Mitochondria are essential eukaryotic organelles, acting as the sites for numerous crucial metabolic and signalling pathways. The biogenesis of mitochondria requires efficient targeting of several hundreds of proteins from the cytosol, to their varied functional locations within the organelle. The translocation of localized proteins across the inner membrane, and their subsequent folding is achieved by the ATP-dependent function of mitochondrial Hsp70 (mtHsp70). It is a bonafide member of the Hsp70 chaperone family, which are involved in a multitude of functions, together aimed at protein quality control and maintenance of cellular homeostasis. These varied functions of Hsp70 proteins require binding to exposed hydrophobic patches in substrate polypeptides thus preventing non-productive associations. The interaction with substrates occurs through the substrate-binding domain (SBD) and is regulated by the ATPase activity of the nucleotide-binding domain (NBD), through a series of conformational changes. Conversely, substrate binding to the SBD also stimulates ATP hydrolysis, and thereby the core activities of the two domains are regulated by mutual allosteric signalling. This mechanism of bidirectional inter-domain communication is indispensable for Hsp70 function, which is characterized by cycles of substrate binding and release, coupled to cycles of ATP binding and hydrolysis. The process of allosteric regulation in Hsp70 proteins has been comprehensively investigated, especially in the bacterial homolog, DnaK. However, the in vivo functional significance of inter-domain communication in the eukaryotic mtHsp70 system and the mechanism of its regulation remain unexplored. Furthermore, the complex physiological implications of impairment in allosteric communication and their correlation with diverse disease conditions, including Myelodysplastic syndrome (MDS), and Parkinson’s disease (PD), are yet to be elucidated. Based on this brief introduction, the primary research objectives set out in the present thesis were to: 1. uncover the regulation of ligand-modulated allosteric communication between the two domains of mtHsp70; and its in vivo significance in the context of protein import into the organelle. (Chapter 2) 2. understand the role of mtHsp70 in progression of Parkinson’s disease; and to study the modulation of α-synuclein toxicity by the protein quality control function of the mtHsp70 chaperone network. (Chapters 3 and 4) We have employed a battery of genetic and biochemical approaches to investigate the above questions using the Saccharomyces cerevisiae mtHsp70 protein, Ssc1; an essential protein that is involved in a plethora of critical functions in this eukaryotic model system. Objective 1: Structural studies, primarily in bacterial DnaK, have yielded mechanistic insights into its interactions with ligands and cochaperones, as well as conformational transitions in different ligand-bound states. In recent years, the availability of crystal structures of full-length DnaK and detailed information from NMR studies and single-molecule resolution spectroscopic analyses (both DnaK and eukaryotic Hsp70s), have significantly contributed to our understanding of the inter-domain interface, critical residues and contacts, and the energetics of the entire process of ligand-modulated conformational changes. Although eukaryotic mtHsp70s have a high degree of conservation with DnaK, they possess significant differences in their conformational and biochemical properties. They are essential for a vast repertoire of physiological functions, which are distinctly different from their bacterial counterpart. Using a combined in vivo and in vitro approach, we have uncovered specific structural elements within mtHsp70s, which are required for allosteric modulation of the chaperone cycle and maintenance of in vivo functions of the protein. Foremost, we demonstrate that a conserved SBD loop, L4,5 plays a critical role in inter-domain communication, and multiple mutations in this loop result in significant growth and protein translocation defects. The mutants are associated with a specific set of altered biochemical properties, which are indicative of impaired inter-domain communication. Using the loop L4,5 mutant, E467A as a template for genetic screening, we report a series of intragenic suppressor mutations, which are capable of correcting a distinct subset of the altered properties, and thereby leading to restoration of in vivo functions, including growth, preprotein import and mitochondria biogenesis. The suppressors modify the altered conformational landscape associated with E467A, and also provide us with information regarding unique aspects governing the regulation of allosteric communication, especially in physiological contexts. Strikingly, they reveal that restoration of communication in the NBD to SBD direction is sufficient for function, when the protein is primed in a high ATPase activity state. In this unique scenario, the requirement for ATPase stimulation upon substrate binding is rendered unnecessary, thereby making conformational changes in the SBD to NBD direction, dispensable for function. Further, we provide evidence to show that loop L4,5 functions synergistically with the linker region, working in tandem for organization of the inter-domain interface and propagation of communication. Together, our analyses provide the first insights into regulation of allosteric inter-domain communication in vivo and their implications in mitochondrial protein translocation and organelle biogenesis. Objective 2: Point mutations in the loop L4,5 have been associated with Myelodysplastic syndrome. Additionally, a mutation