/j.neuropharm.2011.08.022. [132] A. Fisas, X. Codony, G. Romero, A. Dordal, J. Giraldo, R. Merce, J. Holenz, N. Vrang, R.V. Sorensen, D. Heal, H. Buschmann, and P.J. Pauwels, Chronic 5-HT6 receptor modulation by E-6837 induces hypophagia and sustained weight loss in diet-induced obese rats. Br J Pharmacol 148 (2006) 973-83. doi: 10.1038/sj.bjp.0706807. [133] V. Compan, M. Zhou, R. Grailhe, R.A. Gazzara, R. Martin, J. Gingrich, A. Dumuis, D. Brunner, J. Bockaert, and R. Hen, Attenuated response to stress and novelty and hypersensitivity to seizures in 5-HT4 receptor knock-out mice. J Neurosci 24 (2004) 412-9. doi: 10.1523/JNEUROSCI.280603.2004. [134] A. Jean, G. Conductier, C. Manrique, C. Bouras, P. Berta, R. Hen, Y. Charnay, J. Bockaert, and V. Compan, Anorexia induced by activation of serotonin 5-HT4 receptors is mediated by increases in CART in the nucleus accumbens. Proc Natl Acad Sci U S A 104 (2007) 16335-40. doi: 10.1073/pnas.0701471104. [135] A. Jean, L. Laurent, J. Bockaert, Y. Charnay, N. Dusticier, A. Nieoullon, M. Barrot, R. Neve, and V. Compan, The nucleus accumbens 5-HTR(4)-CART pathway ties anorexia to hyperactivity. Transl Psychiatry 2 (2012) e203. doi: 10.1038/tp.2012.131. [136] P.J. Fletcher, M. Tampakeras, J. Sinyard, and G.A. Higgins, Opposing effects of 5-HT(2A) and 5HT(2C) receptor antagonists in the rat and mouse on premature responding in the five-choice serial reaction time test. Psychopharmacology (Berl) 195 (2007) 223-34. doi: 10.1007/s00213007-0891-z. [137] G.A. Higgins, M. Enderlin, M. Haman, and P.J. Fletcher, The 5-HT2A receptor antagonist M100,907 attenuates motor and 'impulsive-type' behaviours produced by NMDA receptor antagonism. Psychopharmacology (Berl) 170 (2003) 309-319. doi: 10.1007/s00213-003-1549-0. Chapter 1. 30 [138] P.L. Rock, C.J. Harmer, S.F. McTavish, G.M. Goodwin, and R.D. Rogers, Short-term quetiapine treatment alters the use of reinforcement signals during risky decision-making and promotes the choice of negative expected values in healthy adult males. J Neurosci 33 (2013) 15588-95. doi: 10.1523/JNEUROSCI.5721-11.2013. [139] C.A. Winstanley, D.E. Theobald, J.W. Dalley, J.C. Glennon, and T.W. Robbins, 5-HT2A and 5-HT2C receptor antagonists have opposing effects on a measure of impulsivity: interactions with global 5-HT depletion. Psychopharmacology (Berl) 176 (2004) 376-85. doi: 10.1007/s00213-0041884-9. [140] P.M. Pitychoutis, A. Belmer, I. Moutkine, J. Adrien, and L. Maroteaux, Mice Lacking the Serotonin Htr2B Receptor Gene Present an Antipsychotic-Sensitive Schizophrenic-Like Phenotype. Neuropsychopharmacology 40 (2015) 2764-73. doi: 10.1038/npp.2015.126. [141] J. Bockaert, S. Claeysen, V. Compan, and A. Dumuis, 5-HT(4) receptors: history, molecular pharmacology and brain functions. Neuropharmacology 55 (2008) 922-31. doi: 10.1016/j.neuropharm.2008.05.013. [142] J. Bockaert, S. Claeysen, V. Compan, and A. Dumuis, 5-HT(4) receptors, a place in the sun: act two. Curr Opin Pharmacol 11 (2011) 87-93. doi: 10.1016/j.coph.2011.01.012. [143] E.S. Mitchell, and J.F. Neumaier, 5-HT6 receptors: a novel target for cognitive enhancement. Pharmacol Ther 108 (2005) 320-33. doi: 10.1016/j.pharmthera.2005.05.001. [144] I. Kendall, H.A. Slotten, X. Codony, J. Burgueno, P.J. Pauwels, J.M. Vela, and K.C. Fone, E-6801, a 5-HT6 receptor agonist, improves recognition memory by combined modulation of cholinergic and glutamatergic neurotransmission in the rat. Psychopharmacology (Berl) 213 (2011) 413-30. doi: 10.1007/s00213-010-1854-3. [145] M. Pereira, B.J. Martynhak, R. Andreatini, and P. Svenningsson, 5-HT6 receptor agonism facilitates emotional learning. Front Pharmacol 6 (2015) 200. doi: 10.3389/fphar.2015.00200. [146] S. Woods, N.N. Clarke, R. Layfield, and K.C. Fone, 5-HT(6) receptor agonists and antagonists enhance learning and memory in a conditioned emotion response paradigm by modulation of cholinergic and glutamatergic mechanisms. Br J Pharmacol 167 (2012) 436-49. doi: 10.1111/j.1476-5381.2012.02022.x. [147] G. Sarkisyan, and P.B. Hedlund, The 5-HT7 receptor is involved in allocentric spatial memory information processing. Behav Brain Res 202 (2009) 26-31. doi: 10.1016/j.bbr.2009.03.011. [148] M.G. Frank, M.P. Stryker, and L.H. Tecott, Sleep and sleep homeostasis in mice lacking the 5-HT2c receptor. Neuropsychopharmacology 27 (2002) 869-73. doi: 10.1016/S0893-133X(02)003536. [149] P.B. Hedlund, S. Huitron-Resendiz, S.J. Henriksen, and J.G. Sutcliffe, 5-HT7 receptor inhibition and inactivation induce antidepressantlike behavior and sleep pattern. Biol Psychiatry 58 (2005) 8317. doi: 10.1016/j.biopsych.2005.05.012. [150] S. Kantor, R. Jakus, B. Balogh, A. Benko, and G. Bagdy, Increased wakefulness, motor activity and decreased theta activity after blockade of the 5-HT2B receptor by the subtype-selective antagonist SB-215505. Br J Pharmacol 142 (2004) 1332-42. doi: 10.1038/sj.bjp.0705887. [151] J.M. Monti, and H. Jantos, Effects of the 5-HT(7) receptor antagonist SB-269970 microinjected into the dorsal raphe nucleus on REM sleep in the rat. Behav Brain Res 167 (2006) 245-50. doi: 10.1016/j.bbr.2005.09.025. [152] D.R. Thomas, S. Melotto, M. Massagrande, A.D. Gribble, P. Jeffrey, A.J. Stevens, N.J. Deeks, P.J. Eddershaw, S.H. Fenwick, G. Riley, T. Stean, C.M. Scott, M.J. Hill, D.N. Middlemiss, J.J. Hagan, G.W. Price, and I.T. Forbes, SB-656104-A, a novel selective 5-HT7 receptor antagonist, modulates REM sleep in rats. Br J Pharmacol 139 (2003) 705-14. doi: 10.1038/sj.bjp.0705290. [153] J.M. Monti, Serotonin control of sleep-wake behavior. Sleep Med Rev 15 (2011) 269-81. doi: 10.1016/j.smrv.2010.11.003. Chapter 1. 31 [154] A.F. Almeida-Santos, R.C. Ferreira, I.D. Duarte, D.C. Aguiar, T.R. Romero, and F.A. Moreira, The antipsychotic aripiprazole induces antinociceptive effects: Possible role of peripheral dopamine D2 and serotonin 5-HT1A receptors. Eur J Pharmacol 765 (2015) 300-6. doi: 10.1016/j.ejphar.2015.08.053. [155] L. Fei, L. Abrardi, and R.D. Mediati, Unexpected effect of aripiprazole on nociceptive pain. Ther Adv Psychopharmacol 2 (2012) 211-2. doi: 10.1177/2045125312451271. [156] T. Meuser, C. Pietruck, A. Gabriel, G.X. Xie, K.J. Lim, and P. Pierce Palmer, 5-HT7 receptors are involved in mediating 5-HT-induced activation of rat primary afferent neurons. Life Sci 71 (2002) 2279-89. doi: 10.1016/s0024-3205(02)02011-8. [157] N.M. Goodfellow, C.D. Bailey, and E.K. Lambe, The native serotonin 5-HT(5A) receptor: electrophysiological characterization in rodent cortex and 5-HT(1A)-mediated compensatory plasticity in the knock-out mouse. J Neurosci 32 (2012) 5804-9. doi: 10.1523/JNEUROSCI.4849-11.2012. [158] G. Becker, R. Bolbos, N. Costes, J. Redoute, A. Newman-Tancredi, and L. Zimmer, Selective serotonin 5-HT1A receptor biased agonists elicitdistinct brain activation patterns: a pharmacoMRI study. Sci Rep 6 (2016) 26633. doi: 10.1038/srep26633. [159] A. Newman-Tancredi, Biased agonism at serotonin 5-HT1A receptors: Preferential postsynaptic activity for improved therapy of CNS disorders. Neuropsychiatry 1 (2011) 149-164. doi: 10.2217/npy.11.12. [160] P. De Deurwaerdere, R. Bharatiya, A. Chagraoui, and G. Di Giovanni, Constitutive activity of 5-HT receptors: Factual analysis. Neuropharmacology 168 (2020) 107967. doi: 10.1016/j.neuropharm.2020.107967. [161] U. Shah, H. Pincas, S.C. Sealfon, and J. González-Maeso, Chapter 11 - Structure and function of serotonin GPCR heteromers. in: C.P. Müller, and K.A. Cunningham, (Eds.), Handbook of Behavioral Neuroscience, Elsevier, 2020, pp. 217-238.Book Section doi: 10.1016/B978-0-44464125-0.00011-6. [162] S. Brenner, The genetics of Caenorhabditis elegans. Genetics 77 (1974) 71-94. http://www.ncbi.nlm.nih.gov/pubmed/4366476 [163] J.G. White, E. Southgate, J.N. Thomson, and S. Brenner, The structure of the nervous system of the nematode Caenorhabditis elegans. Philos Trans R Soc Lond B Biol Sci 314 (1986) 1-340. doi: 10.1098/rstb.1986.0056. [164] S.J. Cook, T.A. Jarrell, C.A. Brittin, Y. Wang, A.E. Bloniarz, M.A. Yakovlev, K.C.Q. Nguyen, L.T. Tang, E.A. Bayer, J.S. Duerr, H.E. Bulow, O. Hobert, D.H. Hall, and S.W. Emmons, Whole-animal connectomes of both Caenorhabditis elegans sexes. Nature 571 (2019) 63-71. doi: 10.1038/s41586-019-1352-7. [165] O. Hobert, Neurogenesis in the nematode Caenorhabditis elegans. WormBook (2010) 1-24. doi: 10.1895/wormbook.1.12.2. [166] J.F. Etchberger, A. Lorch, M.C. Sleumer, R. Zapf, S.J. Jones, M.A. Marra, R.A. Holt, D.G. Moerman, and O. Hobert, The molecular signature and cis-regulatory architecture of a C. elegans gustatory neuron. Genes Dev 21 (2007) 1653-74. doi: 10.1101/gad.1560107. [167] C.e.S. Consortium, Genome sequence of the nematode C. elegans: a platform for investigating biology. Science 282 (1998) 2012-8. doi: 10.1126/science.282.5396.2012. [168] C.H. Lai, C.Y. Chou, L.Y. Ch'ang, C.S. Liu, and W. Lin, Identification of novel human genes evolutionarily conserved in Caenorhabditis elegans by comparative proteomics. Genome Res 10 (2000) 703-13. doi: 10.1101/gr.10.5.703. [169] W. Kim, R.S. Underwood, I. Greenwald, and D.D. Shaye, OrthoList 2: A New Comparative Genomic Analysis of Human and Caenorhabditis elegans Genes. Genetics 210 (2018) 445-461. doi: 10.1534/genetics.118.301307. Chapter 1. 32 [170] T. Kaletta, and M.O. Hengartner, Finding function in novel targets: C. elegans as a model organism. Nat Rev Drug Discov 5 (2006) 387-98. doi: 10.1038/nrd2031. [171] A. Fire, S. Xu, M.K. Montgomery, S.A. Kostas, S.E. Driver, and C.C. Mello, Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391 (1998) 80611. doi: 10.1038/35888. [172] D.J. Dickinson, J.D. Ward, D.J. Reiner, and B. Goldstein, Engineering the Caenorhabditis elegans genome using Cas9-triggered homologous recombination. Nat Methods 10 (2013) 1028-34. doi: 10.1038/nmeth.2641. [173] K.M. Van Pelt, and M.C. Truttmann, Caenorhabditis elegans as a model system for studying agingassociated neurodegenerative diseases. Translational Medicine of Aging 4 (2020) 60-72. doi: 10.1016/j.tma.2020.05.001. [174] X. Chen, J.W. Barclay, R.D. Burgoyne, and A. Morgan, Using C. elegans to discover therapeutic compounds for ageing-associated neurodegenerative diseases. Chem Cent J 9 (2015) 65. doi: 10.1186/s13065-015-0143-y. [175] A. Teixeira-Castro, M. Ailion, A. Jalles, H.R. Brignull, J.L. Vilaca, N. Dias, P. Rodrigues, J.F. Oliveira, A. Neves-Carvalho, R.I. Morimoto, and P. Maciel, Neuron-specific proteotoxicity of mutant ataxin3 in C. elegans: rescue by the DAF-16 and HSF-1 pathways. Hum Mol Genet 20 (2011) 29963009. doi: 10.1093/hmg/ddr203. [176] F. Pohl, A. Teixeira-Castro, M.D. Costa, V. Lindsay, J. Fiuza-Fernandes, M. Goua, G. Bermano, W. Russell, P. Maciel, and P. Kong Thoo Lin, GST-4-Dependent Suppression of Neurodegeneration in C. elegans Models of Parkinson's and Machado-Joseph Disease by Rapeseed Pomace Extract Supplementation. Front Neurosci 13 (2019) 1091. doi: 10.3389/fnins.2019.01091. [177] D. Vilasboas-Campos, M.D. Costa, A. Teixeira-Castro, R. Rios, F.G. Silva, C. Bessa, A.C.P. Dias, and P. Maciel, Neurotherapeutic effect of Hyptis spp. leaf extracts in Caenorhabditis elegans models of tauopathy and polyglutamine disease: Role of the glutathione redox cycle. Free Radic Biol Med 162 (2021) 202-215. doi: 10.1016/j.freeradbiomed.2020.10.018. [178] S. Duarte-Silva, A. Silva-Fernandes, A. Neves-Carvalho, C. Soares-Cunha, A. Teixeira-Castro, and P. Maciel, Combined therapy with m-TOR-dependent and -independent autophagy inducers causes neurotoxicity in a mouse model of Machado-Joseph disease. Neuroscience 313 (2016) 162-73. doi: 10.1016/j.neuroscience.2015.11.030. [179] A. Silva-Fernandes, S. Duarte-Silva, A. Neves-Carvalho, M. Amorim, C. Soares-Cunha, P. Oliveira, K. Thirstrup, A. Teixeira-Castro, and P. Maciel, Chronic treatment with 17-DMAG improves balance and coordination in a new mouse model of Machado-Joseph disease. Neurotherapeutics 11 (2014) 433-49. doi: 10.1007/s13311-013-0255-9. [180] L.F. Lei, G.P. Yang, J.L. Wang, D.M. Chuang, W.H. Song, B.S. Tang, and H. Jiang, Safety and efficacy of valproic acid treatment in SCA3/MJD patients. Parkinsonism Relat Disord 26 (2016) 55-61. doi: 10.1016/j.parkreldis.2016.03.005. [181] Z.F. Altun, D.H. Hall, and L.A. Herndon, WormAtlas Hermaphrodite Handbook - Nervous System - General Description. WormAtlas (2005). doi: 10.3908/wormatlas.1.18. [182] P.J. Brockie, and A.V. Maricq, Ionotropic glutamate receptors: genetics, behavior and electrophysiology. WormBook (2006) 1-16. doi: 10.1895/wormbook.1.61.1. [183] J.B. Rand, Acetylcholine. WormBook (2007) 1-21. doi: 10.1895/wormbook.1.131.1. [184] E.M. Jorgensen, Gaba. WormBook (2005) 1-13. doi: 10.1895/wormbook.1.14.1. [185] C. Li, and K. Kim, Neuropeptides. WormBook (2008) 1-36. doi: 10.1895/wormbook.1.142.1. [186] D.L. Chase, and M.R. Koelle, Biogenic amine neurotransmitters in C. elegans. WormBook (2007) 1-15. doi: 10.1895/wormbook.1.132.1. [187] M. de Bono, and A.V. Maricq, Neuronal substrates of complex behaviors in C. elegans. Annu Rev Neurosci 28 (2005) 451-501. doi: 10.1146/annurev.neuro.27.070203.144259. Chapter 1. 33 [188] M. de Bono, Molecular approaches to aggregation behavior and social attachment. J Neurobiol 54 (2003) 78-92. doi: 10.1002/neu.10162. [189] A.C. Giles, J.K. Rose, and C.H. Rankin, Investigations of learning and memory in Caenorhabditis elegans. Int Rev Neurobiol 69 (2006) 37-71. doi: 10.1016/S0074-7742(05)69002-2. [190] Y. Zhang, H. Lu, and C.I. Bargmann, Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegans. Nature 438 (2005) 179-84. doi: 10.1038/nature04216. [191] J.C. Brewer, A.C. Olson, K.M. Collins, and M.R. Koelle, Serotonin and neuropeptides are both released by the HSN command neuron to initiate Caenorhabditis elegans egg laying. PLoS Genet 15 (2019) e1007896. doi: 10.1371/journal.pgen.1007896. [192] S. Srinivasan, L. Sadegh, I.C. Elle, A.G. Christensen, N.J. Faergeman, and K. Ashrafi, Serotonin regulates C. elegans fat and feeding through independent molecular mechanisms. Cell Metab 7 (2008) 533-44. doi: 10.1016/j.cmet.2008.04.012. [193] G. Gurel, M.A. Gustafson, J.S. Pepper, H.R. Horvitz, and M.R. Koelle, Receptors and other signaling proteins required for serotonin control of locomotion in Caenorhabditis elegans. Genetics 192 (2012) 1359-71. doi: 10.1534/genetics.112.142125. [194] C.H. Rankin, Nematode behavior: the taste of success, the smell of danger! Curr Biol 16 (2006) R89-91. doi: 10.1016/j.cub.2006.01.025. [195] L. Segalat, D.A. Elkes, and J.M. Kaplan, Modulation of serotonin-controlled behaviors by Go in Caenorhabditis elegans. Science 267 (1995) 1648-51. doi: 10.1126/science.7886454. [196] K.M. Collins, A. Bode, R.W. Fernandez, J.E. Tanis, J.C. Brewer, M.S. Creamer, and M.R. Koelle, Activity of the C. elegans egg-laying behavior circuit is controlled by competing activation and feedback inhibition. Elife 5 (2016). doi: 10.7554/eLife.21126. [197] L.E. Waggoner, G.T. Zhou, R.W. Schafer, and W.R. Schafer, Control of alternative behavioral states by serotonin in Caenorhabditis elegans. Neuron 21 (1998) 203-14. doi: 10.1016/s08966273(00)80527-9. [198] B.M. Song, and L. Avery, The pharynx of the nematode C. elegans: A model system for the study of motor control. Worm 2 (2013) e21833. doi: 10.4161/worm.21833. [199] B.M. Song, and L. Avery, Serotonin activates overall feeding by activating two separate neural pathways in Caenorhabditis elegans. J Neurosci 32 (2012) 1920-31. doi: 10.1523/JNEUROSCI.2064-11.2012. [200] N.F. Trojanowski, D.M. Raizen, and C. Fang-Yen, Pharyngeal pumping in Caenorhabditis elegans depends on tonic and phasic signaling from the nervous system. Sci Rep 6 (2016) 22940. doi: 10.1038/srep22940. [201] E.R. Sawin, R. Ranganathan, and H.R. Horvitz, C. elegans locomotory rate is modulated by the environment through a dopaminergic pathway and by experience through a serotonergic pathway. Neuron 26 (2000) 619-31. doi: 10.1016/s0896-6273(00)81199-x. [202] S.W. Flavell, D.M. Raizen, and Y.