Rozprawy doktorskie na temat „DisA protein”

Understanding how proteins fold and bind is interesting since these processes are central to most biological activity. Protein folding and protein-protein interaction are by themselves very complex but using a good and robust system to study them could ease some of the hurdles. In this thesis I have tried to answer some of the fundamental questions of protein folding and binding. I chose to work with PDZ domains, which are protein domains consisting of 90-100 amino acids. They are found in more than 400 human proteins and function mostly as protein-protein interaction units. These proteins are very stable, easy to express and purify and their folding reaction is reversible under most laboratory conditions. I have characterized the interaction of PSD-95 PDZ3 domain with its putative ligand under different experimental conditions and found out that its binding kinetics is sensitive to salt and pH.  I also demonstrated that the two conserved residues R318 and H372 in PDZ3 are responsible for the salt and pH effect, respectively, on the binding reaction. Moreover, I determined that for PSD 95 PDZ3 coupling of distal residues to peptide binding was better described by a distance relationship and there was a very weak evidence of an allosteric network. Further, I showed that another PDZ domain, SAP97 PDZ2 undergoes conformational change upon ligand binding. Also, I characterized the binding mechanism of a dimeirc ligand/PDZ1-2 tandem interaction and showed that despite its apparent complexity the binding reaction is best described by a square scheme. Additionally, I determined that for the SAP 97 PDZ/HPV E6 interaction that all three PDZ domains each bind one molecule of the E6 protein and that a set of residues in the PDZ2 of SAP 97 could operate in an unexpected long-range manner during E6 interaction. Finally, I showed that perhaps all members in the PDZ family could fold via a three state folding mechanism. I characterized the folding mechanism of five different PDZ domains having similar overall fold but different primary structure and the results indicate that all five fold via an intermediate with two transition states. Transition state one is rate limiting at low denaturant concentration and vice versa for transition state two. Comparing and characterizing the structures of the transition states of two PDZ domains using phi value analysis indicated that their early transition states are less similar as compared to their late transition states.

The Z-disc of striated muscle cells is a highly specialized three-dimensional structure which delineates the boundary of the individual sarcomeres. It accomplishes a unique role by anchoring actin filaments and acts as a molecular trigger for contraction. Beyond a well-defined structural role, in recent years it is emerging the hypothesis that Z-disc may be directly involved in the perception and transmission of muscular stress signals. To achieve these complex functions, many Z-disc proteins are involved in multiple protein interactions. The importance of these interactions is indicated by the fact that mutations in several Z-disc proteins can result in muscular dystrophies and/or cardiomyopathies in human and mice. The knowledge of Z-disc interactome and its regulation would improve by far the comprehension of the Z-disc biology and the onset of muscular disorders. The main goal of my project was to understand the complex network of protein-protein interactions occurring at the Z-disc of skeletal and cardiac muscle. In particular, my work was focused on two groups of Z-disc proteins: the FATZ and myotilin protein families on one hand, and some proteins belonging to the enigma family on the other hand. This work led to the identification of a specific interaction between the PDZ domains of enigma family members and the C-terminal five amino acids of the FATZ and myotilin families. The work of this thesis is part of a wider project involving the groups of Dr. G. Faulkner at ICGEB, Trieste, and Prof. O. Carpen at University of Turku, Finland. Together with our collaborators we noted that the C-terminal five amino acids of FATZ-1 (ETEEL), FATZ-2 (ESEDL), FATZ-3 (ESEEL), myotilin (ESEEL), palladin (ESEDL) and myopalladin (ESDEL) are highly similar. Searches in protein sequence database revealed that this E-[S/T]-[D/E]-[D/E]-L motif is restricted in Vertebrates to the FATZ and myotilin families of proteins, and it is evolutionary conserved from zebrafish to humans, indicating its importance for their biological function. The ELM program (a source for predicting functional sites in eukaryotic proteins) predicted that the terminal four amino acids of the FATZ family, myotilin, palladin and myopalladin constitute a binding motif for class III PDZ domain proteins (X-[D/E]-X-[V/I/L]). The first object of my work was to determine if the proteins with this new type of class III PDZ binding motif at their C-terminal could effectively bind PDZ domains. We knew from the literature that ZASP binds to all the three members of the FATZ family by means of its N-terminal PDZ domain, and that the C-terminal region of myotilin interacts with ZASP. In addition to ZASP, other two members of the enigma family of PDZ proteins, ALP and CLP-36, were included in this study. Both the full-length and the truncated (lacking the last five amino acids) version of the FATZ and myotilin families were produced as native proteins and tested for PDZ binding using the AlphaScreen (Amplified Luminescence Proximity Homogeneous Assay) technique. Biotinylated phosphorylated and non-phosphorylated peptides corresponding to the C-terminal five amino acids of the FATZ family, myotilin, palladin and myopalladin were also used in AlphaScreen interaction experiments, as well as a control peptide having E instead of L as its last amino acid (ESEEE). The results presented in this thesis show that the final five amino acids of the FATZ and myotilin families of proteins are responsible for the binding to the PDZ domains of ZASP, ALP and CLP-36, and that the nature of the last amino acid of the motif is crucial for the interaction. We also show that phosphorylation of the ligand sequence modulates the ability of the peptides to bind to the PDZ domains of the enigma family. ?-actinin-2 was included in this study as its C-terminus (GESDL) is classified as a class I PDZ binding motif that is able to bind to ZASP and ALP PDZ domains. AlphaScreen experiments confirm the binding of both the full-length and the C-terminal phosphorylated and non-phosphorylated peptides of ?-actinin-2 to the PDZ domains of ZASP and ALP, and they also reveal an interaction with the PDZ domain of CLP-36. These interactions were verified using another in vitro binding technique, the TranSignal PDZ Domain Array. Based on the results of the PDZ arrays, RIL was found to be another member of the enigma family capable to bind to the E-[S/T]-[D/E]-[D/E]-L motif. Therefore, these final five amino acids can be considered a novel type of class III PDZ binding motif specific for the PDZ domains of enigma proteins. To better quantify the strength of the noted interactions, SPR (Surface Plasmon Resonance) experiments were performed in the laboratory of Dr. A. Baines at University of Kent, UK. The affinities of the interactions between the PDZ domain of ZASP and some of the phosphorylated and non-phosphorylated peptides of the FATZ and myotilin families result to be in the nM range. The SPR results also demonstrate a new interaction between the PDZ domain of ZASP and ANKRD2. This protein is a member of the MARP family and it is thought to be involved in muscle stress response pathways. ANKRD2 localizes both in the sarcomeric I-band and the nucleus, and it is able to bind to several transcription factors, including YB-1, PML and p53. This interaction strengthens the hypothesis that, besides a structural function, Z-disc could have a role in cell signalling. The fact that at the Z-disc many proteins can interact with the same partners, it would be helpful to define the pattern and level of expression of the individual proteins in different muscle tissues. Another aim of my work was to measure the abundance of mRNAs of some Z-disc proteins using the Real-Time PCR technique. Four different muscles from adult mice were considered: tibialis (a fast-twitch skeletal muscle), soleus (a slow-twitch skeletal muscle), gastrocnemius (a skeletal muscle with mixed fibers) and heart (cardiac muscle). The different distribution of the FATZ proteins, myotilin and the alternatively spliced variants of ZASP suggest that, at least in mouse, the interactions between these proteins could be compartmentalized in distinct fiber types.