isolated in clinical cases of Parkinson’s disease was found to be impaired in allosteric communication. These observations further highlight the importance of efficient inter-domain communication in mtHsp70 in the complex physiological scenario of eukaryotic cells. Independent clinical screens of PD patients have revealed unique point mutations in the mtHsp70 and a strong association of the gene locus with the disease progression. This is also correlated with decreased mtHsp70 levels in affected neurons and the interactions of this protein with established PD-candidate proteins like α-synuclein and Dj-1. Further, mitochondrial dysfunction is a common phenomenon associated with neurodegenerative disorders. To understand the specific role of mtHsp70 in PD, we have developed a yeast model for studying the disease variants in isolation from other players of the multifactorial disease, and in complete absence of the wild type protein. We generated two analogous PD-mutations in Ssc1, R103W and P486S; which recapitulated the symptoms of mitochondrial dysfunction in affected neurons, including cell death, inner membrane depolarization, increased generation of ROS, and respiratory incompetence. At the molecular level, we observed an increased aggregation propensity of R103W, while P486S exhibited futile enhanced interaction with J-protein cochaperone partners thereby resulting in loss of chaperoning activity and impaired mitochondrial protein quality control. Remarkably, these altered biochemical properties mimicked similar defects in the human mtHsp70 variants, therefore, affirming the involvement of mtHsp70 in PD progression. To further investigate the relevance of impaired mitochondrial protein quality control in PD, we have explored whether mtHsp70 can act as a genetic modifier of α-synuclein toxicity. It is known that α-synuclein can act as an unfolded substrate for the Hsp70 chaperone system and also deposits as intracellular aggregates in PD-affected brains. Intriguingly, it is known to translocate into mitochondria under conditions of neuronal stress in spite of lacking a canonical mitochondrial signal sequence. Utilizing our yeast-PD model, we find that targeting of α-synuclein A30P disease variant into mitochondria leads to a severe mitochondrial dysfunction phenotype in the wild type Ssc1 background, but not the P486S mutant background. This results in multiple cellular manifestations, which are reversed upon overexpression of the Ssc1 chaperone. Significantly, increasing the J-protein cochaperone availability also leads to reversal of the mutant-associated defects. However, the simultaneous overexpression of both together does not additively improve the protective effects; highlighting the importance of the relative availability of chaperone and cochaperone proteins in preventing aggregation. Our analyses further reveal that while both the wild type and P486S Ssc1 proteins are equally capable of delaying aggregation of α-synuclein, only the wild-type chaperone is better able to prevent aggregation in the presence of its J-protein cochaperone, leading to accumulation of soluble oligomeric species. These observations raised the intriguing possibility, that the reduced chaperoning ability of the proline to serine PD-mutant is, in fact, a compensatory adaptation, favoring the aggregation of α-synuclein over its more toxic soluble oligomeric form. We verify this hypothesis with the aggregation kinetics of A30P α-synuclein, whose intrinsically lower aggregation tendency results in a pronounced delay in aggregation with the wild-type chaperone, thereby strongly favoring the toxic oligomeric species and correlating with the observed lethality in yeast cells. In conclusion, our study provides a model of α-synuclein aggregation-related toxicity and its modulation by the extent of protein quality control within the mitochondrial matrix, through the action of the mtHsp70 chaperone network.

Src family kinases (SFKs) are key regulators of cell proliferation, differentiation, and survival. The expression of these non-receptor tyrosine kinases is strongly correlated with cancer development and tumor progression. Thus, this family of proteins serves as an attractive drug target. The activation of SFKs can occur via multiple signaling pathways, yet many of them are poorly understood. Here, we summarize the current knowledge on G protein-coupled receptor (GPCR)- mediated regulation of SFKs, which is of considerable interest because GPCRs are among the most widely used pharmaceutical targets. This type of activation can occur through a direct interaction between the two proteins or be allosterically regulated by arrestins and G proteins. We postulate that a rearrangement of binding motifs within the active conformation of arrestin-3 mediates Src regulation by comparison of available crystal structures. Therefore, we hypothesize a potentially different activation mechanism compared to arrestin-2. Furthermore, we discuss the probable direct regulation of SFK by GPCRs and investigate the intracellular domains of exemplary GPCRs with conserved polyproline binding motifs that might serve as scaffolding domains to allow such a direct interaction. Large intracellular domains in GPCRs are often understudied and, in general, not much is known of their contribution to different signaling pathways. The suggested direct interaction between a GPCR and a SFK could allow for a potential immediate allosteric regulation of SFKs by GPCRs and thereby unravel a novel mechanism of SFK signaling. This overview will help to identify new GPCR–SFK interactions, which could serve to explain biological functions or be used to modulate downstream effectors.