-J. You, Behavioral States. Genetics 216 (2020) 315-332. doi: 10.1534/genetics.120.303539. [203] C.I. Nussbaum-Krammer, and R.I. Morimoto, Caenorhabditis elegans as a model system for studying non-cell-autonomous mechanisms in protein-misfolding diseases. Dis Model Mech 7 (2014) 31-9. doi: 10.1242/dmm.013011. [204] V. Prahlad, T. Cornelius, and R.I. Morimoto, Regulation of the cellular heat shock response in Caenorhabditis elegans by thermosensory neurons. Science 320 (2008) 811-4. doi: 10.1126/science.1156093. [205] V. Prahlad, and R.I. Morimoto, Neuronal circuitry regulates the response of Caenorhabditis elegans to misfolded proteins. Proc Natl Acad Sci U S A 108 (2011) 14204-9. doi: 10.1073/pnas.1106557108. Chapter 1. 34 [206] M.C. Tatum, F.K. Ooi, M.R. Chikka, L. Chauve, L.A. Martinez-Velazquez, H.W.M. Steinbusch, R.I. Morimoto, and V. Prahlad, Neuronal serotonin release triggers the heat shock response in C. elegans in the absence of temperature increase. Curr Biol 25 (2015) 163-174. doi: 10.1016/j.cub.2014.11.040. [207] K.M. Berendzen, J. Durieux, L.W. Shao, Y. Tian, H.E. Kim, S. Wolff, Y. Liu, and A. Dillin, Neuroendocrine Coordination of Mitochondrial Stress Signaling and Proteostasis. Cell 166 (2016) 1553-1563 e10. doi: 10.1016/j.cell.2016.08.042. [208] Q. Zhang, X. Wu, P. Chen, L. Liu, N. Xin, Y. Tian, and A. Dillin, The Mitochondrial Unfolded Protein Response Is Mediated Cell-Non-autonomously by Retromer-Dependent Wnt Signaling. Cell 174 (2018) 870-883 e17. doi: 10.1016/j.cell.2018.06.029. [209] G. Jafari, Y. Xie, A. Kullyev, B. Liang, and J.Y. Sze, Regulation of extrasynaptic 5-HT by serotonin reuptake transporter function in 5-HT-absorbing neurons underscores adaptation behavior in Caenorhabditis elegans. J Neurosci 31 (2011) 8948-57. doi: 10.1523/JNEUROSCI.169211.2011. [210] H.R. Horvitz, M. Chalfie, C. Trent, J.E. Sulston, and P.D. Evans, Serotonin and octopamine in the nematode Caenorhabditis elegans. Science 216 (1982) 1012-4. doi: 10.1126/science.6805073. [211] Z.F. Altun, L.A. Herndon, C.A. Wolkow, C. Crocker, R. Lints, N.E. Schroeder, and D.H. Hall, WormAtlas, http://www.wormatlas.org 2002-2021; Accessed March 2021, [212] K.A. Cunningham, Z. Hua, S. Srinivasan, J. Liu, B.H. Lee, R.H. Edwards, and K. Ashrafi, AMPactivated kinase links serotonergic signaling to glutamate release for regulation of feeding behavior in C. elegans. Cell Metab 16 (2012) 113-21. doi: 10.1016/j.cmet.2012.05.014. [213] Y. Ishita, T. Chihara, and M. Okumura, Serotonergic modulation of feeding behavior in Caenorhabditis elegans and other related nematodes. Neurosci Res 154 (2020) 9-19. doi: 10.1016/j.neures.2019.04.006. [214] N. Bhatla, An Interactive Visualization of the C. elegans Neural Network, http://wormweb.org/neuralnet#c=NSM&m=1, 2009 - accessed March 2021, [215] M. Carre-Pierrat, D. Baillie, R. Johnsen, R. Hyde, A. Hart, L. Granger, and L. Segalat, Characterization of the Caenorhabditis elegans G protein-coupled serotonin receptors. Invert Neurosci 6 (2006) 189-205. doi: 10.1007/s10158-006-0033-z. [216] G.P. Harris, V.M. Hapiak, R.T. Wragg, S.B. Miller, L.J. Hughes, R.J. Hobson, R. Steven, B. Bamber, and R.W. Komuniecki, Three distinct amine receptors operating at different levels within the locomotory circuit are each essential for the serotonergic modulation of chemosensation in Caenorhabditis elegans. J Neurosci 29 (2009) 1446-56. doi: 10.1523/JNEUROSCI.458508.2009. [217] R.J. Hobson, J. Geng, A.D. Gray, and R.W. Komuniecki, SER-7b, a constitutively active Galphas coupled 5-HT7-like receptor expressed in the Caenorhabditis elegans M4 pharyngeal motorneuron. J Neurochem 87 (2003) 22-9. doi: 10.1046/j.1471-4159.2003.01967.x. [218] B. Olde, and W.R. McCombie, Molecular cloning and functional expression of a serotonin receptor from Caenorhabditis elegans. J Mol Neurosci 8 (1997) 53-62. doi: 10.1007/BF02736863. [219] C.M. Dempsey, S.M. Mackenzie, A. Gargus, G. Blanco, and J.Y. Sze, Serotonin (5HT), fluoxetine, imipramine and dopamine target distinct 5HT receptor signaling to modulate Caenorhabditis elegans egg-laying behavior. Genetics 169 (2005) 1425-36. doi: 10.1534/genetics.104.032540. [220] J. Xie, S. Dernovici, and P. Ribeiro, Mutagenesis analysis of the serotonin 5-HT2C receptor and a Caenorhabditis elegans 5-HT2 homologue: conserved residues of helix 4 and helix 7 contribute to agonist-dependent activation of 5-HT2 receptors. J Neurochem 92 (2005) 375-87. doi: 10.1111/j.1471-4159.2004.02867.x. Chapter 1. 35 [221] R. Ranganathan, S.C. Cannon, and H.R. Horvitz, MOD-1 is a serotonin-gated chloride channel that modulates locomotory behaviour in C. elegans. Nature 408 (2000) 470-5. doi: 10.1038/35044083. [222] S. Dernovici, T. Starc, J.A. Dent, and P. Ribeiro, The serotonin receptor SER-1 (5HT2ce) contributes to the regulation of locomotion in Caenorhabditis elegans. Dev Neurobiol 67 (2007) 189-204. doi: 10.1002/dneu.20340. [223] R.W. Fernandez, K. Wei, E.Y. Wang, D. Mikalauskaite, A. Olson, J. Pepper, N. Christie, S. Kim, S. Weissenborn, M. Sarov, and M.R. Koelle, Cellular Expression and Functional Roles of All 26 Neurotransmitter GPCRs in the C. elegans Egg-Laying Circuit. J Neurosci 40 (2020) 7475-7488. doi: 10.1523/JNEUROSCI.1357-20.2020. [224] R.J. Hobson, V.M. Hapiak, H. Xiao, K.L. Buehrer, P.R. Komuniecki, and R.W. Komuniecki, SER-7, a Caenorhabditis elegans 5-HT7-like receptor, is essential for the 5-HT stimulation of pharyngeal pumping and egg laying. Genetics 172 (2006) 159-69. doi: 10.1534/genetics.105.044495. [225] L. Carnell, J. Illi, S.W. Hong, and S.L. McIntire, The G-protein-coupled serotonin receptor SER-1 regulates egg laying and male mating behaviors in Caenorhabditis elegans. J Neurosci 25 (2005) 10671-81. doi: 10.1523/JNEUROSCI.3399-05.2005. [226] H. Xiao, V.M. Hapiak, K.A. Smith, L. Lin, R.J. Hobson, J. Plenefisch, and R. Komuniecki, SER-1, a Caenorhabditis elegans 5-HT2-like receptor, and a multi-PDZ domain containing protein (MPZ-1) interact in vulval muscle to facilitate serotonin-stimulated egg-laying. Developmental Biology 298 (2006) 379-391. doi: 10.1016/j.ydbio.2006.06.044. [227] V.M. Hapiak, R.J. Hobson, L. Hughes, K. Smith, G. Harris, C. Condon, P. Komuniecki, and R.W. Komuniecki, Dual excitatory and inhibitory serotonergic inputs modulate egg laying in Caenorhabditis elegans. Genetics 181 (2009) 153-63. doi: 10.1534/genetics.108.096891. [228] H. Murakami, and S. Murakami, Serotonin receptors antagonistically modulate Caenorhabditis elegans longevity. Aging Cell 6 (2007) 483-8. doi: 10.1111/j.1474-9726.2007.00303.x. 36 Chapter 2. Filter retardation and biochemical fractionation assays for detecting and quantifying neuronal polyglutamine aggregates using C. elegans protein lysates: protocol development attempts Chapter 2. 37 Data from this chapter is included in the following manuscripts: Jalles, A.*; Vieira, C.*; Pereira-Sousa, J.; Vilasboas-Campos, D.; Mota, A.F.; Vasconcelos, S.; Ferreira-Lomba, B.; Da Silva, J. D.; Maciel P.; Teixeira-Castro, A. “Aripiprazole offsets mutant ATXN3induced motor dysfunction by targeting dopamine D2 and serotonin 1A and 2A receptors in C. elegans ” (under revision – Frontiers in Molecular Neuroscience ) Duarte-Silva, S.; Vilasboas-Campos, D.; Neves-Carvalho, A.; Pereira-Sousa, J.; Teixeira-Castro, A. and Maciel, P. “The current lead molecular tweezer CLR-01: impact on in vivo models of MJD” (manuscript in preparation) Note: All data presented in this chapter was generated by the author of this thesis. Importantly, the techniques developed here were also used in the article presented in Chapter 3. Chapter 2. 38 Filter retardation and biochemical fractionation assays for detecting and quantifying neuronal polyglutamine aggregates using C. elegans protein lysates: protocol development attempts Joana Pereira-Sousa1,2,3, Daniela Vilasboas-Campos1,2, Patricia Maciel1,2, Andreia Teixeira-Castro1,2 1Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Braga 2ICVS/3B’s - PT Government Associate Laboratory Braga/Guimarães, Portugal 3Behavioural and Molecular Lab (Bn’ML), University of Minho, Braga Abstract Alzheimer’s, Parkinson’s and Machado-Joseph diseases are a set of neurodegenerative disorders collectively known as proteinopathies. The commonality of these diseases resides in the accumulation of specific proteins in the brain. Regardless of whether proteinopathies share, or not, a common disease mechanism is generally accepted that the strong tendency to aggregation of the specific proteins is tightly linked to pathogenesis and disease progression. For this reason, the ability to detect and quantify protein aggregation is of utmost importance. Several techniques allow the study of aggregation in vitro and in vivo , however assessment of the aggregation process continues to be very challenging. The most used in vivo technologies are based on the imaging of fluorescent proteins. These methods include confocal dynamic imaging, Förster/Fluorescence Resonance Energy Transfer (FRET) or Bimolecular Fluorescence Complementation (BiFC) and explore the propensity of the monomeric proteins to form immobile aggregates with fluorescent emission and puncta formation. Nonetheless, other characteristics of aggregates can be evaluated using biochemical techniques. Filter retardation assay and biochemical fractionation techniques explore the differential size and solubility of the different multimeric and aggregation species. Although these biochemical methods have been already applied to study aggregation in several model systems, they were generally used in a qualitative manner. Here, we optimised filter retardation assay and biochemical fractionation protocols, and by adding internal loading controls to the assays, made them more suitable to semi-quantification and statistical analysis. Furthermore, we explored the amenability of a combinatory strategy comprising confocal dynamic imaging, filter retardation assay and biochemical fractionation to assess aggregation of mutant ATXN3 in C. elegans , through a systematic approach that can be easily applied to other proteins . Using a C. elegans model of ATXN3 proteotoxicity and taking advantage of a known genetic modifier of ATXN3 aggregation, mod-5 , previously described by us, we optimised the biochemical methods as proposed. We showed a decrease in ATXN3 SDS-insoluble large aggregates, measured by filter retardation assay, with a concomitant decrease in SDS-resistant species assessed by biochemical fractionation. These Chapter 2. 45 As loading control, 5 µg of protein extracts were transferred to an Eppendorf tube with a screw cap, supplemented with Laemmli Buffer to a final concentration of 1X and boiled for 15 minutes. The samples were resolved in a 10% SDS-gel, as previously described [22; 33], for drug treatments. Gel was then incubated with AzureRed Fluorescent Total Protein Stain (Azure biosystems), following the manufacturer instructions, to determine total protein staining (TPS) signals. Total protein quantification was done in AzureSpot Analysis Software (Azure biosystems), according to the manufacturer’s instructions. For mod-5 KO animals, loading controls were resolved in an SDS-PAGE gel prepared with TGX™ FastCast™ Acrylamide Kit, 10% (Bio-Rad) to detect total protein staining, following the manufacturer’s instructions. Quantification of the gel images obtained in Image lab software (BioRad) in Tiff format, was performed using AzureSpot Analysis Software (Azure biosystems), according to the manufacturer’s instructions. After quantification, filter retardation ATXN3 densitometry values were normalized by TPS ratio. 2.6.2. Biochemical fractionation Four-day-old animals were collected from 96-well plates and washed three times with M9, to eliminate OP50 bacteria and progeny, the last wash being supplemented with an EDTA-free protease inhibitor cocktail (Roche), and flash-frozen in liquid nitrogen. For protein/ aggregate extraction 150 µl of cold RIPA Buffer (50 mM Tris pH 7.5, 150 mM NaCl, 25 mM EDTA, 0.2% Triton X-100) freshly supplemented with an EDTA-free protease inhibitor cocktail (Roche) was added and animals were disrupted by mechanic force using glass beads (Sigma) in a FastPrep®-24 (MP Biomedicals) (six cycles of 6 ms for 25 s followed by 5 min on ice). After a centrifugation at 500 g during 1 min, the supernatant was collected, and the total protein quantified by the Bradford assay (Bio-Rad). Protein concentration was adjusted to 4 µg/µl and a loading control fraction, corresponding to 50 µg of protein, was collected. The rest of the protein sample was submitted to a 22000 g centrifugation for 30 min at 4ºC. The pellet fraction was separated from the supernatant (soluble fraction). Pellets were then homogenized in 150 µl RIPA buffer supplemented with 2% SDS, at RT, by pipetting up-and-down. A second centrifugation step at RT was performed at 22000 g for 30 min. The supernatant (SDS-soluble fraction) was collected, and the remaining pellet were incubated for 16 hours in 100% formic acid at 50ºC, allowing formic acid evaporation [24; 36]. This sample was dissolved in 50 µl of Laemmli buffer (0.1 M Tris, pH 6.8, 16% glycerol, 2% SDS, β-mercaptoethanol and bromophenol blue powder) (SDS-resistant fraction). Fractions collected during the biochemical fractionation protocol were submitted to SDS-PAGE and Western-blot against ATXN3. To do so, 50 µg of protein of loading control fraction, 50 µg of protein of soluble fraction, Chapter 2. 46 40 µl of SDS-soluble fraction and the complete SDS-resistant fraction were loaded in a 10% SDS gel. After wet transfer, TPS was determined using AzureRed Fluorescent Total Protein dye (Azure biosystems), according to the manufacturer’s instructions. The membrane was then blocked with 5% non-fat milk in TBS 1X during 1h at room temperature. To detect ATXN3 protein, the membrane was incubated overnight at 4ºC with mouse anti-ATXN3 antibody (1:1000) (Millipore Cat# MAB5360). After washing with TBS 1X, the membrane was incubated 1h at RT with horseradish peroxidase-coupled anti-mouse antibody (BioRad). Chemiluminescence (ECL western-blotting detection reagents, Bio-Rad or ECL Radiance, Azure biosystems) was used and detected using Sapphire Biomolecular Imager (Azure biosystems). TPS was detected in the same apparatus following the manufacturer’s instructions, prior to the chemiluminescence assay. Quantification was done in AzureSpot Analysis Software (Azure biosystems), according to the manufacturer’s instructions. After quantification, the ATXN3 signal of each fraction was normalized to TPS of the loading control fraction of the corresponding treatment. The values shown in the violin plots are the ratio of ATXN3 (normalised to TPS of the loading control fraction) divided by ATXN3 values of the ATXN3 mutant animals (for mod-5 ) or vehicle-treated mutant ATXN3 animals (for drug treatments), for each fraction. Quartiles (Q1 and Q3) and median (M) were represented. 