Il disco-Z del muscolo striato è una struttura molecolare altamente specializzata a livello della quale si instaurano numerose interazioni proteina-proteina. Il disco-Z delinea il confine dei singoli sarcomeri, fornendo un punto di ancoraggio per i filamenti sottili di actina; il loro scorrimento sui filamenti spessi di miosina produce la forza meccanica responsabile della contrazione. Uno dei ruoli chiave del disco-Z, dunque, è quello di trasmettere la tensione generata dalla struttura seriale dei sarcomeri lungo le miofibrille e, di conseguenza, lungo tutto il muscolo. Al di là di un evidente significato strutturale, negli ultimi anni sta diventando sempre più consistente l’ipotesi di un suo coinvolgimento anche nella percezione e nella trasmissione di segnali. L’importanza delle interazioni tra le proteine del disco-Z è indicata dal fatto che mutazioni in molte di queste proteine possono risultare in distrofie muscolari e/o cardiomiopatie sia in uomo sia in topo. Una più ampia conoscenza delle interazioni che si articolano a livello del disco-Z e, in generale, degli eventi che le regolano, aiuterebbe a chiarire la biologia del disco-Z e l’insorgenza di eventuali patologie associate. Il mio progetto di Dottorato è stato incentrato su due gruppi di proteine sarcomeriche e sulle loro interazioni: le proteine delle famiglie FATZ e miotilina da un lato, e alcune proteine appartenenti alla famiglia enigma dall’altro. Questo lavoro ha portato all’identificazione di un’interazione specifica tra i domini PDZ delle proteine della famiglia enigma e gli ultimi cinque residui aminoacidici presenti nelle proteine delle famiglie FATZ e miotilina. Il lavoro di questa tesi fa parte di un progetto più ampio che coinvolge i gruppi coordinati dalla Dr.ssa G. Faulkner dell’ICGEB, Trieste, e il Prof. O. Carpen dell’Università di Turku, Finlandia. Grazie alla loro collaborazione, è stato possibile notare che i cinque residui C-terminali delle proteine FATZ-1 (ETEEL), FATZ-2 (ESEDL), FATZ-3 (ESEEL), miotilina (ESEEL), palladina (ESEDL) e miopalladina (ESDEL) sono molto simili. Una ricerca effettuata in database di sequenze proteiche ha rivelato che questo motivo, E-[S/T]-[D/E]-[D/E]-L, è quasi esclusivamente ristretto nei Vertebrati alle proteine delle famiglie FATZ e miotilina; inoltre, esso sembra essere conservato da zebrafish ad uomo, suggerendo la sua importanza per le proteine che lo contengono. Il programma ELM (che predice siti funzionali in proteine eucariotiche) ha predetto che gli ultimi quattro amino acidi delle proteine FATZ, miotilina, palladina e miopalladina costituiscono un motivo di legame per le proteine con domini PDZ di classe III (X-[D/E]-X-[V/I/L]). Il mio primo obiettivo è stato quello di verificare se le proteine caratterizzate da questo nuovo motivo C-terminale potessero effettivamente legare domini PDZ. E’ noto dalla letteratura che tutti e tre i componenti della famiglia FATZ legano il PDZ di ZASP, e che l’interazione tra ZASP e miotilina è mediata dalla regione C-terminale di quest’ultima. Oltre a ZASP, altri due membri della famiglia enigma, ALP e CLP-36, sono stati inclusi nello studio. Le proteine della famiglia FATZ e miotilina sono state prodotte sia in versione full-length sia priva degli ultimi cinque amino acidi per essere utilizzate in saggi di interazione AlphaScreen (Amplified Luminescence Proximity Homogeneous Assay). Peptidi biotinilati, fosforilati e non, corrispondenti ai cinque amino acidi finali delle FATZ, miotilina, palladina e miopalladina sono stati inoltre impiegati nei saggi AlphaScreen, così come un peptide di controllo avente in ultima posizione un acido glutammico (E) invece che una leucina (L). I risultati riportati in questa tesi dimostrano che gli ultimi cinque amino acidi delle proteine delle famiglie FATZ e miotilina sono responsabili del legame ai domini PDZ di ZASP, ALP e CLP-36, e che la natura dell’ultimo residuo aminoacidico è cruciale per questa interazione. Inoltre, la fosforilazione del residuo di serina o treonina del ligando C-terminale può influenzare il legame dei peptidi nei confronti dei domini PDZ della famiglia enigma. La proteina ?-actinina-2 è stata introdotta nello studio, poiché la sua sequenza C-terminale (GESDL) è classificata come motivo di legame per i domini PDZ di classe I (X-[S/T]-X-[V/I/L]). Gli esperimenti AlphaScreen hanno confermato l’interazione di ?-actinina-2 (sia della forma full-length sia dei peptidi C-terminali, fosforilati e non) con i PDZ di ZASP e ALP, e hanno fatto emergere una nuova interazione con il PDZ di CLP-36. Molte di queste interazioni sono state verificate con un altro metodo di interazione proteina-proteina in vitro, il TranSignal PDZ Domain Array. Sulla base dei risultati di PDZ array è stato possibile identificare un altro membro della famiglia di proteine enigma, RIL, in grado di legare il motivo E-[S/T]-[D/E]-[D/E]-L. Possiamo considerare questi cinque amino acidi C-terminali come un nuovo motivo di legame per le proteine con domini PDZ di classe III, specifico per i domini PDZ delle proteine enigma. Per poter meglio quantificare la forza delle interazioni studiate, alcuni esperimenti di SPR (Surface Plasmon Resonance) sono stati eseguiti nel laboratorio del Dr. A. Baines all’Università di Kent, UK. Le affinità delle interazioni tra il dominio PDZ di ZASP e alcuni dei peptidi fosforilati e non-fosforilati delle famiglie di proteine FATZ e miotilina risultano essere nell’ordine del nM. Gli esperimenti di SPR hanno portato anche all’identificazione di un’interazione tra il PDZ di ZASP e ANKRD2. Si pensa che questa proteina, membro della famiglia MARP, sia coinvolta nelle