[EN] Carbamoyl phosphate synthetase 1 (CPS1), a 1462-residue mitochondrial enzyme, catalyzes the entry of ammonia into the urea cycle, which converts ammonia, the neurotoxic waste product of protein catabolism, into barely toxic urea. The urea cycle inborn error and rare disease CPS1 deficiency (CPS1D) is inherited with mendelian autosomal recessive inheritance, being due to CPS1 gene mutations (>200 mutations reported), and causing life-threatening hyperammonemia. We have produced recombinantly human CPS1 (hCPS1) in a baculovirus/insect cell expression system, isolating the enzyme in active and highly purified form, in massive amounts. This has allowed enzyme crystallization for structural studies by X-ray diffraction (an off-shoot of the present studies). This hCPS1 production system allows site-directed mutagenesis and enzyme characterization as catalyst (activity, kinetics) and as protein (stability, aggregation state, domain composition). We have revealed previously unexplored traits of hCPS1 such as its domain composition, the ability of glycerol to replace the natural and essential CPS1 activator N-acetyl-L-glutamate (NAG), and the hCPS1 protection (chemical chaperoning) by NAG and by its pharmacological analog N-carbamyl-L-glutamate (NCG). We have exploited this system to explore the effects on the activity, kinetic parameters and stability/folding of the enzyme, and to test the disease-causing nature, of mutations identified in patients with CPS1 deficiency (CPS1D). These results, supplemented with those obtained with other non-clinical mutations, have provided novel information on the functions of three non-catalytic domains of CPS1. We have introduced three CPS1D-associated mutations and one trivial polymorphism in the glutaminase-like domain of CPS1, supporting a stabilizing and an activity-enhancing function of this non-catalytic domain. Two mutations introduced into the bicarbonate phosphorylation domain have shed light on bicarbonate binding and have directly confirmed the importance of this domain for NAG binding to the distant (in the sequence) C-terminal CPS1 domain. The introduction of 18 CPS1D-associated missense mutations mapping in a clinically highly eloquent central non-catalytic domain have proven the disease-causing nature of most of these mutations while showing that in most of the cases they trigger enzyme misfolding and/or destabilization. These results, by proving an important role of this domain in the structural integration of the multidomain CPS1 protein, have led us to call this domain the Integrating Domain. Finally, we have examined the effects of eight CPS1D-associated mutations, of one trivial polymorphism and of five non-clinical mutations, all of them mapping in the C-terminal domain of the enzyme where NAG binds, whereas we have re-analyzed prior results with another four clinical and five non-clinical mutations affecting this domain. We have largely confirmed the pathogenic nature of the clinical mutations, predominantly because of decreased activity, in many cases due to hampered NAG binding. A few mutations had substantial negative effects on CPS1 stability/folding. Our analysis reveals that NAG activation begins with a movement of the final part of the ß4-¿4 loop of the NAG site. Transmission of the activating signal to the phosphorylation domains involves helix ¿4 from this domain and is possibly transmitted by the mutually homologous loops 1313-1332 and 778-787 (figures are residue numbers) belonging, respectively, to the carbamate and bicarbonate phosphorylation domains. These two homologous loops are called from here on Signal Transmission Loops. / [ES] La carbamil fosfato sintetasa 1 (CPS1), una enzima mitocondrial, cataliza la entrada del amonio en el ciclo de la urea, que convierte esta neurotoxina derivada del catabolismo de las proteínas en urea, mucho menos tóxica. El déficit de CPS1 (CPS1D) es un error innato del ciclo de la urea, una enfermedad rara autosómica recesiva, que se debe a mutaciones en el gen CPS1 (>200 mutaciones descritas) y que cursa con hiperamonemia. Hemos producido CPS1 humana recombinante (hCPS1) en un sistema de expresión de células de insecto y baculovirus, y la hemos aislado en forma activa, muy pura y en cantidad elevada. Este sistema de producción de hCPS1 permite la realización de mutagénesis dirigida y la caracterización de la enzima como catalizador (actividad, cinética) y como proteína (estabilidad, estado de agregación y composición de dominios). Hemos revelado características de la hCPS1 antes no exploradas como es la composición de dominios, la capacidad que tiene el glicerol para reemplazar al activador natural y esencial de la CPS1, N-acetil-L-glutamato (NAG), y la protección de la hCPS1 por NAG y por su análogo farmacológico N-carbamil-L-glutamato (NCG) (chaperonas químicas). Hemos utilizado este sistema para explorar los efectos en actividad, parámetros