2.7. Statistical analysis All statistical analysis was performed using the IBM SPSS statistics 23 software (SPSS) and GraphPad Prism 8.0.1 software. Normal distribution of continuous variables was analysed according with the Shapiro-Wilk or Kolmogorov-Smirnov normality test, depending on sample size, and histogram plot distribution. Levene’s test assessment was performed to confirm homogeneity of variances between groups. When both assumptions were verified, groups were compared using One-Way ANOVA followed by post-hoc analysis by Tukey (Fig. 1C), Dunnett (Fig. 2A, Fig. 3D, Fig. 4B – number of aggregates/ TA assay 2 and area occupied by aggregates/ TA assay 2, Fig. 4D, Supplementary fig. 2, Supplementary fig. 3A, Supplementary fig. 4 – genetic modifiers), or using independent sample t-test when just two groups were compared (Fig. 3B – assay 2 and 3, Fig. 3B - area occupied by aggregates/ TA assay 1, Fig. 3C, Fig. 4B – assay 1 and 3, Fig. 4C – total area and area occupied by aggregates/ TA, Supplementary fig. 3B and 3C, Supplementary fig. 4 – CLR-01). For biochemical fractionation, a one-sample t-test with Bonferroni correction versus the value of 1 was applied (Fig. 2B, Fig. 3E, Fig. 4E). Chapter 2. 47 For samples with normal distribution but not displaying homogeneity of variances, Welch correction for independent sample t-test was reported (Fig. 4C – number of aggregates/ TA, Supplementary fig. 4 – APZ). When the data did not display a normal distribution, bootstrap with BCA correction was performed. Bootstrap sampling was followed by an independent sample t-test test with bias correction (Fig. 3B – number of aggregates/ TA assay 1). Detailed statistics for all experiments are shown in Supplementary table 1. 3. Results 3.1. Total protein determination – detection of smaller variations with high dynamic range To compare total protein staining (TPS) and tubulin loading controls in C. elegans lysates, we conducted SDS-PAGE followed by Western blot analysis using α-tubulin antibody in parallel with detection of TPS, using AzureRed® total protein dye in lysates from WT, ATXN3 mutant, mod-5 Δ; ATXN3 mutant and ATXN3 mutant animals treated with APZ. In all samples, tubulin load linearly correlated with TPS with a coefficient of determination (r2) of 0.7203 that supported the direct proportional relationship between the two variables, however without a perfect fit (Figure 1A). Interestingly, the evaluation of the coefficient of variation between the two variables, that measures the distance of each individual value from the average, showed a smaller variation of TPS of ≈14.8%, in contrast with a 23.8% variation in α-tubulin, a difference of almost 9% between the two methods (Figure 1B). Quantification of α-tubulin levels across the different genotypes showed a significant increase in the expression of this protein in mutant ATXN3 animals when compared to WT (Figure 1C). Furthermore, analysis of serial dilutions of protein samples from WT animals revealed a limited dynamic range of tubulin that was overcome by the dynamic range of TPS (Figure 1D), which allowed the detection of analytical differences among the close protein concentrations tested. Evaluation of the curves slope showed that TPS measurements are almost 10 times more sensitive allowing clear discrimination of slightly increasing protein loads within its linear range. Taking together, we opted for TPS as a standard internal loading control, at the expense of the use of α-tubulin as an internal loading control due to its low dynamic range, higher variability, and its marginal increased expression in the disease model (ATXN3 mutant strain). Chapter 2. 48 Figure 1. Tubulin versus total protein staining signals as loading control strategies in western-blotting of C. elegans lysates. (A) Tubulin intensity correlates with total protein staining (TPS) intensity, however without a perfect fit presenting a r2 of 0.7203 being (B) TPS much less variable than tubulin. (C) Tubulin intensity and expression is increased in ATXN3-mutant animals when compared with WT animals, when corrected to TPS. (D) Evaluation of the dynamic range of each loading control strategy showed that TPS (red) presented ten times higher dynamic range than tubulin (green) (105 in tubulin versus to 106 in TPS). (C) (n = 4, ± SEM), *P = 0.050, (ANOVA, Tukey). 3.2. Confocal dynamic imaging, filter retardation assay and biochemical fractionation to assess ATXN3 aggregation in C. elegans Previously, our laboratory showed that the ATXN3-proteotoxicity C. elegans model has a neuronspecific pattern of aggregation, as in some neurons mutant ATXN3 proteins remain in their soluble state, whereas in others there is the formation of protein aggregates [22]. Since in these animals mutant ATXN3 proteins are expressed in fusion with YFP, aggregation can be assessed by dynamic confocal imaging and Chapter 2. 49 quantified using MeVisLab [23]. Dynamic confocal imaging has great advantages, including the fact that can be performed in live animals and allows the study of the dynamics of protein foci in vivo . However, this technique is also time consuming, as it evaluates the aggregation load in individual animals, being highly dependent on the efficacy of the anaesthetic during acquisition to avoid animal movement. So, to complement the study of aggregation in this model, we explored two biochemical techniques: (a) filter retardation assay that captures large agglomerates of proteins which, because of their size, cannot pass through a membrane and (b) biochemical fractionation assay, that through the usage of different buffers separates protein species by their solubility. 3.2.1. Impact of genetic modifiers Previous studies showed that serotonergic signalling modulation, by citalopram (CIT) treatment, impacts on MJD pathogenesis, decreasing ATXN3 aggregation in several disease models. Interestingly, genetic ablation of the serotonin transporter in C. elegans , mod-5 , was shown to mimic CIT treatment [24], mod-5 mutation being a potent suppressor of ATXN3 aggregation. To evaluate the ability of the biochemical techniques in study to detect alterations in ATXN3 aggregation in C. elegans , we performed filter retardation assay and biochemical fractionation of ATXN3 mutant animals in the background of mod5 ablation . To fulfil the need of quantification, we optimised the filter retardation assay protocol by collecting a loading sample of 5 µg of total protein prior to the samples loading into the cellulose acetate membrane. These loading control samples were resolved into SDS-PAGE gels and stained to total protein (Supplementary figure 1A – bottom panel). The ratios between TPS within the gels were used to normalize the ATXN3 signal in filter retardation membranes. We detected higher ATXN3 aggregation in ATXN3 mutant animals that decreased in the absence of mod-5 (Figure 2A and Supplementary figure 1A). Normalization of ATXN3 signals by TPS did not alter the statistical results but allowed important correction for eventual errors in protein loading, which can significantly affect results interpretation (Supplementary figure 2). Similarly, to turn biochemical fractionation assay more prone to quantification and statistical analysis, a loading control sample was added to the SDS-PAGE gel and TPS signals were measured. Also, to avoid variations among biological replicates (due to chemiluminescence-based detection of ATXN3 intensity), intensity values of each fraction were normalized to TPS-corrected ATXN3 control levels (ATXN3 mutant animals), having control always a ratio of 1. For this reason, statistical analysis was based on one- Chapter 2. 50 sample t-test against value of 1. With this change in protocol, we managed to semi-quantify biochemical fractionation experiments and showed that mutation in mod-5 gene in mod-5 Δ; ATXN3 mutant animals consistently decreased ATXN3 SDS-resistant species of mutant ATXN3 animals (Figure 2B), this shift in ATXN3 aggregated species being accompanied by improvement of motor function of the animals (Supplementary figure 3A), but not altering ATXN3 steady-state levels in whole animal lysates (Supplementary figure 4). Figure 2. Biochemical methods, filter retardation assay and biochemical fractionation, are suitable to measure ATXN3 aggregation. Evaluation of ATXN3 aggregation by (A) filter retardation assay upon mod-5 ablation corroborates the suppression of aggregation previously reported by confocal live imaging techniques. (B) Biochemical fractionation of mod-5 Δ; ATXN3 mutant showed a marked decrease in ATXN3 SDS-resistant species. For filter retardation assay (A): ATXN3 densitometry values in each sample were plotted as the mean of each sample (normalized to TPS) divided by the average of ATXN3 mutant animals. (n = 3 - 4, ± SD), *P < 0.05 ***P < 0.001, (ANOVA, Dunnett); For biochemical fractionation (B): ratio of ATXN3 divided by ATXN3 values of the ATXN3 mutant animals (control), for each fraction, were plotted (n = 3, Q1 – M – Q3), **P < 0.01 (one sample t-test, with Bonferroni correction vs value 1). Chapter 2. 51 3.2.2. Impact of drug treatment To test the impact of APZ treatment on a C. elegans model of mutant ATXN3 proteotoxicity, we evaluated aggregation of ATXN3 by confocal dynamic imaging, as well as using filter retardation assay and biochemical fractionation. Live confocal imaging of at least 8 animals per condition was performed in triplicates (three independent assays). Assay 2 presented an increase in the number and size of mutant ATXN3 foci upon APZ treatment, however assays 1 and 3 showed limited impact of drug treatment (Figure 3A and 3B). Pool of the three assays, normalized to control, revealed no overall impact of APZ treatment on mutant ATXN3 foci (Figure 3C). Drug efficacy was ensured in the same animals by measuring motor function of the animals upon treatment (Supplementary figure 3B). When aggregation was assessed by the filter retardation assay, using a significantly higher number of animals per assay (≈ 3000 animals) and with a total of six independent replicates tested, no significant statistical effect on aggregation was observed, however, we showed a tendency for APZ treatment to increase ATXN3 aggregation (p = 0.054) with four out of six replicates presenting higher aggregation compared to control (Figure 3D and Supplementary figure 1B). To further explore the effects of APZ treatment on ATXN3 aggregation, we performed a biochemical fractionation assay of ATXN3 mutant animals in the presence of the compound. Although no statistical differences were found when comparing treated and untreated animals, there was also a tendency to a decrease in ATXN3 SDS-soluble species, with a parallel increase in ATXN3 SDS-resistant species upon treatment (Figure 3E). No alterations in ATXN3 steady-state level were found upon APZ treatment (Supplementary figure 4). Chapter 2. 52 Figure 3. APZ treatment had a limited impact on mutant ATXN3 aggregation. (A) Mutant ATXN3 expression pattern in the animals’ head upon treatment, obtained by confocal dynamic imaging. (B) One of the three independent assays performed by confocal dynamic imaging showed an increase in the number of ATXN3 aggregates (top) and size (bottom), however, (C) the pool of the three assays, normalized by the results of daily control, did not reach statistical significance. (D) Filter retardation analysis showed that four out of six samples presented higher aggregation than average of control animals with a borderline tendency to APZ treatment-mediated increase in ATXN3 Chapter 2. 53 aggregation. (E) APZ treatment had a limited impact on ATXN3 aggregation, when evaluated by biochemical fractionation, however three out of four assays point to an increase of SDS-resistant species. For confocal dynamic imaging: (B) Assay 1 (n = 9 - 10, ± SD) (unpaired t-test corrected for BCA); Assay 2 (n = 9, ± SD) *P < 0.05 (unpaired t-test); Assay 3 (n = 8 - 13, ± SD) (unpaired t-test). (C) Total area (n = 28 - 32, ± SD) (unpaired t-test); number of aggregates / Total area (n = 27 to 31, ± SD), (unpaired t-test); area occupied by aggregates / Total area (n = 26 to 31, ± SD), (unpaired t-test). For filter retardation assay (D): ATXN3 densitometry values were normalized to TPS measured by AzureRed® (Azure Biosystems) and the results were plotted dividing each individual normalized sample value by the vehicle average. (n = 4 - 6, Q1 – M – Q3), *P < 0.05, (ANOVA, Dunnett); For biochemical fractionation (E): ratio of ATXN3 divided by ATXN3 values of the vehicle, for each fraction, were plotted. (n = 3 - 4, Q1 – M – Q3), (one sample t-test, with Bonferroni correction vs value 1). Similarly, to address the possible impact of the molecular tweezer CLR-01 on ATXN3 aggregation we employed the same protocols as for APZ treatment. One out of the three assays, obtained by confocal dynamic imaging, showed a decrease in number of aggregates upon CLR-01 treatment (Figure 4A and B). The pool of those three assays showed a decrease in the number of ATXN3 foci (p = 0.049) (Figure 4C). However, aggregation of ATXN3 was not significantly changed when assessed by filter retardation assay (Figure 4D and Supplementary figure 1C), differences in ATXN3 solubility not being found upon CLR-01 treatment, by biochemical fractionation, however the majority of the assays show a tendency to a decrease in all ATXN3 species (soluble, SDS-soluble and SDS-resistant) (Figure 4E). As for APZ, CLR01 treatment efficacy was ensured by determining the impact of drug treatment in animals motor function, prior to animals’ collection for the aggregation assays (Supplementary figure 3C). Alterations of ATXN3 steady-state levels were not found upon CLR-01 treatment (Supplementary figure 4). Overall, these results suggested that the techniques used here can detect potential alterations in mutant ATXN3 aggregation and inform about ATXN3 solubility, in semi-quantitative manner. These techniques are limited by inter-individual variability among a population, which is clearly lower upon genetic modifications of strong effect, such as the deletion of mod-5 , in comparison to drug treatments, in which each individual may respond differently to compound treatment. Chapter 2. 54 Figure 4. CLR-01 treatment showed a limited impact on ATXN3 aggregation. (A) Mutant ATXN3 expression pattern in the animals’ head upon CLR-01 treatment, obtained by confocal dynamic imaging. (B) One of the three independent assays performed by confocal dynamic imaging showed a decrease in ATXN3 aggregates number (top) without altering the area occupied by the aggregates (bottom). (C) Pool of the three assays normalized by the respective control animals showed a marginal decrease in the number of ATXN3 aggregates in the animals’ head, however (D) filter retardation and (E) biochemical fractionation analyses showed a limited impact of CLR-01 treatment in ATXN3 Chapter 2. 