vie di risposta a stress muscolari. ANKRD2 può trovarsi sia nella banda-I del sarcomero sia nel nucleo ed è in grado di legare diversi fattori di trascrizione, come YB-1, PML e p53. La scoperta di questa interazione rafforza l’ipotesi che il disco-Z, oltre ad un ruolo specificamente strutturale, potrebbe essere coinvolto in vie di segnalazione. Dal momento che a livello del disco-Z molte proteine hanno più di un partner proteico, sarebbe utile cercare di definire il livello e il profilo di espressione delle singole proteine in tessuti muscolari con diverse caratteristiche. Un altro obiettivo del mio lavoro è stato quindi quello di valutare l’abbondanza degli mRNA di alcune delle proteine del disco-Z da me studiate con la Real-Time PCR. Allo scopo sono stati presi in considerazione quattro tessuti muscolari di topo adulto: il tibiale (un muscolo scheletrico a contrazione rapida), il soleo (un muscolo scheletrico a contrazione lenta), il gastrocnemio (un muscolo scheletrico con fibre miste) e il muscolo cardiaco. La differente distribuzione delle FATZ, miotilina e ZASP (con le sue varianti di splicing) suggerisce che, almeno in topo, le interazioni tra queste proteine potrebbero essere compartimentalizzate in distinte fibre muscolari.

Low back pain (LBP) is associated with intervertebral disc (IVD) degeneration and exerts enormous socioeconomic burdens on the society. The nucleus pulposus (NP) is the structural and functional core of the IVD, and plays vital roles in its homeostasis. Although the etiology of IVD degeneration is not fully understood, the cellular changes of the NP have been proposed to be associated with degeneration. Conventional management for IVD degeneration primarily targets to relieve LBP and other symptoms without restoring or preserving disc function. Novel therapeutic strategies have emerged with an aim to retard or even reverse disc degeneration. In particular, the use of growth factors, such as the bone morphogenetic proteins (BMP), has received considerable attention due to their anabolic effects on extracellular matrix (ECM) synthesis by NP cells. BMP-2 and BMP-7 are of great interest for their involvement in osteogenesis, chondrogenesis, and development and maintenance of the IVD. To date, the benefits of BMP-2 on disc degeneration are controversial, given the inconsistent findings from animal model studies. The effectiveness of BMP-7 in disc repair, however, has been well demonstrated both in vitro and in vivo. A better understanding of the differences between BMP-2 and BMP-7 regulatory action on NP cells may facilitate future applications of BMP in disc repair/regeneration. This study hypothesized that BMP-2 and BMP-7 act differentially on human NP cells via different signal transduction processes. The differential effect of BMP-2 and BMP-7 was first tested in bovine NP cells using a three-dimensional culture system (alginate beads). Both BMP-2 and BMP-7 enhanced ECM production and phenotypic characteristics of bovine NP cells. Notably, BMP-7 was significantly more potent than BMP-2 in this regard. The effects of BMPs were further tested on non-degenerated (ND-NP) and degenerated (D-NP) human NP cells. The DMMB assay revealed that BMP-7 exerted a superior up-regulatory action on GAG production of D-NP cells compared to BMP-2. Furthermore, the overall response of D-NP cells to BMP-2 and BMP-7 was significantly lower than ND-NP cells. Immunohistochemical staining and quantitative RT-PCR assays demonstrated that D-NP cells possess a more fibroblastic and less chondrocyte-like phenotype than ND-NP cells. At the mRNA level, the BMP receptor BMPR1A was not expressed in D-NP cells. BMP-7, but not BMP-2, induced expression of BMPR1A in D-NP cells. On the other hand, gene expression of selected TGF-β pathway components and hypoxia pathway components were significantly up-regulated by BMP-2 but down-regulated by BMP-7. These findings suggest that D-NP cells can activate differential molecular cascades in response to BMP-2 and BMP-7. In conclusion, this study showed a superior effect of BMP7 in up-regulation of classical BMP signaling components including BMP receptor BMPR1A. The reduced responsiveness of D-NP cells to BMP-2 and BMP-7 stimulation may be related to a different expression pattern of BMP receptors. This study provides insights into the differential regulatory actions of BMP-2 and BMP-7 on human NP cells and facilitates the future application of BMPs in managing disc degeneration.
published_or_final_version
Orthopaedics and Traumatology
Doctoral
Doctor of Philosophy

The main focus of my PhD studies was aimed at understanding the general mechanism of nuclear translocation and isolating novel components of the nuclear translocation pathway in neurons. Using the Drosophila visual system as an in vivo model to study nuclear motility in developing photoreceptor cells (R-cells), I have identified a novel role for the Ser/Thr kinase Misshapen (Msn) and the endocytic trafficking pathway in regulating the nuclear translocation process.
The development of R-cells in the Drosophila eye disc is an excellent model system for the study of nuclear motility owing to its monolayer organization and the stereotypical translocation of its differentiating R-cell nuclei along the apical-basal plane. Prior to my thesis work, several laboratories had identified dynein and its associating proteins in R-cell nuclear translocation, however nothing was known about the signalling pathway that controlled their function in nuclear migration. Thus, one of my thesis goals was to elucidate the signalling mechanism controlling nuclear translocation in R-cells.