61 [37] S. Gregoire, J. Irwin, and I. Kwon, Techniques for Monitoring Protein Misfolding and Aggregation in Vitro and in Living Cells. Korean J Chem Eng 29 (2012) 693-702. doi: 10.1007/s11814-0120060-x. [38] R. Li, and Y. Shen, An old method facing a new challenge: re-visiting housekeeping proteins as internal reference control for neuroscience research. Life Sci 92 (2013) 747-51. doi: 10.1016/j.lfs.2013.02.014. [39] R.E. Ferguson, H.P. Carroll, A. Harris, E.R. Maher, P.J. Selby, and R.E. Banks, Housekeeping proteins: a preliminary study illustrating some limitations as useful references in protein expression studies. Proteomics 5 (2005) 566-71. doi: 10.1002/pmic.200400941. [40] C.P. Moritz, Tubulin or Not Tubulin: Heading Toward Total Protein Staining as Loading Control in Western Blots. Proteomics 17 (2017). doi: 10.1002/pmic.201600189. [41] A.J. Rodrigues, M. do Carmo Costa, T.L. Silva, D. Ferreira, F. Bajanca, E. Logarinho, and P. Maciel, Absence of ataxin-3 leads to cytoskeletal disorganization and increased cell death. Biochim Biophys Acta 1803 (2010) 1154-63. doi: 10.1016/j.bbamcr.2010.07.004. [42] S. Mazzucchelli, A. De Palma, M. Riva, A. D'Urzo, C. Pozzi, V. Pastori, F. Comelli, P. Fusi, M. Vanoni, P. Tortora, P. Mauri, and M.E. Regonesi, Proteomic and biochemical analyses unveil tight interaction of ataxin-3 with tubulin. Int J Biochem Cell Biol 41 (2009) 2485-92. doi: 10.1016/j.biocel.2009.08.003. [43] J. Hinz, L. Lehnhardt, S. Zakrzewski, G. Zhang, and Z. Ignatova, Polyglutamine expansion alters the dynamics and molecular architecture of aggregates in dentatorubropallidoluysian atrophy. J Biol Chem 287 (2012) 2068-78. doi: 10.1074/jbc.M111.318915. [44] L.O. Narhi, J. Schmit, K. Bechtold-Peters, and D. Sharma, Classification of protein aggregates. J Pharm Sci 101 (2012) 493-8. doi: 10.1002/jps.22790. [45] E.E. Wanker, E. Scherzinger, V. Heiser, A. Sittler, H. Eickhoff, and H. Lehrach, Membrane filter assay for detection of amyloid-like polyglutamine-containing protein aggregates. Methods Enzymol 309 (1999) 375-86. doi: 10.1016/s0076-6879(99)09026-6. [46] M. Herrera-Vaquero, D. Bouquio, M. Kallab, K. Biggs, G. Nair, J. Ochoa, A. Heras-Garvin, C. Heid, I. Hadrovic, W. Poewe, G.K. Wenning, F.G. Klarner, T. Schrader, G. Bitan, and N. Stefanova, The molecular tweezer CLR01 reduces aggregated, pathologic, and seeding-competent alphasynuclein in experimental multiple system atrophy. Biochim Biophys Acta Mol Basis Dis 1865 (2019) 165513. doi: 10.1016/j.bbadis.2019.07.007. [47] R. Malik, H. Meng, P. Wongkongkathep, C.I. Corrales, N. Sepanj, R.S. Atlasi, F.G. Klarner, T. Schrader, M.J. Spencer, J.A. Loo, M. Wiedau, and G. Bitan, The molecular tweezer CLR01 inhibits aberrant superoxide dismutase 1 (SOD1) self-assembly in vitro and in the G93A-SOD1 mouse model of ALS. J Biol Chem 294 (2019) 3501-3513. doi: 10.1074/jbc.RA118.005940. [48] T. Vopel, K. Bravo-Rodriguez, S. Mittal, S. Vachharajani, D. Gnutt, A. Sharma, A. Steinhof, O. Fatoba, G. Ellrichmann, M. Nshanian, C. Heid, J.A. Loo, F.G. Klarner, T. Schrader, G. Bitan, E.E. Wanker, S. Ebbinghaus, and E. Sanchez-Garcia, Inhibition of Huntingtin Exon-1 Aggregation by the Molecular Tweezer CLR01. J Am Chem Soc 139 (2017) 5640-5643. doi: 10.1021/jacs.6b11039. [49] O. Sin, T. de Jong, A. Mata-Cabana, M. Kudron, M.A. Zaini, F.A. Aprile, R.I. Seinstra, E. Stroo, R.W. Prins, C.N. Martineau, H.H. Wang, W. Hogewerf, A. Steinhof, E.E. Wanker, M. Vendruscolo, C.F. Calkhoven, V. Reinke, V. Guryev, and E.A. Nollen, Identification of an RNA Polymerase III Regulator Linked to Disease-Associated Protein Aggregation. Mol Cell 65 (2017) 1096-1108 e6. doi: 10.1016/j.molcel.2017.02.022. [50] T.A. Scott, L.M. Quintaneiro, P. Norvaisas, P.P. Lui, M.P. Wilson, K.Y. Leung, L. Herrera-Dominguez, S. Sudiwala, A. Pessia, P.T. Clayton, K. Bryson, V. Velagapudi, P.B. Mills, A. Typas, N.D.E. Chapter 2. 62 Greene, and F. Cabreiro, Host-Microbe Co-metabolism Dictates Cancer Drug Efficacy in C. elegans. Cell 169 (2017) 442-456 e18. doi: 10.1016/j.cell.2017.03.040. [51] K.B. Pho, and L.T. MacNeil, Biology is the root of variability: cautionary tales in Caenorhabditis elegans biology. Biochem Soc Trans 47 (2019) 887-896. doi: 10.1042/BST20190001. [52] W. Lenhard, and A. Lenhard, Computation of Effect Sizes (2017), http://www.psychometrica.de/effect_size.html Accessed June 2021, doi: 10.13140/RG.2.2.17823.92329. [53] J.O. Uanhoro, Effect size calculators (2017), https://effect-size-calculator.herokuapp.com/ Accessed June 2021. Chapter 2. 63 Supplementary Material Chapter 2. 64 Supplementary figure 1. Filter retardation membranes and SDS-page total protein gels. Filter retardation membranes (upper panel) and SDS-PAGE total protein loading controls (bottom panel) for (A) genetic modifier mod-5 , (B) APZ and (C) CLR-01 treatments after ATXN3 Western blot probing. Highlighted rectangles are shown in the main Figures 2A, 3D and 4D. * Discarded samples, due to probable membrane clogging (assumed by reduced solvent flux upon vacuum initiation). Chapter 2. 65 Supplementary figure 2. Quantification and statistical analysis of filter retardation assay of mod-5 Δ; ATXN3 mutant animals. (A) ATXN3 intensity signals as a measure of ATXN3 aggregation, without TPS normalization, were decreased in mod-5 Δ; ATXN3 mutant animals in comparison with ATXN3 mutant animals. (B) The difference is maintained upon normalization of ATXN3 signals, using TPS. (C) TPS did not significantly differ among genotypes, however some experimental variability can be detected. (A): (n = 3 - 4, ± SD), **P < 0.01 ***P < 0.001, (ANOVA, Dunnett); (B): (n = 3 - 4, ± SD), *P < 0.05 ***P < 0.001, (ANOVA, Dunnett); (C): (n = 3 - 4, ± SD), (ANOVA, Dunnett). Chapter 2. 66 Supplementary figure 3. Motility assay performed in animals collected for filter retardation assay and biochemical fractionation experiments. Motor assessment demonstrating motor improvement in (A) genetic ablation of mod-5 in ATXN3 mutant background and drug efficacy of (B) APZ and (C) CLR-01 treatments. For mod-5 Δ; ATXN3 mutant (A) (n = 3, ± SD) ***P < 0.001 (ANOVA, Dunnett test); for drug treatments (B) (n = 9, ± SD) ***P < 0.001 (unpaired t-test); (C) (n = 9, ± SD) ***P < 0.001 (unpaired t-test). Chapter 2. 67 Supplementary figure 4. Total steady-state levels of ATXN3 protein promoted by the genetic modifier mod-5 or drug treatments. No alterations were found in ATXN3 full-length protein levels. For genetic modifier (n = 4, ± SEM) ns – non-significant differences (ANOVA, Dunnett test); for drug treatments (APZ) (n = 6, ± SEM) ns – non-significant differences (unpaired t-test with welch correction); (CLR-01) (n = 6, ± SEM) ns – non-significant differences (unpaired t-test). Chart representation (top). Membrane of ATXN3 and total protein staining obtained (bottom). Chapter 2. 68 Supplementary table 1. Statistical report. Effect sizes reported calculated as [52; 53]. Figure number Statistics Sample size Fig. 1 C F (2, 9) = 3,946, p = 0.0588, ω2p = 0.329 4 Fig. 2 A F (2, 8) = 17.98, p = 0.0011, ω2p = 0.755 3 to 4 B - Soluble T(2) = 1.555, pcorrected = 0.7803, d = 0.898 3 B - SDS soluble T(2) = 2.356, pcorrected = 0.4278, d = 1.360 3 B - SDS resistant T(2) = 19.00, pcorrected = 0.0084, d = -10.97 3 Fig. 3 B – number of aggregates / Total area - Assay 1 T(17) = -0.611, p = 0.552, d = -0.281 9 to 10 B – number of aggregates / Total area - Assay 2 T(16) = -2.420, p = 0.028, d = -1.141 9 B – number of aggregates / Total area - Assay 3 T(19) = 0.369, p = 0.716, d = 0.166 8 to 13 B – area occupied by aggregates / Total area Assay 1 T(16) = -1.624, p = 0.124, d = -0.766 9 B – area occupied by aggregates / Total area Assay 2 T(16) = -2.292, p = 0.036, d = -1.08 9 B – area occupied by aggregates / Total area Assay 3 T(19) = 1.095, p = 0.287, d = 0.492 8 to 13 C – Total area T(58) = 0.9583, p = 0.342, d = 0.248 28 to 32 C – number of aggregates / Total area T(56) = 1.186, p = 0.240, d = 0.312 27 to 31 C – area occupied by aggregates / Total area T(55) = 0.8746, p = 0.3856, d = 0.233 26 to 31 D F (2, 13) = 11.53, p = 0.0013, ω2p = 0.568 4 to 6 E - Soluble T(2) = 0.8012, pcorrected > 0.999, d = 0.462 3 E – SDS soluble T(3) = 3.119, pcorrected = 0.157, d = -1.560 4 E – SDS resistant T(3) = 1.820, pcorrected = 0.499, d = 0,910 4 Fig. 4 B – number of aggregates / Total area - Assay 1 T(14) = 0.6492, p = 0.5267, d = 0.350 5 to 11 B – number of aggregates / Total area - Assay 2 F (2, 29) = 8.332, p = 0.0014, ω2p = 0.314 10 to 11 B – number of aggregates / Total area - Assay 3 T(13) = 0.2388, p = 0.8150, d = 0.126 6 to 9 Chapter 2. 69 B – area occupied by aggregates / Total area Assay 1 T(14) = 2.014, p = 0.0636, d = 1.086 5 to 11 B – area occupied by aggregates / Total area Assay 2 F (2, 29) = 6.512, p = 0.0046, ω2p = 0.256 10 to 11 B – area occupied by aggregates / Total area Assay 3 T(13) = 1.732, p = 0.1069, d = 0.913 6 to 9 C – Total area T(50) = 0.7581, p = 0.4520, d = 0.211 24 to 28 C – number of aggregates / Total area T(45.66) = 2.020, p = 0.0493, d = 0.562 24 to 28 C – area occupied by aggregates / Total area T(35) = 1.890, p = 0.067, d = 0.633 15 to 22 D F (3, 11) = 0.9075, p= 0.4686, ω2p = -0.019 6 E - Soluble T(5) = 1.459, pcorrected = 0.6132, d = -0,595 6 E – SDS soluble T(5) = 0.9864, pcorrected > 0.9999, d = -0,403 6 E – SDS resistant T(5) = 1.945, pcorrected = 0.3192, d = -0,794 6 Supplementary Fig. 2 A F(2, 8) = 22.80, p < 0.001, ω2p = 0.798 3 to 4 B F(2, 8) = 17.98, p = 0.0011, ω2p = 0.755 3 to 4 C F(2, 8) = 0.2736, p = 0.767, ω2p = -0.152 3 to 4 Supplementary Fig. 3 A F(2, 5) = 74.78, p < 0.001, ω2p = 0.949 3 B T(16) = 9.454, p < 0.001, d = 4.457 9 C T(16) = 11.18, p < 0.001, d = 5.270 9 Supplementary Fig. 4 Genetic modifier F(2, 9) = 19.50, p < 0.001, ω2p = 0.755 4 APZ T(5.985) = 0.5101, p = 0.6282, d = 0.295 6 CLR-01 T(10) = 0.9033, p = 0.3876, d = 0.522 6 70 Chapter 3. Identification of the 5-HT1A serotonin receptor as a novel therapeutic target in a C. elegans model of Machado-Joseph disease Chapter 3. 77 3. Results 3.1. Chronic and acute treatment with befiradol ameliorates mutant ATXN3-mediated neuronal dysfunction 3.2. Treatment with befiradol decreases aggregation of mutant ATXN3 protein 3.3. Befiradol acts via 5-HT1AR orthologue SER-4 to mitigate motor dysfunction in C. elegans Figure 3. Persistence of effect of befiradol treatment of ATXN3 mutant animals upon drug removal. Chapter 3. 78 Figure 4. C. elegans SER-4 receptors are molecular targets of befiradol. Chapter 3. 79 3.4. Addressing 5HT1A autoand heteroreceptor contribution to befiradol’s therapeutic action Figure 5. 5-HT1A autoand heteroreceptors’ function contribute to befiradol treatment efficacy of ATXN3 mutant animals. Chapter 3. 80 4. Discussion Figure 6. Hypothetical model of befiradol action in a C. elegans model of ATXN3 proteotoxicity. Chapter 3. 81 Chapter 3. 82 Contributors Data sharing statement Declaration of interests Acknowledgements References Chapter 3. 83 Chapter 3. 84 Chapter 3. 85 Supplementary Material Chapter 3. 86 Supplementary figure 1. Befiradol and tandospirone toxicity evaluated using the food clearance assay. The optical density of the E. coli OP50 suspension was measured daily in WT animals treated with (a) befiradol or (b) tandospirone at a 500 to 0.0001 µM concentration range. The mean OD was calculated for each day from triplicate samples and plotted during the 7 days of the experiment. DMSO 1% (green line) corresponds to drug vehicle and was used as negative (safe) control, conversely DMSO 5% (red line) was used as a positive (toxic) control. Despite the fact that the curves for all concentrations of both drugs are not different from the respective negative control in a statistically significant manner, visual inspection of animals treated with 500 µM of befiradol revealed some deficits in body fitness (data not shown) consistent with mild drug toxicity, being therefore this concentration classified as not safe. The other concentrations were classified as safe, as the OD decrease paralleled the one of vehicle control samples and visual inspection confirmed normal growth of the animals. WT – Wild-type. Statistical analysis: Application of a non-linear regression model for sigmoidal curves against DMSO 1% presented unique model for LogIC50 and HillSlope values, suggestive of no statistical differences between the curves (befiradol P = 0.1352; tandospirone P = 0.0836). Chapter 3. 93 Supplementary table 1. Strains used in the study. *Strains generated in this work ID number Strain name Genotype N2 (Bristol) WT WT MAC037/ AM520 ATXN3 WT (AT3q75) rmls238 [P regf-1 ::AT3v1-1q75::YFP] MAC001/ AM685 ATXN3 mutant (AT3q130) rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II DA1814 ser-1 Δ ser-1 ( ok345 ) X MAC075 ser-1 Δ; ATXN3 mutant rmls263 [P rgef-1 ::AT3v1-1q130::YFP]; ser-1 ( ok345 ) X AQ866 ser-4 Δ ser-4 ( ok512 ) III MAC072 ser-4 Δ; ATXN3 mutant rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; ser-4 ( ok512 ) III tm2654 ser-5 Δ ser-5 ( tm2654 ) I MAC013* ser-5 Δ; ATXN3 mutant ser-5 ( tm2654 ) I; rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II MAC038* ATXN3 mutant ( ser-5 WT) ser-5 (+) I; rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II DA2100 ser-7 Δ ser-7 ( tm1325 ) X MAC017* ser-7 Δ; ATXN3 mutant rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; ser-7 ( tm1325 ) X MAC039* ATXN3 mutant ( ser-7 WT) rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; ser-7 (+) X MT9668 mod-1 Δ mod-1 ( ok103 ) V MAC019* mod-1 Δ; ATXN3 mutant rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; mod-1 ( ok103 ) V MAC040* ATXN3 mutant ( mod-1 WT) rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; mod-1 (+) V LX636 dop-1 Δ dop-1 ( vs101 ) X MAC171* dop-1 Δ; ATXN3 mutant rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; dop-1 ( vs101 ) X MAC168* ATXN3 mutant ( dop-1 WT) rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; dop-1 (+) X LX702 dop-2 Δ dop-2 ( vs105 ) V MAC222* dop-2 Δ; ATXN3 mutant rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; dop-2 ( vs105 ) V MAC167* ATXN3 mutant ( dop-2 WT) rmls263 [P rgef-1 ::AT3v1-1q130::YFP] II; dop-2 (+) V MAC390* tph-1 Δ; ATXN3 mutant rmls263 [P rgef-1 ::AT3v1-1q130::YFP] ; tph-1 ( mnh5 ) II LX2004 HSN GCaMP5, mCherry lite-1(ce314), vsIs183 lin-15(n765ts) X Chapter 3. 94 Supplementary table 2. List of primers used in the generation of double mutant strains. Gene Primer ID Sequence ser-5 ser-5 F1 5’- CTTCGTCGATTTTGCTCACA - 3’ ser-5 rev 3 5’- TGCGCCAGTAAAGTGTTAATGT - 3’ ser-5 R2 5’- CCAACCAACTTGCTACATCG - 3’ ser-7 ser-7_for1 5’- TGGATCTGCTAGCACTGTGG - 3’ ser-7_rev1 5’- CCGGAGATCCTAAGGTAGGC - 3’ ser-7_rev2 5’- ACCCCATTTTGGGCCTATAC - 3’ mod-1 mod-1_for1 5’- AATACTCCAGGTGCCAATGC - 3’ mod-1_rev2 5’- TTCCGTCGTTTTCCTTTAGC - 3’ mod-1_rev3 5’- TCGAATTGAATACCAACACAATG - 3’ dop-1 dop-1_for1 5’- TGTGCTGAAATGAACGAATGA - 3’ dop-1_rev1 5’- GGATTGGCAGAAGAGTTTGC - 3’ dop-1_rev2 5’- CGGAATGGTTTCCTCGTTAT - 3’ dop-2 dop-2_for1 5’- ACGATTCCTTGCGATTCTGG - 3’ dop-2_rev1 5’- CACAGCCAGGCCTAGGATAA - 3’ dop-2_rev2 5’- GGAGCCTATGCTGCTATGGA - 3’ tph-1 tph-1_fulldel_F1 5’- ACCACGCCATCGGATATCTA - 3’ tph-1_fulldel_F2 5’- GGATCGTGTTGTTGAGCAAG - 3’ tph-1_fulldel_R1 5’- GGAGAATCAATGGTCAACTCG - 3’ Chapter 3. 95 Supplementary table 3. List of primers used in reverse-transcriptase quantitative PCR. Gene Primer ID Sequence ser-1 ser-1 F1 5’- CGTACACTTGCTCGAGGTCA - 3’ ser-1 R1 5’- GTTGATGCCTCTGTCGTTGC - 3’ ser-5 ser-5 F1 5’- TTGCACTTGCTCAGTGGTGA - 3’ ser-5 R1 5’- TCCCACCGAACTGTTGTGAG - 3’ ser-7 ser-7 F1 5’- TGTGGATCCCGGATTGGTTG - 3’ ser-7 R1 5’- TTGCGTCGAGATAGAAGCGG - 3’ mod-1 mod-1 F1 5’- TGATCAGCTGCCAGTTTCCA - 3’ mod-1 R1 5’- AGTGCCACCAGTAGACAAGA - 3’ mod-5 mod-5 F1 5’- AACTGCTATCGTGACGCCG - 3’ mod-5 R1 5’- GAGCTTGGGGGTAGACGATG - 3’ Chapter 3. 96 Supplementary table 4. Statistical report. Effect size calculated using Uanhoro, J. O. (2017). Effect size calculators. Available online at: https://effect-sizecalculator.herokuapp.com/ . Figure Statistical Report Sample Size Fig.1a befiradol F(11, 70) = 12.965, p < 0.001, ω2p = 0.616 7 tandospirone F(11, 38.7) = 8.050, p = 0.005, ω2p = 0.605 6 Fig.1b F(4, 173) = 6.115, p < 0.001, ω2p = 0.103 18 to 43 Fig.1c F(4, 24) = 3.598, p = 0.0196, ω2p = 0.264 3 Fig.1d F(10, 22) = 11.896, p < 0.001, ω2p = 0.844 3 Fig.2b (number of aggregates / TA) Acute + TD Chronic H(3) = 10.48, p = 0.0149, η2 = 0.095 11 to 32 befiradol Chronic T(37) = 1.868, p = 0.0697, η2 = 0.086 17 to 22 Fig.2c (area of aggregates / TA) Acute + TD Chronic H(3) = 5.887 , p = 0.1173 , η2 =0.037 11 to 32 befiradol Chronic T(37) = 2.257, p = 0.0300, η2 = 0.121 17 to 22 Fig.2d (total area) Acute + TD Chronic F(3, 79) = 1.802, p = 0.1536, ω2p = 0.028 11 to 32 befiradol Chronic T(37) = 1.804, p = 0.0793, η2 = 0.081 17 to 22 Fig.2e Soluble T(1) = 0.6667, pcorrected < 0.9999, d = 0.471 2 SDS - soluble T(1) = 0.000, pcorrected < 0.9999, d = 0 2 SDS - resistant T(1) = 1.000, pcorrected < 0.9999, d = 0.707 2 Fig.2f Soluble T(4 ) = 0.520 , pcorrected = < 0.9999, d = -0.233 5 SDS - soluble T(3) = 7.437 , pcorrected = 0.015, d = 3.722 4 SDS - resistant T(1) = 0.423, pcorrected = < 0.9999, d = -0.299 2 Fig.2g F(2, 17) = 0.8325, p = 0.4519 , ω2p = 0.000 6 to 7 Fig.2h F(2 , 17) = 2.026, p = 0.1625, ω2p = 0.093 6 to 7 Fig.3 F(9, 52) = 2.398, p = 0.0234, ω2p = 0.168 3 Fig.4b F(8, 34) = 8.683, p < 0.001, ω2p = 0.588 3 Fig.4c F(8, 34) = 11.935, p < 0.001, ω2p = 0.670 3 Fig.4d F(10, 40) = 2.878, p = 0.008, ω2p = 0.269 3 Fig. 4e F(10, 60) = 3.237, p = 0.002, ω2p = 0.240 4 Fig. 4f F(10, 40) = 4.257, p < 0.001, ω2p = 0.390 3 Fig. 5a F(9, 6.160) = 6.039, p = 0.019, ω2p = 0.737 2 -3 Fig. 5b F(6, 14) = 102.703, p < 0.001, ω2p = 0.966 3 Fig. 5c F(2, 18.651) = 5.117, p = 0.017, ω2p = 0.275 11 - 12 S1a F(40, 44) = 1.406, p = 0.1352, ω2p = 0.160 5 Chapter 3. 97 S1b F(36, 40) = 1.568, p = 0.0836, ω2p = 0.210 5 S2a F (2, 20) = 0.277, p = 0.7609, ω2p = 0 7 - 8 S2b F (2, 9) = 0.471, p = 0.6387, ω2p = 0 4 S2c F (2, 3.353) = 13.72, p = 0.0243, ω2p = 0.800 4 S2d F (4, 20) = 8.233, p = 0.0004, ω2p = 0.536 3 - 6 S4a H(3) = 9.571, p = 0.0226, η2 = 0.411 5 S4b H(3) = 14.12, p = 0.0027, η2 = 0.695 5 S4c F(3, 16) = 8.179, p = 0.0016, ω2p = 0.518 5 S4d F(3, 16) = 5.598, p = 0.0081, ω2p = 0.408 5 S4e F(3, 16) = 14.201, p < 0.001, ω2p = 0.664 5 S5a F(4, 13) = 33.983, p < 0.001, ω2p = 0.880 4 S5b F(4, 24) = 16.341, p < 0.001, ω2p = 0.680 6 - 7 S6a F(16, 34.183) = 8.428, p < 0.001, ω2p = 0.699 4 S6b F(16, 12.372) = 22.505, p < 0.001, ω2p = 0.921 3 98 Chapter 4. Addressing the role of serotonin receptors in the suppression of mutant ataxin-3 proteotoxicity Chapter 4. 99 Addressing the role of serotonin receptors in the suppression of mutant ataxin-3 proteotoxicity Joana Pereira-Sousa1,2,3, Daniela Vilasboas-Campos1,2, Carmén Vieira1,2, Patricia Maciel1,2, Andreia TeixeiraCastro1,2 1Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Braga 2ICVS/3B’s - PT Government Associate Laboratory Braga/Guimarães, Portugal 3Behavioural and Molecular Lab (Bn’ML), University of Minho, Braga Abstract Serotonin is a monoamine neurotransmitter produced, in the brain, by serotonergic neurons. These neurons project to virtually all brain regions exerting a tonic modulatory effect. The pleotropic effects exerted by serotonin, in the extensive target sites, is ensured by a vast repertoire of serotonin receptors (5-HTRs). In humans, seven types of 5-HTRs were reported, 5-HT1-7, that can present subtypes, in a total of 14 receptors. Except for 5-HT3, all the 5-HTRs are metabotropic G protein-coupled receptors (GPCRs). In C. elegans , the serotonergic system is composed by four metabotropic GPCRs, namely SER1, SER-4, SER-5 and SER-7, orthologs of the mammalian receptors 5-HT2, 5-HT1A, 5-HT6 and 5-HT7, respectively, and by MOD-1, the nematode specific channel-gated serotonin receptor. Previously, our laboratory described the modulation of serotonergic signalling as a potential therapeutic strategy for Machado-Joseph disease (MJD). Citalopram (CIT) treatment improved motor coordination and decreased mutant ATXN3 aggregation in several animal models of the disease. CIT’s effect showed to be dependent on SERT, its molecular target, with an important contribution of the 5HT1A/SER-4, and also of 5-HT2/SER-1 receptors in C. elegans . Furthermore, direct and specific modulation of the 5-HT1AR alone was sufficient to reduce mutant ATXN3-mediated proteotoxicity. This led us to ask whether other 5-HTRs could also play a role in the suppression of MJD pathogenesis. Here, we found a potential partial contribution of the 5-HT6/SER-5 and 5-HT7/SER-7 receptors, as in the absence of these receptors therapeutic effect of CIT was reduced. In contrast, ablation of MOD1 seemed to be dispensable for CIT-mediated suppression of motor dysfunction in C. elegans . Next, using tool compounds likely targeting the different receptors, we observed that, surprisingly, antagonism of 5HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7 ameliorated motor function of the mutant animals. Although these findings are promising, further studies are needed to support the contribution of 5HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7 in MJD therapeutics and to dissect the specific pathways by which each receptor impacts on ATXN3 proteotoxicity. Chapter 4. 100 1. Introduction Serotonin is a monoamine neurotransmitter synthesised by the conversion of the amino acid Ltryptophan by tryptophane hydroxylase [1]. Serotonin can be found in central nervous system as well as in the enterochromaffin cells of the gastrointestinal tract, and in blood platelets [2; 3; 4]. Within the brain, serotonin is produced by cell bodies of the serotonergic neurons in the raphe nucleus [5; 6]. These neurons project, almost, to all brain regions exerting a tonic modulatory effect on its extensive target cells (reviewed in [7]). The pleiotropic effect exerted by serotonin in target cells leads to the modulation of several systems and to physiological and phenotypic responses. These events are ensured by the high diversity and number of serotonin receptors. In humans, seven types of 5-HT receptors (5-HTRs) were reported, 5-HT1-7. Over this, some receptor types present subtypes, including the 5-HT1A-F, 5-HT2A-C, 5-HT5A-B, in a total of 14 subtypes. 5-HTRs are encoded by 17 genes, 12 of them encode metabotropic 5-HTRs and five encode ionotropic receptor subunits of the 5-HT3R [8; 9]. 5-HT metabotropic receptors are G protein-coupled receptors (GPCRs) that elicit, upon serotonin binding, canonical signalling pathways by association with G proteins [10]. Depending on the G protein subunit associated with the receptor, the signalling cascade elicited by its activation can have an inhibitory or excitatory effect on neurons (reviewed on [8; 11]). 5-HT1 and 5-HT5 receptors are associated with the Gi/o subunit, which activation results in inhibition of adenylyl cyclase and decrease cAMP formation [12]. Furthermore, activation of these receptors opens G protein-gated inwardly rectifying potassium channels (GIRKs) leading to hyperpolarisation of the neuron and inhibits the opening of voltage-gated calcium channels [13]. Therefore, activation of 5-HT1R and 5HT5R produce an inhibitory effect on neurons. In contrast, 5-HT2R, 5-HT4R, 5-HT6R and 5-HT7R families are associated with Gq/11 or Gs subunits, eliciting an excitatory response. Conversely to Gi/o, Gs-coupled signalling activates adenylyl cyclase and increase cAMP formation. 5-HT2Rs are the ones interacting with Gαq proteins, leading to the activation of phospholipase C to promote increasing formation of inositol trisphosphate (IP3) and diacylglycerol (DAG). This results in the activation of protein kinase C (PKC) by DAG and a cytosolic raise in Ca2+ levels by IP3 [14; 15; 16]. In addition to these canonical pathways, 5-HT metabotropic receptors reveal non-canonical signalling by diverse molecules, except for the 5-HT5R. All 5-HTRs act through ERK, but also through small Chapter 4. 101 G proteins (for 5-HT1R and 5-HT7R), PLA2, PLD, Src/Akt (for 5-HT2R), and through mTOR, Cdk5 (for 5HT6R) [8; 11]. The C. elegans serotonergic system is composed of neurons expressing five types of receptors: SER-1, SER-4, SER-5, SER-7 and MOD-1 [17]. As in mammalian, these receptors are divided in metabotropic and ionotropic receptors. SER-1, SER-4, SER-5 and SER-7 are GPCRs with a high degree of homology to the mammalian receptors 5-HT2, 5-HT1A, 5-HT6 and 5-HT7 receptors, respectively [18; 19; 20; 21]. Therefore, throughout this chapter, the orthology of the receptors will be represented as follows: mammalian receptor/ C. elegans orthologue. MOD-1 is a serotonin-gated chloride channel specific of C. elegans nematode [22]. Previous work from our laboratory showed that modulation of serotonergic signalling supressed MJD pathogenesis [23; 24]. Citalopram (CIT) is a selective serotonin reuptake inhibitor (SSRI) that blocks the serotonin reuptake transporter (SERT), increasing serotonin availability in the synaptic cleft. Longterm administration of CIT to pre-symptomatic MJD animal models ameliorated motor dysfunction and decreased ATXN3 aggregation [24]. The effect of CIT showed to be dependent on MOD-5, the C. elegans orthologue of SERT, but also on 5-HT2R/SER-1 and 5-HT1AR/SER-4, suggesting a role for the serotonin receptors in the suppression of MJD pathogenesis. 5-HT1AR/SER-4 involvement was further supported by pharmacogenetic and pharmacological approaches. Acute and chronic administration of befiradol, a 5HT1AR/SER-4 potent and selective agonist [25; 26], ameliorated motor dysfunction and decreased mutant ATXN3 aggregation in a C. elegans model of mutant ATXN3 proteotoxicity [23]. Here, we started to explore the contribution of the different serotonin receptors to the suppression of MJD pathogenesis in C. elegans using tool compounds. Tool compounds (also known as chemical probes) are small molecules that are potent, selective to the target protein and with a known mode-ofaction, that allows researchers to ask mechanistic and phenotypic questions about their molecular targets in biochemistry-based or animal studies. Arrowsmith et al defined a chemical probe based on the following criteria: (i) in vitro potency of <100 nM at the protein target, (ii) >30-fold selectivity against other protein targets, and (iii) demonstration of on-target effect in cells at <1 μM [27]. These compounds comprise different classes of molecules with distinct ability to influence receptor activity (intrinsic efficacy). Agonists are defined as compounds that, upon binding to the receptor, change receptor activity to produce a response. Antagonists are compounds that, although they can bind to the receptor with high affinity, do not produce alterations in receptor response, having no intrinsic efficacy. In biological systems, however, several principles may be considered to define the action of tool compounds. In organism-based research, Chapter 4. 102 the presence of the endogenous ligand of the receptor will produce a baseline response with biological relevance. In the presence of a competitor, as a tool compound, the receptor-ligand equilibrium will be altered by increasing concentrations of the agonist or antagonist, which compete for the same receptor binding site, changing the receptor’s response [28]. In addition, it has also been demonstrated that several receptors can signal even in the absence of a ligand, displaying a constitutive activity, independent of the binding of the endogenous ligand [11; 28]. Therefore, the impact of a compound on the receptor’s response, in biological models, should consider the presence of the endogenous ligand and the presence (or absence) of a constitutive activity of the receptor, thus defining three major categories of tool compounds: a) agonists, b) neutral antagonists and c) inverse agonists. In such scenario, when added to the system, full agonists will elicit the maximum response by the receptor (Figure 1). Furthermore, assuming a receptor that does not have constitutive activity, both neutral antagonists and inverse agonists will compete with the endogenous ligand by the receptor binding site, decreasing the receptor’s response to the same extent (Figure 1A). However, in the presence of a receptor with constitutive activity, the impact of neutral antagonists or inverse agonists will be different. Neutral antagonists decrease the response of the receptor by competition with endogenous ligand, but only inverse agonists will impact the constitutive activity of the receptor, further suppressing its response (Figure 1B) [28]. Figure 1. Schematic representation of the impact of tool compounds on a receptor’s response. In complex biological systems, receptors response is elicited by the endogenous ligand, however some receptors can also present a constitutive activity that influences basal response independently of the endogenous ligand stimulation. With increasing concentrations of full agonists (blue line) receptors response increase to a maximum, while the impact on receptors response by neutral antagonists (red line) or inverse agonists (green line) depends on the presence or absence of receptors’ constitutive activity. (A) In the absence of constitutive activity of the receptor, neutral antagonists and inverse agonists compete by the binding site occupied by the endogenous ligand, that by displacement of endogenous ligand decrease receptors basal Chapter 4. 109 Chapter 4. 110 Figure 3. Chronic administration of antagonists, but not agonists of 5-HT2R/SER-1, 5-HT6R/SER-5 and 5HT7R/SER-7, ameliorates motor dysfunction of ATXN3 proteotoxicity model. Dose-response curves of ATXN3 mutant animals treated with different concentrations of agonists (blue) and antagonists (red). (A) Antagonism of the 5-HT2R/SER-1 positively impacted on motor dysfunction of mutant ATXN3-expressing animals in several concentrations tested, contrary to chronic treatment with an agonist. (B) 5-HT6R/SER-5 and (C) 5-HT7R/SER-7 treatment with agonists and antagonists ameliorated motor dysfunction of ATXN3 proteotoxicity model with an efficacy between 20% to 40%, in the absence of standard dose-response curves. The most efficient concentration of the compounds did not alter the motor behaviour of wild-type (WT) and non-expanded ATXN3 expressing animals (ATXN3 WT). Chronic treatment with antagonists, but not with agonists, of (D) 5-HT2R/SER-1, (E) 5-HT6R/SER-5 and (F) 5-HT7R/SER-7 ameliorated motor dysfunction of ATXN3 mutant animals, using new drug batches. Agonist: blue squares/bars; Antagonists: red squares/bars. Statistical analysis: (A) TCB-2 (n = 3, ± SD) (ANOVA, Dunnett), LY 266097 (n = 4, ± SD) (ANOVA, Dunnett T3); (B) EDMT (n = 3, ± SD) (ANOVA, Games-Howell), SB399885 (n = 3, ± SD) (ANOVA, Sidak test corrected for BCA); (C) LP44 (n = 3, ± SD) (ANOVA, Sidak test corrected for BCA), DR 4485 (n = 3, ± SD) (ANOVA, Sidak test corrected for BCA). (D) (n =3, ± SD), ns – non significant, *P<0.05 (Two-Way ANOVA, Dunnett); (E) (n =3, ± SD), ns – non significant, **P<0.01 (Two-Way ANOVA, Dunnett); (F) (n = 5, ± SD), ns – non significant, *P<0.05 (Two-Way ANOVA, Dunnett). Table 3. Most efficient concentration determined in dose-response curves, of each tool compound, used in the experiments summarized in figure 3. Mammalian receptor C. elegans orthologue Tool compound Concentration 5-HT2B SER-1 LY 266097 0.1 µM 5-HT6 SER-5 EMDT oxalate SB 399885 hydrochloride 50 µM 50 µM 5-HT7 SER-7 LP44 DR 4485 hydrochloride 0.01 µM 10 µM To further explore the role of 5-HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7 in the suppression of ATXN3 proteotoxicity, we used compounds with inverse agonism properties. MDL 100907, SB-742457 and SB-269970 were described as selective receptor blockers, with high potency on receptor binding, displaying inverse-agonism on 5-HT2AR, 5-HT6R and 5-HT7R, respectively [11; 37; 40; 44]. In contrast with what was observed for 5-HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7 antagonists (LY 266097, SB 399885 and DR 4485 respectively), the inverse agonists used failed to impact the motor deficits of the animals at the three concentrations tested (Figure 4). Chapter 4. 