Using a combination of molecular and genetic approaches, I identified Msn as a key component of a novel signalling pathway regulating R-cell nuclear translocation. Loss of msn causes a failure of R-cell nuclei to migrate apically. Msn appears to control R-cell nuclear translocation by regulating the localization of dynein and Bicaudal-D (Bic-D). My results also show that Msn enhances Bic-D phosphorylation in cultured cells, suggesting that Msn regulates R-cell nuclear migration by modulating the phosphorylation state of Bic-D. Consistently, my results show that a Bic-D-phosphorylation-defective mutation disrupted the apical localization of both Bic-D and dynein. I propose a model in which Msn induces the phosphorylation of Bic-D, which in turn modulates the activity and/or subcellular localization of dynein leading to the apical migration of R-cell nuclei.
In addition to studying Msn, I have also searched for additional players in R-cell nuclear migration. From a gain-of-function approach, I found that the misexpression of the GTPase-activating-protein (GAP) RN-Tre caused a severe defect in R-cell nuclear migration. Since mammalian RN-Tre is involved in negatively regulating Rab protein activity, I speculated that the RN-Tre misexpression phenotype reflected a role for Rab-mediated vesicular transport in regulating R-cell nuclear migration. I systematically examined the potential role of Rab family proteins in R-cell nuclear migration and found that interfering with the function of Rab5, Rab11 or Shibire caused a similar nuclear migration phenotype. I propose that an endocytic pathway involving these GTPases is required for the targeting of determinants to specific subcellular locations, which in turn drive the apical migration of R-cell nuclei during development.

The aim of the study we present in this PhD thesis is to gain a deeper insight into structure of Z-disc proteins FATZ 1 and y-filamin. Z-discs are multi-protein complexes which are the primary conduits of the force generated by striated muscle contraction. The protein composition of Z-disc is not well defined and new proteins are continuously being reported. FATZ 1 is expressed early during myofibrilogenesis and it is presumed to have an important role in Z-disc assembly and functioning. At the moment, it is the only Z-disc protein for which a direct connection with the signal transducer protein phosphatase calcineurin has been demonstrated. The protein sequence of FATZl does not account for its interaction with numerous proteins as it does not contain any known protein interaction domains. For this reason we investigated the structural characteristics of this Z-disc protein. We studied also y-filamin, a binding partner of FATZl. The role of filamins in cytoskeleton organization and signal transduction has been well documented. y-filamin is the muscle specific isoform of filamins and it has not been studied as extensively as other filamin isoforms. y-filamin can be found both at sarcolemma and at the Z-disc of striated muscle cells and for this reason is considered to be a structural and functional link between contractile apparatus and sarcolemma. y-filamin repeat 23 was reported to be necessary for the interaction of y-filamin and FATZ. We determined crystal structure of y-filamin repeat 23, and found that y-filamin repeat 23 bound nickel. Detailed analysis of y-filamin protein sequence indicated that other filamin repeats could bind nickel or similar divalent cations. Physiological importance of nickel in Eukaryotes is not well studied, but in Prokaryotes nickel is an essential component of many enzymes. Nickel binding to a protein with mainly structural role would be an exception, since nickel is bound mainly to proteins with enzymatic activities. We analyzed also structural characteristics of y-filamin repeat 20. The y-filamin repeat 20 has the unique muscle specific insertion responsible for protein localization in the Z-disc. The muscle specific insertion presents no homology with any sequence with known protein structure. Our results from structural studies of repeat 20 indicate that the insertion in repeat 20 influences the fold of this repeat and makes it structurally different from y-filamin repeat 23.

Abstract The Z-disk is a sophisticated structure that connects adjacent sarcomeres in striated muscle myofibrils. α-Actinin provides strength to the Z-disks by crosslinking the actin filaments of adjacent sarcomeres. α-Actinin is an antiparallel homodimer, composed of an N-terminal actin binding domain (ABD), the central rod domain, and two pairs of C-terminal EF-hands. The PDZ-LIM domain proteins interact with α-actinin at the Z-disk. Of these proteins, only the actinin-associated LIM protein (ALP), Z-band alternatively spliced PDZ-containing protein (ZASP/Cypher) and C-terminal LIM protein (CLP36) have a ZASP/Cypher-like (ZM) motif consisting of 26-27 conserved residues in the internal region between the PDZ and LIM domains. The aim of this work was to understand the molecular interplay between the ZM-motif containing members of the PDZ-LIM proteins and α-actinin. To unveil the biological relevance of the interaction between the PDZ-LIM proteins and α-actinin, naturally occurring human ZASP/Cypher mutations were analyzed. Two interaction sites were found between ALP, CLP36 and α-actinin using recombinant purified proteins in surface plasmon resonance (SPR) analysis. The PDZ domain of ALP and CLP36 recognized the C-terminus of α-actinin, whereas the internal regions bound to the rod domain. Further characterization showed that the ALP internal region adopts and extended conformation when interacting with α-actinin and that the ZM-motif partly mediated the interaction, but did not define the entire interaction area. ZASP/Cypher also interacted and competed with ALP in binding to the rod domain. The internal fragments containing the ZM-motif were important for co-localization of ALP and ZASP/Cypher with α-actinin at the Z-disks and on stress fibers. The absence of ALP and ZASP/Cypher in focal contacts indicates that other interacting molecules, for instance vinculin and integrin, may compete in binding to the rod in these areas or additional proteins are required in targeting to these locations. The co-localization of the ZASP/Cypher with α-actinin could be released by disrupting the stress fibers leading to an accumulation of α-actinin in the cell periphery, whereas ZASP/Cypher was not in these areas. This suggests that an intact cytoskeleton is important for ZASP/Cypher interaction with α-actinin. Earlier studies have shown that mutations in the ZASP/Cypher internal region are associated with muscular diseases. These mutations, however, did not affect ZASP/Cypher co-localization with α-actinin or the stability of ZASP/Cypher proteins. The Z-disk possesses a stretch sensor, which is involved in triggering hypertrophic growth as a compensatory mechanism to increased workloads. α-Actinin is a docking site of molecules that are involved in hypertrophic signaling cascades mediated by calsarcin-calcineurin and protein kinase C (PKC) isoforms. The internal interaction site may be involved in targeting PKCs, which bind to the LIM domains of ZASP/Cypher, to the Z-disks. The similar location of the internal interaction site with calsarcin on the rod suggests that ZASP/Cypher, ALP and CLP36 may regulate calsarcin-mediated hypertrophic signaling.