111 Figure 4. Chronic administration of 5-HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7 inverse agonists fail to impact motor dysfunction on ATXN3 mutant animals. Egg stage to adulthood treatment of ATXN3 mutant animals with (A) 5-HT2R/SER-1 inverse agonist, MLD 100907, (B) 5HT6R/SER-5 inverse agonist, SB-742457 or (C) 5-HT7R/SER-7 inverse agonist, SB-269970 had no impact on locomotion with 10 µM, 1 µM and 0.1 µM concentration. (A) (n = 4, ± SEM), **P<0.01, ***P<0.001 (ANOVA, Dunnett). (B) (n = 3, ± SEM), ***P<0.001 (ANOVA, Dunnett). (C) (n = 4, ± SEM), ***P<0.001 (ANOVA, Dunnett). 3.3. Acute administration of 5-HT2R/SER-1 antagonist, LY 266097, ameliorates motor dysfunction of ATXN3 mutant nematodes Chronic administration of several drugs can lead to the desensitization of the receptors, as has been previously shown extensively [23; 55; 56; 57; 58; 59]. Therefore, we acutely treated the ATXN3 proteotoxicity model (ATXN3 mutant) with the most efficient concentration of 5-HT2R/SER-1, 5-HT6R/SER5 and 5-HT7R/SER-7 antagonists. The 2 hours treatment with 5-HT6R/SER-5 and 5-HT7R/SER-7 antagonists, SB 399885 and DR 4485, had a limited impact on motor function of mutant ATXN3 animals. However, acute administration of the 5-HT2R/SER-1 antagonist LY 266097 attenuated motor deficits of the ATXN3 mutant animals, with an average efficacy of approximately 30% (Figure 5). Chapter 4. 112 Figure 5. Acute administration of LY 266097, a 5-HT2R/SER-1 antagonist, ameliorated mutant ATXN3mediated motor dysfunction. Acute treatment with LY 266097 decreased mutant ATXN3-mediated locomotion defects, although acute treatment with antagonists of 5-HT6R/SER-5 and 5-HT7R/SER-7 had a limited impact on the suppression of mutant ATXN3-induced motor dysfunction. (n = 3, ± SD) *P<0.05 (ANOVA, Dunnett). 4. Discussion In this chapter, we started to address the role of 5-HTRs in the suppression of ATXN3-mediated proteotoxicity. Our results showed that C. elegans G protein-coupled serotonin receptors were indispensable for a maximal therapeutic action of citalopram (CIT). Teixeira-Castro and colleagues previously demonstrated the dependency of SER-1 and SER-4 to citalopram mediated suppression of mutant ATXN3 proteotoxicity [24]. Here, we suggested additional partial contributions of 5-HT6R/SER-5 and 5-HT7R/SER-7, the nematode-specific ion channel receptor MOD-1 being dispensable for this action. Previous reports described alterations of 5-HTR activity or expression levels upon SSRI treatment. Chronic CIT administration increased the density and activity of 5-HT1AR in the rat hippocampus and decreased 5-HT2R density in the frontal cortex [60; 61], without any major changes in 5-HT4R activity [61]. Regarding 5-HT7R, a decrease in receptor binding in rat hypothalamus upon fluoxetine treatment was reported [62; 63]. However, other studies showed no changes in the expression levels of 5-HTRs upon prolonged SSRI treatment [64]. These apparently conflicting results likely can be explained by the different SSRIs used, treatment duration and brain areas in which the analyses were conducted. We further explored the role of each individual 5-HTR in the suppression of ATXN3-mediated motor dysfunction, using tool compounds and a nematode model, and found that chronic antagonism of Chapter 4. 113 5-HT6R/SER-5 and 5-HT7R/SER-7 ameliorated motor impairment of the animals, without any impact at acute administration. These results suggested that a remodelling of the serotonergic system, elicited by the chronic antagonism of 5-HT6R/SER-5 and 5-HT7R/SER-7 is needed for the therapeutic outcome. Interestingly, both chronic and acute antagonism of 5-HT2R/SER-1 ameliorated motor dysfunction of ATXN3 mutant animals suggesting a more direct contribution of this receptor, probably dependent on its downstream signalling. The modulation of the serotonergic system is often associated with depression, being SSRIs the first line of treatment for this condition [65]. In the depression field, alterations in the serotonergic system, as well as anxiolytic and/or antidepressant amenability of drugs targeting 5-HTRs have been extensively described, solo or in combinatory therapy with SSRIs (reviewed in [66] and [67]). Paradoxically, both agonism and antagonism of 5-HT2CR and 5-HT6R elicited antidepressive like behaviour in mice or rats [68; 69; 70; 71; 72; 73; 74; 75]. This may be due to the double action of these receptors in the brain. On the one hand, stimulation of 5-HT2CR and 5-HT6R (as seen upon agonist treatment) may be an important component of the postsynaptic effects triggered by SSRI treatment. On the other hand, receptor antagonism may facilitate the release of other neurotransmitters, such as norepinephrine and dopamine, that are otherwise physiologically repressed by 5-HT2CR activation, and similarly by 5-HT6R activity [66]. In fact, the anxiolytic and antidepressive-like effects of the 5-HT6R antagonist SB 399885 persisted in the absence of serotonin, favouring the hypothesis of an effect through the modulation of other neurotransmitters [76]. Antagonism of 5-HT2AR and 5-HT7R increased the effect of SSRIs and, alone, reduced depressive-like behaviour in mice. The mechanism through which blockage of 5-HT2AR leads to amelioration of depressive symptoms is largely debated since many compounds, including MDL 100907, which display high affinity to 5-HT2AR, augmented the effect of SSRIs [77]. The mechanism proposed was based on the impact, produced by the blockage of 5-HT2AR, on the dopaminergic system [78]. Interestingly and in line with this hypothesis, co-administration of fluoxetine and MDL 100907 did not further increase serotonin levels in the prefrontal cortex of rats compared with fluoxetine alone, suggesting an impact of 5-HT2AR antagonism outside serotonergic transmission [77]. However less explored, 5-HT7R ablation or blockage with antagonists decreased immobility time in mouse models of depression [79; 80; 81] and potentiated the effect of CIT, increasing 5-HT concentration in the prefrontal cortex [82]. In MJD, we showed that not only CIT action was dependent on 5-HT1AR/SER-4, but also that ablation or desensitization of SER-4 autoreceptor ameliorated ATXN3 proteotoxicity [23; 24], likely by increasing 5-HT availability. Furthermore, the contribution of the post-synaptic 5-HT1AR/SER-4 heteroreceptors in suppression of ATXN3 proteotoxicity was shown, since activation of the SER-4 receptor, in the absence of serotonin, improved motor dysfunction and decreased ATXN3 aggregation [23]. Chapter 4. 114 Therefore, it is plausible that modulation of the activity of this and other 5-HT post-synaptic receptors, such as 5-HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7, could also impact on ATXN3 proteotoxicity by remodelling of the serotonergic system or by changing the release of other neurotransmitters. Additionally, we showed here that inverse agonism of 5-HT2R/SER-1, 5-HT6R/SER-5 and 5HT7R/SER-7 - mediated by MDL 100907, SB-742457 and SB-269970, respectively [36; 47; 51; 83] - was inefficient in the suppression of the motor dysfunction seen in ATXN3 mutant animals, contrasting with the positive effect elicited by the antagonism of the same receptors (by LY 266097, SB 399885 and DR 4485). Keeping in mind that the intrinsic efficacy of an antagonist and an inverse agonist differ in the presence of a constitutive activity of the target receptors [28], our results suggest that a) C. elegans serotonin receptors SER-1, SER-5 and SER-7 can display constitutive activity in vivo , consistent with what was previously reported for SER-7 in vitro [21], and b) only a partial suppression of the activity of these receptors was beneficial for the suppression of ATXN3 proteotoxicity, the constitutive activity of the receptor being needed. However, the unspecific targeting of other receptor types, in C. elegans , by the antagonists LY 266097, SB 399885 and DR 4485, that elicited the positive effect in MJD pathogenesis, cannot be discarded. For that reason, although promising, the results presented here should be interpreted carefully and additional experiments are required to validate this hypothesis. The selected tool compounds used in this study generally presented high affinity and selectivity to mammalian 5-HTRs. In functional studies, EMDT oxalate was demonstrated to behave as a 5-HT6R agonist, with a potency equivalent to serotonin, and a reasonable selectivity profile versus other receptors, having 10to 18-fold more selectivity to 5-HT6R than other 5-HTRs [35; 41]. On the other hand, the 5HT6R antagonist and inverse agonist, SB 399885 and SB-742457, both present high affinity to 5-HT6R receptor, the latter being the most potent (Ki = 0.28 nM) [39; 51]. Although both compounds displayed high selectivity (200-fold) to 5-HT6R compared to other receptors [39], SB-742457 also presented some low affinity to 5-HT2AR (Ki = 26 nM, 90-fold than 5-HT6), and an inverse agonist profile for 5-HT6R [36; 51]. For 5-HT7R, LP44 is a potent receptor agonist with a Ki of 0.22 nM and a EC50 of 2.56 μM. This compound presents a 200-fold higher selectivity to 5-HT7R over 5-HT1AR and 1000-fold over 5-HT2AR [45; 48]. However, the antagonist, DR 4485 presented high affinity to 5-HT7R (Ki = 7.24 nM) but also some low affinity for the 5-HT1AR (Ki = 316 nM, 40-fold than 5-HT7R) [43; 50]. The inverse agonist of 5-HT7R SB269970 [30; 47] presents high potency (Ki = 1.3 nM) and a 250-fold selectivity to 5-HT7R comparing with other receptors, except for 5-HT5AR (50-fold) [37; 46]. Chapter 4. 115 For 5-HT2R, both the antagonist and the inverse agonist compounds used in this study proved to be potent and specific in mammals [42; 49]. LY 266097 possesses high affinity for the 5‐HT2BR and at least 100‐fold selectivity over the 5‐HT2AR and 5‐HT2CR, as well as over numerous other receptors [31; 32]. MDL 100907 displays a Ki of 0.36 nM for 5-HT2AR and at least 200-fold selectivity to this over other receptors, such as 5-HT1AR, 5-HT2CR, D2R or α1R [38]. TCB-2, the 5-HT2AR agonist, presents high potency to the receptor [34], although the pharmacological interactions of this compound with other receptors were never explored deeply (reviewed in [33]) and some affinity to the other 5-HT2 subtypes, at least, cannot be discarded. Therefore, and since C. elegans has a single orthologue for all 5-HT2 subtypes, the SER-1 receptor [52], a specific characterization of the binding and activation profile of this tool compound in the nematode model is needed. As in this work, studies of receptor activity and cellular pathway characterization relies on tool compounds that should be potent and selective for the receptor but also have a proven mechanism of action [27]. Although the antagonists that impact ATXN3 proteotoxicity are potent and specific for 5-HT2R, 5-HT6R and 5-HT7R mammalian receptors (LY 266097, SB-399885 and DR4485 respectively) the specificity of these drugs is model organism-dependent, since its mode of action can vary among species. In the case of DR 4485, the possible effect through its weak affinity to 5-HT1AR could not be excluded in C. elegans . Therefore, in the future further studies are needed to confirm the contribution of 5-HT excitatory receptors to the suppression of ATXN3 proteotoxicity in C. elegans and to determine which molecular pathways mediate the beneficial effect. Future experiments To address these issues, we propose, in the future, to 1) using heterologous expression systems, evaluate the affinity and specificity of the ligation of the tool compounds to C. elegans serotonin receptors; 2) using pharmacogenetic approaches and competition pharmacotherapeutic assays, evaluate the contribution of each receptor-drug pair to the amelioration of ATXN3-mediated motor dysfunction and 3) by qRT-PCR and Western blot analysis or broader transcriptomic/ proteomic analysis techniques explore in vivo the activation of specific pathways upon serotonin receptor activation. To explore the specificity of the ligation of the tool compounds to C. elegans receptors, a thermal stability shift assay conjugated with protein identification can be used . In this technique, drug-treated and vehicle-treated cells are subjected to a series of increasing temperatures. Cells are collected and lysed, the soluble proteome being conjugated with fluorophores to distinguish the treated and non-treated Chapter 4. 116 proteomes. After a 2D gel electrophoresis, proteins that presented a thermal shift, that represents protein stabilization by drug-receptor engagement, will be excised from the gel and identified by massspectrometry [84]. This would allow a full characterization of the possible drug targets on the nematode C. elegans . Additionally, and to determine the association and potency of binding of each drug to the receptors we could use Tag-lite® assay platform (Cisbio), a HTRF based assay [85]. This in vitro system allows not only the study of the engagement of the serotonergic C. elegans receptors with each drug but also the measurement of GPCR activation products in canonical and non-canonical pathways (as Ca+, cAMP, AKT and MAPK phosphorylation, etc). After the validation of the specificity of the tool compounds’ target receptor in C. elegans , the contribution of the receptor activation or inactivation by the drug to the suppression of MJD pathogenesis can be evaluated in vivo by pharmacogenetics or competition pharmacological assays, using specific agonists and antagonists, as previously described [23; 24]. Furthermore, the evaluation of the pathways elicited by drug treatment in vivo can be done by qRT-PCR or Western blot by measuring gene expression changes or protein phosphorylation of key molecules associated to each receptor. With these experiments we will be able to assess drug-receptor specificity and discriminate the biological molecular pathways involved in the suppression of ATXN3 proteotoxicity in C. elegans . Acknowledgments We are grateful to members of the Maciel laboratory for sharing reagents, for critical analysis of the data and results discussion. Thanks to the Caenorhabditis Genetics Center (CGC), which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440) for providing some of the nematode strains used in this study. Funding This work was funded by FEDER, through the Competitiveness Internationalization Operational Programme (POCI), and by National funds, through the Foundation for Science and Technology (FCT), under the scope of the project POCI-01-0145-FEDER-031987, NORTE-010145-FEDER-000013 and NORTE-01-0145-FEDER-000023, supported by the Northern Portugal Regional Operational Programme (NORTE 2020), under the Portugal 2020 Partnership Agreement, through the European Regional Development Fund (ERDF); by National funds, through the Foundation for Science and Technology (FCT) - project UIDB/50026/2020 and UIDP/50026/2020. Additionally, this project was supported by Chapter 4. 117 National Ataxia Foundation (NAF) and the FCT individual fellowship PD/BDE/127834/2016 (For PereiraSousa, J). References [1] D.M. Kuhn, and H. Hasegawa, Chapter 12 - Tryptophan hydroxylase and serotonin synthesis regulation. in: C.P. Müller, and K.A. Cunningham, (Eds.), Handbook of Behavioral Neuroscience, Elsevier, 2020, pp. 239-256.Book Section doi: 10.1016/B978-0-444-64125-0.00012-8. [2] V. Erspamer, and B. Asero, Identification of enteramine, the specific hormone of the enterochromaffin cell system, as 5-hydroxytryptamine. Nature 169 (1952) 800-1. doi: 10.1038/169800b0. [3] P.P. Bertrand, and R.L. Bertrand, Serotonin release and uptake in the gastrointestinal tract. Auton Neurosci 153 (2010) 47-57. doi: 10.1016/j.autneu.2009.08.002. [4] T. Audhya, J.B. Adams, and L. Johansen, Correlation of serotonin levels in CSF, platelets, plasma, and urine. Biochim Biophys Acta 1820 (2012) 1496-501. doi: 10.1016/j.bbagen.2012.05.012. [5] A. Parent, L. Descarries, and A. Beaudet, Organization of ascending serotonin systems in the adult rat brain. A radioautographic study after intraventricular administration of [3H]5hydroxytryptamine. Neuroscience 6 (1981) 115-38. doi: 10.1016/0306-4522(81)90050-6. [6] E.C. Azmitia, and M. Segal, An autoradiographic analysis of the differential ascending projections of the dorsal and median raphe nuclei in the rat. J Comp Neurol 179 (1978) 641-67. doi: 10.1002/cne.901790311. [7] B.L. Jacobs, and E.C. Azmitia, Structure and function of the brain serotonin system. Physiol Rev 72 (1992) 165-229. doi: 10.1152/physrev.1992.72.1.165. [8] T. Sharp, and N.M. Barnes, Central 5-HT receptors and their function; present and future. Neuropharmacology 177 (2020) 108155. doi: 10.1016/j.neuropharm.2020.108155. [9] J. Bockaert, S. Claeysen, A. Dumuis, and P. Marin, CHAPTER 1.5 - Classification and Signaling Characteristics of 5-HT Receptors. in: C.P. Müller, and B.L. Jacobs, (Eds.), Handbook of Behavioral Neuroscience, Elsevier, 2010, pp. 103-121.Book Section doi: 10.1016/S15697339(10)70073-4. [10] M.J. Millan, P. Marin, J. Bockaert, and C. Mannoury la Cour, Signaling at G-protein-coupled serotonin receptors: recent advances and future research directions. Trends Pharmacol Sci 29 (2008) 45464. doi: 10.1016/j.tips.2008.06.007. [11] P. De Deurwaerdere, R. Bharatiya, A. Chagraoui, and G. Di Giovanni, Constitutive activity of 5-HT receptors: Factual analysis. Neuropharmacology 168 (2020) 107967. doi: 10.1016/j.neuropharm.2020.107967. [12] M. Noda, H. Higashida, S. Aoki, and K. Wada, Multiple signal transduction pathways mediated by 5HT receptors. Mol Neurobiol 29 (2004) 31-9. doi: 10.1385/MN:29:1:31. [13] A. Montalbano, R. Corradetti, and B. Mlinar, Pharmacological Characterization of 5-HT1A Autoreceptor-Coupled GIRK Channels in Rat Dorsal Raphe 5-HT Neurons. PLoS One 10 (2015) e0140369. doi: 10.1371/journal.pone.0140369. [14] K.A. Berg, S. Maayani, J. Goldfarb, C. Scaramellini, P. Leff, and W.P. Clarke, Effector pathwaydependent relative efficacy at serotonin type 2A and 2C receptors: evidence for agonist-directed trafficking of receptor stimulus. Mol Pharmacol 54 (1998) 94-104. doi: 10.1124/mol.54.1.94. [15] D. Hoyer, D.E. Clarke, J.R. Fozard, P.R. Hartig, G.R. Martin, E.J. Mylecharane, P.R. Saxena, and P.P. Humphrey, International Union of Pharmacology classification of receptors for 5hydroxytryptamine (Serotonin). Pharmacol Rev 46 (1994) 157-203. http://www.ncbi.nlm.nih.gov/pubmed/7938165 Chapter 4. 118 [16] D. Hoyer, J.P. Hannon, and G.R. Martin, Molecular, pharmacological and functional diversity of 5HT receptors. Pharmacol Biochem Behav 71 (2002) 533-54. doi: 10.1016/s00913057(01)00746-8. [17] D.L. Chase, and M.R. Koelle, Biogenic amine neurotransmitters in C. elegans. WormBook (2007) 115. doi: 10.1895/wormbook.1.132.1. [18] B. Olde, and W.R. McCombie, Molecular cloning and functional expression of a serotonin receptor from Caenorhabditis elegans. Journal of Molecular Neuroscience : MN 8 (1997) 53-62. doi: 10.1007/bf02736863. [19] F.F. Hamdan, M.D. Ungrin, M. Abramovitz, and P. Ribeiro, Characterization of a novel serotonin receptor from Caenorhabditis elegans: cloning and expression of two splice variants. J Neurochem 72 (1999) 1372-83. doi: 10.1046/j.1471-4159.1999.721372.x. [20] G.P. Harris, V.M. Hapiak, R.T. Wragg, S.B. Miller, L.J. Hughes, R.J. Hobson, R. Steven, B. Bamber, and R.W. Komuniecki, Three distinct amine receptors operating at different levels within the locomotory circuit are each essential for the serotonergic modulation of chemosensation in Caenorhabditis elegans. J Neurosci 29 (2009) 1446-56. doi: 10.1523/JNEUROSCI.458508.2009. [21] R.J. Hobson, J. Geng, A.D. Gray, and R.W. Komuniecki, SER-7b, a constitutively active Galphas coupled 5-HT7-like receptor expressed in the Caenorhabditis elegans M4 pharyngeal motorneuron. J Neurochem 87 (2003) 22-9. doi: 10.1046/j.1471-4159.2003.01967.x. [22] R. Ranganathan, S.C. Cannon, and H.R. Horvitz, MOD-1 is a serotonin-gated chloride channel that modulates locomotory behaviour in C. elegans. Nature 408 (2000) 470-5. doi: 10.1038/35044083. [23] J. Pereira-Sousa, B. Ferreira-Lomba, A. Bellver-Sanchis, D. Vilasboas-Campos, J.H. Fernandes, M.D. Costa, M.A. Varney, A. Newman-Tancredi, P. Maciel, and A. Teixeira-Castro, Identification of the 5-HT1A serotonin receptor as a novel therapeutic target in a C. elegans model of Machado-Joseph disease. Neurobiol Dis (2021) 105278. doi: 10.1016/j.nbd.2021.105278. [24] A. Teixeira-Castro, A. Jalles, S. Esteves, S. Kang, L. da Silva Santos, A. Silva-Fernandes, M.F. Neto, R.M. Brielmann, C. Bessa, S. Duarte-Silva, A. Miranda, S. Oliveira, A. Neves-Carvalho, J. Bessa, T. Summavielle, R.B. Silverman, P. Oliveira, R.I. Morimoto, and P. Maciel, Serotonergic signalling suppresses ataxin 3 aggregation and neurotoxicity in animal models of Machado-Joseph disease. Brain 138 (2015) 3221-37. doi: 10.1093/brain/awv262. [25] F.C. Colpaert, J.P. Tarayre, W. Koek, P.J. Pauwels, L. Bardin, X.J. Xu, Z. Wiesenfeld-Hallin, C. Cosi, E. Carilla-Durand, M.B. Assie, and B. Vacher, Large-amplitude 5-HT1A receptor activation: a new mechanism of profound, central analgesia. Neuropharmacology 43 (2002) 945-58. doi: 10.1016/s0028-3908(02)00119-3. [26] A. Newman-Tancredi, J.C. Martel, C. Cosi, P. Heusler, F. Lestienne, M.A. Varney, and D. Cussac, Distinctive in vitro signal transduction profile of NLX-112, a potent and efficacious serotonin 5HT1A receptor agonist. J Pharm Pharmacol 69 (2017) 1178-1190. doi: 10.1111/jphp.12762. [27] C.H. Arrowsmith, J.E. Audia, C. Austin, J. Baell, J. Bennett, J. Blagg, C. Bountra, P.E. Brennan, P.J. Brown, M.E. Bunnage, C. Buser-Doepner, R.M. Campbell, A.J. Carter, P. Cohen, R.A. Copeland, B. Cravatt, J.L. Dahlin, D. Dhanak, A.M. Edwards, M. Frederiksen, S.V. Frye, N. Gray, C.E. Grimshaw, D. Hepworth, T. Howe, K.V. Huber, J. Jin, S. Knapp, J.D. Kotz, R.G. Kruger, D. Lowe, M.M. Mader, B. Marsden, A. Mueller-Fahrnow, S. Muller, R.C. O'Hagan, J.P. Overington, D.R. Owen, S.H. Rosenberg, B. Roth, R. Ross, M. Schapira, S.L. Schreiber, B. Shoichet, M. Sundstrom, G. Superti-Furga, J. Taunton, L. Toledo-Sherman, C. Walpole, M.A. Walters, T.M. Willson, P. Workman, R.N. Young, and W.J. Zuercher, The promise and peril of chemical probes. Nat Chem Biol 11 (2015) 536-41. doi: 10.1038/nchembio.1867. Chapter 4. 125 Supplementary table 1. Statistical report. Effect size calculated using Uanhoro, J. O. (2017). Effect size calculators. Available online at: https://effect-sizecalculator.herokuapp.com/. Figure Statistical report Sample size Fig. 2 A F(19, 29.771) = 24.717, p < 0.001, ω2p = 0.900 4 B F(19, 14.324) = 20.101, p < 0.001, ω2p = 0.914 3 C F(19, 18.523) = 14.780, p < 0.001, ω2p = 0.872 3 Fig. 3 A TCB-2 F(8, 18) = 2.830, p = 0.0318, ω2p = 0.351 3 LY 266097 F(11, 16.301) = 6.764, p > 0.001, ω2p = 0.691 4 B EDMT F(10, 22) = 5.807, p > 0.001, ω2p = 0.593 3 SB 399885 F(11, 16.337) = 10.266, p > 0.001, ω2p = 0.782 3 C LP 44 F(9, 11.119) = 8.148, p = 0.001, ω2p = 0.753 3 DR 4485 F(8, 12.252) = 14.780, p = 0.008, ω2p = 0.838 3 D F (4, 23) = 1.414; p = 0.2608, ω2p = 0.056 3 E F (4, 22) = 4.133, p = 0.0120; ω2p = 0.317 3 F F (4, 24) = 1.327, p = 0.2885, ω2p = 0.043 3 - 5 Fig. 4 A F (9, 30) = 8.925, p < 0.001, ω2p = 0.641 4 B F (9, 20) = 15.70, p < 0.001, ω2p = 0.815 3 C F (9, 30) = 9.827, p < 0.001, ω2p = 0.665 4 Fig. 5 F (7, 16) = 2.553, p = 0.0571, ω2p = 0.312 3 126 Chapter 5. General discussion and conclusions Chapter 5. 127 Discussion and future perspectives Machado-Joseph disease (MJD) is one of the so called proteinopathies [1]. This group of diseases, that comprises Parkinson’s, Alzheimer’s and Huntington’s diseases, among others, have protein aggregation as a common hallmark. Although the genetic alterations underlying each disease are different, this common feature establishes the study of protein misfolding and aggregation, and the impact of this phenomenon in cells and organisms, as a central process. In MJD, the abnormal expansion of the CAG-repeat in the ataxin-3 gene translates into a pathological polyglutamine (polyQ) tract that prone ataxin-3 protein (ATXN3) to form aggregates. Several transgenic model organisms expressing the expanded human protein mimic the hallmarks of the disease, including motor impairment and protein aggregation, which correlate with the CAG repeat length and age at onset, as in the human disease [2; 3; 4]. Over the last decade, several efforts have been conducted to unravel the disease mechanism(s) and to find a cure for this untreatable pathology. One of the potential therapeutical approaches discovered was the modulation of serotonergic neurotransmission. Hypothesis-free approaches based on the screening of commercially available drugs showed modulators of serotonergic signalling such as antidepressants (citalopram), antipsychotics (aripiprazole) or anxiolytic drugs (buspirone and tandospirone) as candidate therapeutic agents, having beneficial impact in several dimensions of MJD pathogenesis [5; 6; 7; 8; 9; 10; 11; 12]. However, the exact mechanism by which these drugs impact on MJD is mostly unknown. In this thesis, we contributed to further deepen the understanding of how modulation of the serotonergic system suppresses MJD pathogenesis, assessing its impact on mutant ATXN3-mediated motor dysfunction and aggregation in C. elegans [4]. First, we further developed the existing biochemical techniques filter retardation assay and biochemical fractionation and applied them to a C. elegans model of mutant ATXN3 proteotoxicity, allowing a better assessment of ATXN3 aggregation states in this model (Chapter 2). Using this and other approaches, we explored the contribution of the different serotonin receptors to the suppression of MJD pathogenesis (Chapter 3 and 4). Our results suggest that specific targeting of individual 5-HT receptors, either by the activation of SER-4/5-HT1AR (Chapter 3), or by the chronic antagonism of SER-1/5-HT2R, SER-5/5-HT6R or SER-7/5HT7R (Chapter 4) is beneficial in the context of MJD pathogenesis. Our work thus opens new avenues in the study of the impact of the modulation of serotonergic system in MJD pathogenesis, including the Chapter 5. 128 determination of the specific pathways by which each individual serotonin receptor impacts on ATXN3 proteotoxicity, raising a number of new questions that will be discussed below. Possible serotonergic system deregulation in MJD Although an extensive study of the serotonergic system has never been performed in the context of MJD, the amino acid L-tryptophan, the precursor of serotonin (5-HT), was found to be decreased in the serum of patients [13] and in a transgenic mouse model of the disease [14]. Importantly, L-tryptophan integrated a panel of four serum metabolites that confidently detected the disease state, discriminating MJD patients from controls, being therefore proposed as a promising biomarker [13]. Tryptophan is an essential component of the human diet but, since its storage is low, it is found in low concentrations in the body. This amino acid is involved not only in protein biogenesis but also in the 5-HT and kynurenine biosynthetic pathways [15]. Interestingly, metabolites of the kynurenine pathway have been proposed as novel prognostic markers and/or therapeutic targets for other neurodegenerative diseases, such as Parkinson’s disease (PD) and Huntington’s disease [16; 17; 18; 19; 20; 21]. Tryptophan is the precursor of 5-HT, and in some tissues, such as the pineal gland, 5-HT is further metabolized to melatonin [15]. Although to our knowledge, 5-HT levels were never measured in the serum or cerebrospinal fluid of MJD patients, a study from Takei and colleagues showed a negative correlation between serum melatonin levels and insomnia in patients [22]. This observation suggested that not only tryptophan may be decreased in MJD, but also other tryptophan metabolites, such as kynurenine and serotonin could be altered. MJD patients typically present an atrophy of the cerebellum, pons, and medulla oblongata, and neuropathological studies reveal neuronal loss in the cerebellar dentate nuclei, pons, substantia nigra, thalamus, globus pallidus and in the brainstem nuclei, among others [23; 24]. Volumetric analyses by MRI of these patients demonstrated atrophy of the cerebellum, brainstem, caudate nuclei and thalamus [23; 25; 26]. Importantly, some of these brain regions were also shown to be involved in the pathophysiology of depression, including the thalamus, basal ganglia and brainstem [27; 28]. Brainstem can be considered the nuclei of neurotransmitters. This brain structure comprises the raphe nucleus - that contain serotonergic neurons, the locus coeruleus - with norepinephrinergic neurons, and the substantia nigra and ventral tegmentum - with dopaminergic neurons [27]. Therefore, it is possible that brainstem degeneration, in MJD patients, can contribute to imbalances in neurotransmission, namely by impacting serotonin production. Serotonergic signalling is involved in depression or depressive symptoms Chapter 5. 129 [29; 30]. Indeed, SSRIs are the most common used pharmacological therapies for depression. Interestingly, MJD patients present a higher risk to develop depression than the general population, depression being an established comorbidity in MJD. The inverse correlation between motor performance of these patients and the development of depression was extensively reported, depressive symptoms increasing with disease progression [31; 32; 33; 34; 35]. This has been attributed to the negative psychological impact of the disease. However, it has been proposed that depression could be a primary effect of the neurodegeneration process also [36; 37]. Recent studies in the depression field suggested a more complicated origin of the depressive pathology, involving multiple pathways, among which an imbalance in kynurenine/serotonergic pathways [38; 39; 40; 41; 42]. Even though MJD patients show an increased susceptibility to depression, and a possible impairment in tryptophan levels or in its metabolites, MJD mouse models did not present decreased serotonin levels, at least in the brain regions analysed [11; 43], neither a major anxiousor depressive-like behaviour when assessed postsymptomatically, at 16 weeks of age (Correia J., personal communication). Serotonergic signalling dysfunction was reported in other neurodegenerative disorders, such as PD [44], AD [45; 46] and ALS [47]. Low SERT binding in the brain limbic regions of AD patients [48] and decreased expression of the 5-HT1AR were reported [49]. In the striatum of MJD transgenic mice, the RNA expression profile analysis by RNA sequencing, performed by our laboratory, also showed a downregulation of 5-HT1AR heteroreceptor (Oliveira S., personal communication), without any other major alterations in expression of genes encoding enzymes that play a role in the biosynthetic pathways of serotonin, kynurenine, or melatonin. In summary, more studies with the evaluation of the levels of neurotransmitter and their metabolites, and to dissect serotonergic signalling dysfunction are needed in MJD, particularly in patients. The role of 5-HT receptors on the suppression of MJD pathogenesis As referred above, serotonergic signalling modulation by chronic citalopram (CIT) administration suppressed MJD pathogenesis [11]. We showed that CIT treatment of two different MJD animal models ameliorated motor impairment and suppressed mutant ATXN3 aggregation, and more importantly, that these effects persisted after drug withdrawal [11]. The validation of these findings by an independent laboratory using a distinct transgenic mouse model [5] further supported the theory of serotonergic signalling modulation as a promising therapy for MJD. Chapter 5. 