Intercalated discs (ICDs) are cardiac-specific structures located at the longitudinal termini of cardiomyocytes. Classically, the functions assigned to ICDs include mechanical and electrical communications among adjacent cardiomyocytes. More recently, it has been increasingly realized that ICDs also function in signal transduction and regulation of the surface expression of ion channels. Accordingly, defects of ICD components are shown to cause a number of human cardiac diseases and changes of ICDs are associated with cardiomyopathy, arrhythmias, and heart failure. The expansion of our knowledge about the development, function and maintenance of ICDs are promoted by identification, cataloging and characterization of the molecular components of the ICDs. In this thesis, I characterize a family of Xin repeat-containing proteins, which are striated muscle-specific and localized to the ICDs in the cardiomyocytes. This thesis provides novel insights into the mechanism of the formation, maintenance and functions of ICDs.

Our previous studies showed that the Xin repeat-containing proteins play critical role in cardiac morphogenesis and cardiac function. Knocking down the Xin in chicken embryo collapses the wall of developing heart chambers and leads to abnormal cardiac morphogenesis. In mammals, a pair of paralogous genes, Xinα and Xinβ , exists. Ablation of the mouse Xinα ( mXinα) does not affect heart development. Instead, the mXinα-deficient mice show adult late-onset cardiac hypertrophy and cardiomyopathy with conduction defects. The ICD structural defects in mXinα-null mice occur between 1 and 3 months of age and progressively worsen with aging. The mXinα-deficient hearts up-regulate mXinβ, suggesting a partial compensatory role of mXinβ.

In this thesis, I focus on two questions. First, what are the molecular mechanisms of mXinα's functions that account for the observed phenotypes in the mXinα-deficient hearts? And second, what are the functions of mXinβ? Through biochemical methods and electron microscopy, I demonstrated that mXinα binds and bundles actin filaments. In addition, a direct interaction between mXinα and the adherens junction protein β-catenin facilitates mXinα's interaction with the actin filaments. Based on this in vitro characterization of mXinα, we proposed that mXinα may act as a direct link between the adherens junctions and actin cytoskeleton, thus providing an important means to strengthening the intercellular adhesion at the ICDs. To characterize mXinβ's roles, I generated and characterized mXinβ-knockout mice. I showed that complete loss of mXinβ leads to cardiac morphological defects, diastolic dysfunction and heart failure, which lead to severe growth retardation and early postnatal lethality. I also showed that mXinβ might be involved in a number of cell signaling pathways and provide multiple lines of evidence to support mXinβ's roles in the formation of ICDs.

In summary, this thesis provides novel insights into the specialization of the adherens junctions at the ICDs to withstand the contractile forces, and the molecular mechanisms for the establishment, maintenance and function of ICDs. The knowledge gained from the roles of Xin proteins in cardiac development and function will likely provide new insights for improved therapeutic strategies for human cardiomyopathy, arrhythmias and heart failure.

Intercalated discs (ICDs) are cardiac-specific structures located at the longitudinal termini of cardiomyocytes. Classically, the functions assigned to ICDs include mechanical and electrical communications among adjacent cardiomyocytes. More recently, it has been increasingly realized that ICDs also function in signal transduction and regulation of the surface expression of ion channels. Accordingly, defects of ICD components are shown to cause a number of human cardiac diseases and changes of ICDs are associated with cardiomyopathy, arrhythmias, and heart failure. The expansion of our knowledge about the development, function and maintenance of ICDs are promoted by identification, cataloging and characterization of the molecular components of the ICDs. In this thesis, I characterize a family of Xin repeat-containing proteins, which are striated muscle-specific and localized to the ICDs in the cardiomyocytes. This thesis provides novel insights into the mechanism of the formation, maintenance and functions of ICDs. Our previous studies showed that the Xin repeat-containing proteins play critical role in cardiac morphogenesis and cardiac function. Knocking down the Xin in chicken embryo collapses the wall of developing heart chambers and leads to abnormal cardiac morphogenesis. In mammals, a pair of paralogous genes, Xin&alpha and Xin&beta, exists. Ablation of the mouse Xin&alpha (mXin&alpha) does not affect heart development. Instead, the mXin&alpha-deficient mice show adult late-onset cardiac hypertrophy and cardiomyopathy with conduction defects. The ICD structural defects in mXin&alpha-null mice occur between 1 and 3 months of age and progressively worsen with aging. The mXin&alpha-deficient hearts up-regulate mXin&beta, suggesting a partial compensatory role of mXin&beta. In this thesis, I focus on two questions. First, what are the molecular mechanisms of mXin&alpha's functions that account for the observed phenotypes in the mXin&alpha-deficient hearts? And second, what are the functions of mXin&beta? Through biochemical methods and electron microscopy, I demonstrated that mXin&alpha binds and bundles actin filaments. In addition, a direct interaction between mXin&alpha and the adherens junction protein &beta-catenin facilitates mXin&alpha's interaction with the actin filaments. Based on this in vitro characterization of mXin&alpha, we proposed that mXin&alpha may act as a direct link between the adherens junctions and actin cytoskeleton, thus providing an important means to strengthening the intercellular adhesion at the ICDs. To characterize mXin&beta's roles, I generated and characterized mXin&beta-knockout mice. I showed that complete loss of mXin&beta leads to cardiac morphological defects, diastolic dysfunction and heart failure, which lead to severe growth retardation and early postnatal lethality. I also showed that mXin&beta might be involved in a number of cell signaling pathways and provide multiple lines of evidence to support mXin&beta's roles in the formation of ICDs. In summary, this thesis provides novel insights into the specialization of the adherens junctions at the ICDs to withstand the contractile forces, and the molecular mechanisms for the establishment, maintenance and function of ICDs. The knowledge gained from the roles of Xin proteins in cardiac development and function will likely provide new insights for improved therapeutic strategies for human cardiomyopathy, arrhythmias and heart failure.

PEG-lipid stabilized bilayer disks have been found in lipid mixtures containing polyethylene glycol (PEG)-lipids where the combination of a high bending rigidity and low PEG-lipid/lipid miscibility favours disk formation. The disks are planar and circular in shape and their long-term stability is excellent. Theoretical calculations and experimental observations suggest that the micelle forming PEG-lipid are situated at the rim of the aggregate, protecting the hydrophobic lipid chains in the bulk of the aggregate from contact with water. This thesis deals with fundamental aspects concerning the lipid distribution in the disks, as well as with development, optimization, and initial evaluation of the disks as model membranes in partition and interaction studies. Small angle neutron scattering was used to study the partial segregation of components within the bilayer disk. The experiments verified that the PEG-lipids segregate and accumulate at the bilayer disk rim. The proof of component segregation is important from a fundamental point of view and useful, as exemplified in the below-mentioned study of melittin-lipid interaction, when interpreting partition or binding data obtained from studies based on bilayer disks. Today liposomes are often used as model membranes in partition and interaction studies. Using liposomes to predict, e.g., drug partitioning can however have certain drawbacks. In this thesis the disks were proven to be attractive alternatives to liposomes as model membranes in partition studies. The formation of bilayer disks by a technique based on detergent depletion enabled incorporation of a transmembrane protein in the bilayer disks and opened up for the use of disks as model membranes in membrane protein studies. Further, bilayer disks were used in a comparative study focused on the effect of aggregate curvature on the binding of the peptide melittin. Various techniques were used to perform initial evaluations of the bilayer disks as model membranes. Of these, capillary electrophoresis and biosensor-based technology had not been used before in combination with bilayer disks.

Avaliaram-se diferentes níveis de lisina digestível para frangos de corte, machos, utilizando-se 1050 aves dos 12 aos 22 e 1015 aves dos 37 aos 49 dias de idade. O delineamento experimental foi inteiramente casualizado com cinco tratamentos, sete repetições e as unidades experimentais continham, respectivamente, trinta e vinte e nove, correspondentes às fases inicial e final. Nas duas fases, as dietas dos tratamentos eram isoenergéticas e isoprotéicas à base de milho e farelo de soja, suplementadas dos demais aminoácidos, quando necessário. Na fase inicial, os tratamentos corresponderam aos níveis 1,05; 1,10; 1,15; 1,20 e 1,25% de lisina digestível, em dietas com 3050 kcal de EM/kg e com 19% de PB. Na fase final, os tratamentos corresponderam aos níveis 0,90; 0,95; 1,00; 1,05; 1,10% de lisina digestível, em dietas com 3250 de EM/kg e 18% de PB. Avaliaram-se ganho de peso, consumo de ração, conversão alimentar, composição e deposição de nutrientes corporais, e na fase final as características e rendimento de cortes na carcaça. Na fase inicial, os níveis de lisina dietéticos influenciaram o consumo de ração, constatando-se resposta (P<0,01) quadrática. Observou-se efeito (P<0,01) linear ascendente no peso da carcaça. Dos componentes químicos, expressos em porcentagem na carcaça, houve resposta (P=0,02) quadrática do teor de lisina digestível na concentração de proteína bruta. Efeitos similares foram observados na deposição de proteína e água da carcaça e do corpo vazio, com aumento (P<0,01) linear, em resposta ao acréscimo de lisina na dieta. Da mesma forma, houve efeito (P<0,01) linear ascendente dos níveis de lisina digestível da ração sobre o peso vivo reconstituído. Na composição química das vísceras e sangue observou-se efeito (P<0,10) do nível de lisina apenas na concentração de matéria mineral do corpo vazio e a resposta foi decrescente com o aumento do aminoácido. O nível 1,10% de lisina digestível satisfaz as necessidades do desempenho de frangos de corte machos, dos 12 aos 22 dias de idade, mas ao considerar a composição química e deposição dos nutrientes corporais esta exigência torna-se igual ou maior a 1,25%. Na fase final, das variáveis do desempenho, apenas conversão alimentar foi influenciada pelos níveis de lisina digestível na ração, caracterizando-se efeito (P<0,01) linear descendente. Das características de carcaça e rendimento de cortes, apenas gordura abdominal teve efeito (P=0,02) quadrático em função dos níveis de lisina empregados nas rações. Nas variáveis de composição química da carcaça e do corpo vazio, observou-se efeito (P<0,01) quadrático dos níveis de lisina digestível apenas no teor da matéria mineral. Nas demais frações, vísceras e sangue, não houve efeito dos níveis de lisina digestível na composição química. Observou-se, contudo, indicação de aumento (P=0,09) linear na taxa de deposição de proteína da carcaça e do corpo vazio em resposta a elevação do nível de lisina digestível. Com base no desempenho o nível de lisina digestível deve ser igual ou maior a 1,10%, mas em relação à quantidade de gordura abdominal o indicado seria 1,00%. Essas informações das duas fases estudadas confirmam que a exigência de lisina digestível para desempenho in vivo é inferior às reais demandas para síntese protéica na deposição de massa muscular esquelética
A group of 1050 commercial male broilers, ranging from 12 to 22 days of age, and a group of 1015 commercial male broilers, ranging from 37 to 49 days of age were used to evaluate different digestible lysine levels. A completely randomized trial was used, with 5 treatments (1.05, 1.10, 1.15, 1.20 and 1.25% of digestible lysine, respectively) in Initial Phase and 5 treatments (0.90, 0.95, 1.00, 1.05 and 1.10% of digestible lysine, respectively) in Final Phase, applied in 7 replications and 35 experimental units. The experimental unit was 30 birds in first group and 29 birds in second group. Lysine levels were added in isoenergetic (3050 and 3250 kcal of ME/kg for groups, respectively) and isoproteic (19 and 18% of Crude Protein, for groups, respectively) corn and soy meal rations. The rations were balanced with several amino acids when needed. Weight gain, feed intake, feed: gain rate, body composition, nutrient deposition were measured and carcass characteristics and cut yield in Final Phase. In Initial Phase dietary lysine levels influenced ration consumption (P<0.01), with increasing linear effect on carcass weight (P<0.01). Quadratic effect was observed of the digestible lysine on crude protein concentration (P=0.02), when chemical compounds