130 In this thesis, we further supported these findings by establishing the 5-HT1AR as a new therapeutic target for MJD (Chapter 3, published as [50]). By exploring the effect of befiradol, a specific and potent 5-HT1AR agonist, we found that both chronic and acute administration of the compound ameliorated core features of the disease, including motor impairment and mutant ATXN3 aggregation. Importantly, and as seen for CIT administration, the beneficial effects of befiradol persisted after drug withdrawal [50], suggesting a disease modifying effect of serotonergic signalling modulation. Recent studies in our laboratory are showing promising results of befiradol treatment in MJD transgenic mice, in both chronic or sub-chronic treatment modalities (Ferreira-Lomba B. and Duarte-Silva S., personal communication). Furthermore, our results showed that befiradol’s effect on mutant ATXN3-mediated motor dysfunction was dependent specifically on the 5-HT1AR orthologue in C. elegans SER-4, highlighting the specificity of compound binding in C. elegans [50]. 5-HT1ARs/SER-4 are present in the brain forming two distinct populations: i) the post-synaptic somatodendritic heteroreceptors expressed at 5-HT projecting areas [51; 52] and ii) the pre-synaptic autoreceptors at the cell body and dendrites of 5-HT producing neurons, namely of the dorsal raphe nuclei (DRN), in the case of mice and humans. Serotonin released from the DRN activates 5-HT1A autoreceptors, exerting a feedback inhibition on the firing of 5-HT neurons, decreasing 5-HT production and negatively regulating the 5-HT system [53; 54; 55; 56]. Our results showed that both populations of 5-HT1A autoand heteroreceptors play a role in the suppression of ATXN3 proteotoxicity. First, we observed that in the absence of endogenous serotonin, caused by a full deletion of the tph-1 gene generated by CRISPR-Cas9 technology in the ATXN3-proteotoxicity model, acute treatment with befiradol still ameliorated motor dysfunction of the animals, suggesting the contribution of 5-HT1A heteroreceptors. Interestingly, chronic treatment with befiradol failed to ameliorate motor dysfunction in the absence of serotonin, highlighting a main role for 5-HT autoreceptor desensitization during chronic treatment. Consistently with this, by assessing neuronal activity of a serotonin-producing neuron, by ratiometric calcium imaging in live animals, we found a distinct activity profile between the two treatment modalities. Befiradol acute treatment displaying a neuronal inhibitory profile, comparing with chronic treatment and WT conditions, consistent with SER-4 activation. Taken together, our data led us to propose a model for befiradol’s action through the desensitization of SER-4 autoreceptors and activation of SER-4 heteroreceptors expressed in serotonergic and non-serotonergic neurons, respectively [50]. The exact mechanism by which 5-HT1AR/SER-4 desensitization impacts on MJD pathogenesis remains to be fully determined, however the desensitization of 5-HT1AR has also been associated to SSRI’s therapeutic action. Inhibition of SERT by treatment with SSRIs, likely increases 5-HT availability, with consequent activation of 5-HT1A autoreceptors and the negative feedback against 5-HT production [53; Chapter 5. 131 54; 55; 56], which was suggested to explain the delay in time to first response in the treatment of depression [57]. Upon chronic administration of SSRIs, there is a downregulation of the 5-HT1A autoreceptors, and the subsequent increase in neuronal firing rates [56; 58]. Therefore, it is likely that post-synaptic 5-HT heteroreceptors mediate serotonergic signalling propagation, especially after 5-HT1AR autoreceptor desensitization, as proposed for CIT’s effect on the suppression of MJD pathogenesis. Further exploration of the contribution of the post-synaptic 5-HTRs to the suppression of MJD pathogenesis suggested a potential common involvement of all 5-HT GPCRs (Chapter 4). CIT action showed to be dependent on all 5-HT GPCRs in C. elegans , namely of SER-1/5-HT2R and SER-4/5-HT1AR [11], SER-5/5-HT6R and SER-7/5-HT7R, being likely independent of the ion-channel receptor MOD-1. Surprisingly, the antagonism of SER-1/5-HT2R, SER-5/5-HT6R and SER-7/5-HT7R, but not the inverse agonism of these receptors, were beneficial to motor behaviour of the animals. Antagonists and inverse agonists drug action differ from each other by its intrinsic efficacy. In complex systems, with the presence of the endogenous ligand and when the receptor has constitutive activity, antagonists impact on receptor’s response by competing with endogenous ligand, while inverse agonists further reduce the receptor’s response by the suppression of its constitutive activity [59]. This observation, together with our preliminary findings, suggest that there is an optimal level of activity for 5-HTRs and that their signalling should be tightly controlled/regulated to achieve the beneficial effect in the context of MJD. Interestingly, ERK signalling is a common non-canonical pathway to all 5-HTRs described in our work, 5-HT1AR, 5-HT2R, 5HT6R and 5-HT7R [60; 61]. Long lasting neuronal ERK activation was found associated to neurodegenerative disorders, such as PD and AD [62; 63; 64; 65], and ERK signalling inhibitors were suggested as therapeutic strategies to these diseases [66]. Interestingly, it was shown that ERK1/2 pathway can regulate pathogenic ATXN3 protein levels in cells [67] the ERK signalling status being changed in MJD models [68]. Therefore, in the future we intend to assess the profile of proteins that are involved in ERK signalling upon CIT and befiradol treatments. In order to address if any of the 5-HTRs were more important to the therapeutic effect of SSRIs in depression, Carr and Lucki, in a review paper, analysed behavioural pharmacology studies, selecting 5-HTRs target drugs that suppressed the effect of SSRIs or mimicked antidepressive-like effects in models of depression. They concluded that either agonism of 5-HT1AR, 5-HT1BR, 5-HT2CR, 5-HT4R, and 5-HT6R or antagonism of 5-HT2AR, 5-HT2CR, 5-HT3R, 5-HT6R, and 5-HT7R could produce antidepressive-like behaviours raising the hypothesis that none of the receptors exerted a prominent role in the action of SSRIs, being the beneficial effect probably a result of an orchestration of stimulation and blockage signals of the Chapter 5. 132 different 5-HTRs [69]. Moreover, and aside the redundancy and apparent antagonic capacity of serotoninergic system, some receptors were indeed shown to be modulated, either in expression levels or activation status, by CIT administration [70; 71; 72]. Overall, our findings in MJD resemble what has been previous described in major depression disorders, suggesting that in spite of quite different outcome measures a common mechanism may underlie the therapeutic effects of 5-HT targeting drugs in neurodegenerative and neuropsychiatric disorders. As previously referred, 5-HT1A/SER-4 heteroreceptor agonism [50] and, possibly, acute antagonism of 5-HT2R/SER-1 [Chapter 4] positively impacted on motor impairment of mutant ATXN3expressing animals. CIT action was previously described to modulate these two receptors. On the one hand, as explained above the generally for SSRIs, short time treatment with CIT leads to activation of 5HT1A autoreceptors at DRN, that after a given time threshold become desensitized, with subsequent induction of the firing of serotonergic neurons [71; 72]. However, the desensitization of 5-HT1AR, upon chronic CIT treatment, occurs preferentially at the DRN autoreceptors but not in 5-HT1A heteroreceptors, at least at the hippocampus region, where, at the contrary, the responsivity of 5-HT1AR was found increased [70; 72]. This observation was also verified for another SSRI, paroxetine [73]. The desensitization of 5HT1AR being needed for SSRIs action, it has been proposed that a co-treatment with a 5-HT1AR antagonist would lead to a better outcome in depressive patients [74]. However, clinical trials employing this strategy of co-treatment failed to produce an enhancement in the effect of the SSRI alone [75; 76]. The justification given for this fact was its action on the 5-HT1A heteroreceptors of the forebrain, that, upon treatment with the antagonist, could impact negatively the antidepressive response, regardless of the increased serotonin levels [76]. On the other hand, this study also shows that the positive effect of the SSRIs is maintained in the presence of 5-HT1AR antagonist, suggesting the contribution of other serotonin receptors (other than the 5-HT1AR) in SSRI action [76]. The increased responsivity of the 5-HT1A heteroreceptor was found to be accompanied by a decrease in 5-HT2R activation status or binding sites upon CIT treatment [70; 71]. Also, locomotion activity in mice exposed to a novel environment showed that CIT action was dependent on 5-HT1BR and 5-HT2AR, the response to this SSRI being augmented by 5-HT2CR antagonist SB242084 [77]. In mice that were nonresponders to SSRI treatment, the immobility in forced swimming test was also improved by the antagonism of 5-HT2CR [78], the effect being attributed to the increased serotonin levels in pre-frontal cortex. This increased in serotonin availability was also found in the hippocampus of freely moving rats when treated with SB242084 prior to CIT administration [79]. Interestingly, buspirone, a 5-HT1AR partial Chapter 5. 133 agonist, alone or in combination with CIT, could mediate the alteration of highto lowaffinity status of 5HT2CR in rats [80]. In MJD, we found that CIT chronic treatment decreased the expression levels of 5HT2AR in the striatum, but the impact on receptor activity or binding affinity remains to be elucidated (Oliveira S., personal communication). Less is known regarding the impact of CIT treatment on other 5-HTRs. In rats, co-treatment with CIT and the 5-HT7R antagonist SB269970 augmented serotonin levels in the pre-frontal cortex and culminated in an enhancement of the antidepressive-like behaviour observed, when compared with the single therapies alone [81; 82]. Interestingly, the impact of 5-HT7R antagonism on serotonin levels has been proposed as a cause for the improvement of depression symptoms in animal models, but the 5HT7R desensitization was also suggested to play a role, at least in part, in the mode of action of CIT [83; 84; 85] and, perhaps of other SSRIs, since chronic treatment with fluoxetine and 6-OH-DPAT (a 5-HT1AR agonist) decreased the number of 5-HT7R binding sites in the rats hypothalamus [83; 85]. To our knowledge, there are no descriptions of SSRI-mediated modulation of the 5-HT6 receptor. However, several reports described the antidepressive effects of both 5-HT6R agonists and antagonists [86; 87; 88; 89; 90]. Despite the fact that the mechanism of action of these drugs in depression is also mostly unknown, it was shown that 5-HT6R inhibition, in GABAergic neurons, inhibited GABA release and therefore increased cholinergic, glutamatergic and monoaminergic neurotransmission [91]. Thus, it is hypothesized that the antidepressive properties of agonists and antagonists of this receptor could depend on receptor localization in different neuronal types, through the modulation of other neurotransmitters as acetylcholine, GABA and glutamate [92]. In conclusion, the serotonergic system encompasses a highly complex and robust network, that upon distinct pharmacological or genetical modulation can result in similar biologic outcomes/phenotypes, 5-HTRs possibly being the most relevant receptors for the mode-of-action of SSRIs, albeit the effect of these drugs is greatly dependent on cellular types and treatment modalities. Nevertheless, our results support the hypothesis, previously suggested by others [69; 93], that the increased serotonin levels by chronic CIT treatment, in a disease organism, can lead to a remodelling of serotonin receptors response and to a myriad of stimulations and inhibitions, namely through the desensitization of 5-HT1AR/SER-4 autoreceptors, with augmentation of 5-HT1AR/SER-4 heteroreceptors response and by decreasing 5-HT2R/SER-1, 5-HT6R/SER-5 and 5-HT7R/SER-7 activity (Figure 1). The usage of multimodal drugs that target multiple 5-HTRs, as aripiprazole (APZ) or vortioxetine, helped to further validate this model. We showed that APZ chronic treatment ameliorated motor dysfunction of Chapter 5. 134 mutant ATXN3 expressing animals the effect of the drug being dependent mainly on 5-HT1AR/SER-4, 5HT2R/SER-1 and dopamine D2 receptors in C. elegans (Jalles, A. and Vieira, C. – under revision). Importantly, preliminary results showed a positive impact of vortioxetine on animals’ motor dysfunction (Freitas, B. personal communication), this drug combining SERT/MOD-5 inhibition with antagonism of 5HT3R, 5-HT1DR and 5-HT7R, agonism of 5-HT1AR and partial agonism of 5-HT1BR [94; 95; 96]. Figure 1. Hypothetical model of the beneficial effect induced by the remodelling of serotonergic signalling in the suppression of mutant ATXN3 proteotoxicity. CIT administration, by the blockage of the serotonin transporter MOD-5, likely increases serotonin availability in the synaptic cleft and induces activation of the SER-4 autoreceptor. Prolonged activation of this autoreceptor likely results in its desensitization, decreasing the inhibitory signal exerted by SER-4 on 5-HT production. These actions on the pre-synaptic serotonergic neuron, may lead to a remodelling of the serotonergic system, impacting post-synaptic 5-HTR-mediated signalling. Agonism of SER-4 and/or antagonism of SER-1, SER-5 and SER-7 was beneficial to the suppression of ATXN3 proteotoxicity, however the downstream signalling pathways involved in 5-HTRs-mediated effects were not yet elucidated [11; 50] (Jalles, A. and Vieira, C. – under revision). In spite of the clear need for further confirmatory studies and to dissect the molecular pathways involved in the suppression of MJD pathogenesis by 5-HTRs-mediated signalling, with this work, we proposed a new therapeutic target in MJD, the 5-HT1AR. Moreover, we strengthened the finding of a possible role for other receptor types, such as 5-HT2R, 5-HT6R and 5-HT7R, which can also constitute valuable targets, to be validated in the future, using a single molecule treatment or in combinatory therapies for MJD.