were analyzed. Protein and water deposition was observed (P<0.01), in carcass and empty body, with linear increase, due to the lysine addition in ration. Same effects were observed on reconstituted body weight. Chemical composition of blood and offal were statistically different (P<0.10) in empty body mineral matter concentrations with decreasing values when increasing lysine levels. Lysine levels of 1.10% was sufficient for 12 to 22 day old male broiler development requirements, however, considering body chemical composition, the level needed would not be down 1.25%. In Final Phase, dietary lysine levels influenced feed: gain rate (P<0.01), with decreasing linear effect. Quadratic effect was observed (P=0.02) due to lysine levels used when evaluating carcass characteristics and cuts, abdominal fat deposition. Chemical composition of carcass was statistically different (P<0.01) in empty body mineral matter, with squared effect due to the lysine level considered. Blood and offal chemical composition had no effects, however, linear increase (P=0.09) was shown in carcass protein deposition and empty body due to the increase of lysine levels. Considering the performance, digestible lysine level might be 1.10% or higher, however, for abdominal fat composition, suggested level is 1.00%. According to the given results, the needs for in vivo performance are lower than the real requirements for protein synthesis in skeletal muscle formation

Introduction. Arrhythmogenic cardiomyopathy (ACM) is a predominantly genetically determined disease characterized by fibrofatty replacement which leads to right ventricular failure, arrhythmias, and sudden cardiac death (SCD). ACM is inherited as an autosomal dominant trait with incomplete penetrance. Advances in genetic technology have revealed substantial genetic heterogeneity: at least 15 independent loci and 13 disease genes have now been identified associated with the disease, with the large involvement of the genes encoding for desmosomal and area composita proteins. Since causative mutations in ACM genes have been detected in about 60% of probands, additional and still unknown disease-genes could be involved. Aim of the study. This study aimed at detecting causative mutations underlying the disease expression in a cohort of 59 Italian unrelated index cases through a new targeted next generation sequencing (NGS) approach. Moreover, the identification of novel disease loci and genes in four families with recurrence of ACM was attempted by integrating different genetic approaches. Methods. Three different NGS approaches have been applied: targeted gene panels (TGP), whole exome sequencing (WES), and whole genome sequencing (WGS). Two custom targeted gene panels, including 56 genes associated with different cardiomyopathies or 69 known and candidate ACM genes have been used to screen both familial and sporadic cases. Genetic analysis was extended to available family members to evaluate the segregation of each mutation identified in the index case. The whole exome of 11 subjects belonging to 4 families was sequenced. In one of these families (Family#6) also WGS in 3 subjects in addition to multipoint linkage analysis was performed. Results. Targeted gene panels resulted a valuable tool for mutation screening in patients affected with ACM. At least a mutation was found in 15 out of 19 probands screened with the ‘Cardiomyopathies gene panel’. Moreover, by using the ‘ACM known and candidate genes panel’, a mutation in a candidate gene has been identified in 40.6% of probands negative for mutations in known-disease genes. Two novel missense variants in TJP1 gene (p.R265W, p.Y669C), encoding for protein ZO-1 and two in CDH2 gene (p.E493G, p.V491G), encoding for N-cadherin resulted of particular interest, other than a stop mutation in TP63 gene (p.R266*). Both TJP1 mutations affected highly conserved amino acid residues and in silico analysis of p.Y669C mutation, which segregates within the proband’s family, showed its predicted damaging effect into the protein structure. WES analysis allowed to identify a pathogenic mutation in DSP gene (p.Q1297*) not previously detected by dHPLC in Family#3, and two putative TTN mutations (p.R32573C and p.L32198M) that segregate in Family#4 and Family#5. In Family#5, a rare stop mutation in a candidate gene (CMYA5, p.K3597*) was also identified. In Family #6, a multi-step approach was applied. The availability of many affected and unaffected family members allowed to perform linkage analysis in an ‘affected only’ approach. The analysis pointed out the presence of two regions with positive pLOD score values on chromosome 19p13.3 and 11q21. The first corresponded to a 7cM region and was shared by all the affected subjects, but one. The 3cM region on chromosome 11q21 segregated in all the affected subjects exception made for two, who carried a PKP2 splice site variant (c.2578-3 T>C). WES performed in 4 patients of the family failed to identified a possible shared pathogenic mutation neither inside the critical regions, nor in the whole exome. These results were confirmed with WGS performed in 2 affected and a healthy individual. Moreover WGS pointed out the presence of a huge amount of complex variants located in repetitive intronic or intragenic regions. Discussion. The identification of disease-causing mutations facilitates timely diagnosis, allows the prevention of complications and determines the potential risk in close relatives of a proband. For hereditary cardiomyopathies, as well as for most mendelian diseases, in the last ten years NGS technology improved this process allowing to parallel sequence a large amount of genes in a time and cost-efficient manner. Since mutations in the most index cases sequenced by using a targeted gene panel have been detected, it resulted a valid approach not only for genetic testing but also to identify putative novel disease genes. In the present study, novel mutations in genes encoding for intercalated disc proteins (TJP1, CDH2) have been identified, confirming the idea that ACM has to be considered a ‘junctional disease’ rather than only a ‘desmosomal disease’. Despite TGP remains the most commonly used approach for hereditary cardiomyopathies, WES, applied in familial ACM cases, allowed to identify novel variants in candidate genes that would be not detected with TGP and could have a possible pathogenic role or a modifier effect in the phenotypic expression. Finally, even though WGS didn’t allow to make a genotype-phenotype correlation in Family #6, the huge amount of data produced represent permanent data that could be re-analysed in the future with new insights and the discovery of novel disease genes.
Introduzione. La cardiomiopatia aritmogena (ACM) è una patologia ereditaria del muscolo cardiaco, caratterizzata da una progressiva sostituzione adiposa o fibroadiposa a carico prevalentemente del miocardio del ventricolo destro. Dal punto di vista clinico, è una patologia eterogenea con ampia variabilità clinica inter- ed intra- familiare; presenta infatti sia forme completamente asintomatiche sia forme molto gravi con rischio di morte improvvisa. Questa patologia, geneticamente eterogenea, è trasmessa come carattere autosomico dominante a penetranza incompleta ed espressività variabile. Ad oggi 15 loci e 13 geni sono stati associati alla malattia, di cui gran parte codificano per proteine desmosomali e proteine della cosiddetta area composita dei dischi intercalari. Poiché sono state identificate mutazioni causative in geni noti sono nel 60% dei casi, altri geni, non ancora identificati, potrebbero essere coinvolti nella comparsa del fenotipo patologico. Scopo della ricerca. Nello studio descritto nella presente tesi, il DNA di 59 casi indice è stato analizzato attraverso due diversi pannelli di geni tramite sequenziamento di nuova generazione (NGS). Inoltre, in quattro famiglie con ricorrenza di casi di ACM, in cui non sono state identificate mutazioni nei geni desmosomali, sono state integrate diverse tecniche allo scopo di identificare nuovi loci e geni malattia. Metodi. Nell’ambito del sequenziamento di nuova generazione (NGS) tre diversi approcci sono stati utilizzati: il sequenziamento di pannelli di geni target (TGP), il sequenziamento dell’intero esoma (WES) e il sequenziamento dell’intero genoma. Due sono i pannelli di geni considerati, uno comprendente 56 geni associati a diverse cardiomiopatie, l’altro comprendente 69 geni, tra cui i 13 geni associati alla cariomiopatia aritmogena e 56 geni candidati. L’analisi genetica è stata poi estesa ai familiari dei probandi, ove disponibili, per valutare la segregazione delle mutazioni identificate. L’intero esoma è stato poi sequenziato in 11 soggetti appartenenti a 4 diverse famiglie. Infine, in una di queste famiglie (Famiglia#6) è stata eseguita un’analisi di linkage ed è stato sequenziato l’intero genoma di tre soggetti. Risultati. L’utilizzo dei pannelli di geni target si è rivelato una buona strategia per lo screening di mutazioni in pazienti affetti da ACM: almeno una mutazione è stata trovata in 15 probandi su un totale di 19 analizzati con il pannello di geni associati a diverse cardiomiopatie; inoltre, è stata identificata una mutazione in uno dei geni candidati nel 40.6% dei probandi sequenziati con il secondo pannello e risultati negativi per mutazioni nei geni noti. Tra queste mutazioni, risultano di particolare interesse 2 nuove mutazioni missenso localizzate nel gene TJP1 (p.R265W, p.Y669C), 2 mutazioni nel gene che codifica per l’N-caderina (CDH2, p.E493G, p.V491G) e una mutazione di stop nel gene TP63 (p.R266*). Entrambe le mutazioni di TJP1 riguardano aminoacidi altamente conservati, inoltre un’analisi in silico degli effetti della mutazione p.Y669C, che segrega all’interno della famiglia, evidenzia un riarrangiamento della struttura proteica della proteina che riporta la mutazione rispetto alla proteina wild-type. L’analisi dell’esoma nella Famiglia#3 ha permesso di identificare una mutazione patogena di stop nel gene DSP, che non era stato possibile individuare al precedente screening tramite dHPLC. Inoltre, tale approccio ha permesso di identificare due putative mutazioni nel gene TTN (p.R32573C and p.L32198M) che segregano nelle Famiglie #4 e #5, rispettivamente. Nella Famiglia #5 inoltre è stata identificata una rara mutazione di stop in un gene candidato (CMYA5, p.K3597*). Per lo studio della Famiglia#6, invece, sono stati utilizzati diversi approcci. Data la disponibilità di molti soggetti affetti e sani appartenenti alla famiglia, in seguito alla genotipizzazione degli stessi, è stata eseguita un’analisi di linkage. L’analisi ha evidenziato la presenza di due loci che riportano valori di lod score positivi a livello del cromosoma 19p13.3 e del cromosoma 11q21. Il primo locus corrisponde ad una regione di 7 cM ed è condiviso da tutti i soggetti affetti della famiglia, eccetto uno. La regione di 3cM nel cromosoma 11q21 invece segrega in tutti i soggetti affetti ad eccezione di due, che portano una mutazione in un sito di splicing del gene PKP2 (c.2578-3 T>C). Il sequenziamento dell’esoma in 4 soggetti affetti della famiglia non ha permesso di identificare nuove mutazioni condivise né all’interno delle due regioni critiche, né all’interno dell’intero esoma. Tali risultati sono stati confermati dal sequenziamento del genoma effettuato in due affetti e un soggetto sano della famiglia. Il sequenziamento del genoma ha inoltre messo in luce la presenza di un’enorme quantità di varianti complesse localizzate in regioni introniche o intrageniche. Discussione. L’identificazione di mutazioni causative nella cardiomiopatia aritmogena facilita la diagnosi tempestiva, permette di prevenire eventuali complicazioni e determina il rischio di sviluppare la malattia nei familiari di un soggetto affetto. Per le cardiomiopatie ereditarie, come per la maggior parte delle malattie mendeliane, negli ultimi 10 anni le tecniche NGS hanno apportato grossi miglioramenti nel processo di identificazione di mutazioni, permettendo di sequenziare una grande quantità di geni parallelamente, in modo più rapido e meno costoso. In questo studio, dal momento che sono state identificate mutazioni nella maggior parte di probandi analizzati, l’utilizzo di pannelli di geni target si è dimostrato un valido approccio non solo per un test genetico ma anche per l’identificazione di nuovi geni malattia. Sono state infatti identificate nuove mutazioni in geni che codificano per proteine dei dischi intercalari dei cardiomiociti, confermando l’idea che l’ACM deve essere considerata una ‘malattia delle giunzioni’ piuttosto che una ‘malattia dei desmosomi’. Nonostante l’utilizzo di pannelli di geni target rimanga l’approccio più comunemente usato nella ricerca di mutazioni nelle cardiomiopatie ereditarie, il sequenziamento dell’intero esoma, applicato in questo studio solo a casi familiari, ha permesso di identificare varianti in geni candidati non inclusi nei pannelli di geni e che potrebbero avere un ruolo nell'espressione del fenotipo patologico. Infine, nonostante il sequenziamento del genoma nella Famiglia#6 non abbia permesso al momento di stabilire una correlazione tra genotipo e fenotipo al momento, la mole di dati prodotti potrà essere rianalizzata in futuro alla luce di nuovi geni annotati e nuovi geni malattia identificati.