cinéticos y estabilidad/plegamiento de la enzima, y para comprobar la naturaleza patogénica de mutaciones identificadas en pacientes con CPS1D. Estos resultados, junto con los obtenidos con otras mutaciones no clínicas, han aportado información novedosa sobre tres de los dominios no catalíticos de CPS1. Las observaciones realizadas tras introducir en el dominio de tipo glutaminasa de la enzima tres mutaciones asociadas a CPS1D y un polimorfismo trivial, apoyan la contribución de este dominio no catalítico a la estabilidad y a aumentar la actividad de la enzima. Dos mutaciones introducidas en el dominio de fosforilación de bicarbonato han arrojado luz sobre el modo de unión del bicarbonato (un sustrato). Los resultados de estas mutaciones también han confirmado la contribución de este dominio para la unión de NAG, cuyo sitio de unión se encuentra en el dominio C-terminal de CPS1, bastante alejado (en la secuencia) del dominio de fosforilación de bicarbonato. Además, hemos introducido 18 mutaciones de cambio de sentido asociadas a CPS1D, las cuales están localizadas en un dominio no catalítico, central y de elevada elocuencia clínica. Estos resultados han demostrado la naturaleza patogénica de estas mutaciones, ya que en la mayoría de los casos estas mutaciones producen un mal plegamiento o/y desestabilización de la enzima. Debido a que estos resultados han puesto de manifiesto el importante papel de este dominio en la integración estructural de la proteína multidominio CPS1, lo hemos llamado Dominio Integrador. Finalmente, hemos examinado los efectos de 8 mutaciones asociadas a CPS1D, de un polimorfismo trivial y de 5 mutaciones no clínicas, todas localizadas en el dominio C-terminal de la enzima, donde se une NAG. Además, hemos reanalizado resultados anteriores con otras 4 mutaciones clínicas y 5 no clínicas afectando a este dominio. Hemos confirmado el carácter patogénico de las mutaciones clínicas, las cuales predominantemente causan una disminución en la actividad enzimática, en muchos casos debida a que la unión de NAG se encuentra obstaculizada. Unas pocas mutaciones mostraron efectos negativos en la estabilidad/plegamiento de CPS1. Nuestros análisis revelan que la activación por el NAG empieza con un movimiento de la parte final del bucle ß4-¿4 del sitio de NAG. La transmisión de la señal activadora a los dominios de fosforilación implica a la hélice ¿4 de este dominio y posiblemente se transmite a través de los bucles homólogos 1313-1332 y 778-787 (numeración de residuos) pertenecientes, respectivamente, a los dominios de fosforilación de carbamato y bicarbonato. Por ello, hemos llamado a ambos bucles Bucles de / [CAT] La carbamil fosfat sintetasa 1 (CPS1), un enzim mitocondrial, catalitza l'entrada d'amoni en el cicle de la urea, que convertix l'amoni, producte neurotòxic del catabolisme de les proteïnes, en urea, una molècula molt poc tòxica. El dèficit de CPS1 (CPS1D) és un error innat del cicle de la urea, una malaltia rara autosòmica recessiva, que es deu a mutacions en el gen CPS1 (>200 mutacions descrites) i que cursa amb hiperamonièmia. Hem produït CPS1 humana recombinant (hCPS1) en un sistema d'expressió de cèl·lules d'insecte i baculovirus, i l'hem aïllada en forma activa, molt pura i en gran quantitat. Això ha permés la cristal·lització de l'enzim per a estudis estructurals amb difracció de raios-X (treball no inclòs en esta tesi Aquest sistema de producció de hCPS1 permet la realització de mutagènesi dirigida i la caracterització de l'enzim com a catalitzador (activitat, cinètica) i com a proteïna (estabilitat, estat d'agregació i composició de dominis). Hem revelat característiques de la hCPS1 no explorades abans com és la composició de dominis, la capacitat que té el glicerol per a reemplaçar l'activador natural i essencial de CPS1, N-acetil-L-glutamat (NAG), i la protecció de la hCPS1 per NAG i pel seu anàleg farmacològic N-carbamil-L-glutamat (NCG) (xaperones químiques) . Hem utilitzat aquest sistema per a explorar els efectes en l'activitat, els paràmetres cinètics i l'estabilitat/plegament de l'enzim, i per a comprovar la naturalesa patogènica de mutacions identificades en pacients amb CPS1D. Aquestos resultats, junt amb els obtinguts amb altres mutacions no clíniques, han aportat informació nova sobre tres dels dominis no catalítics de la CPS1. Les observacions, després d'introduir tres mutacions associades a CPS1D i un polimorfisme trivial en el domini tipus glutaminasa de CPS1, recolzen la contribució d'aquest domini no catalític a l'estabilitat i a l'optimització de l'activitat enzimàtica. Dues mutacions introduïdes en el domini de fosforilació de bicarbonat han esclarit el mode d'unió de bicarbonat. Els resultats d'aquestes mutacions també han confirmat la contribució d'aquest domini per a la unió de NAG, el lloc d'unió de la qual es troba en el domini C-terminal de CPS1, prou allunyat (en la seqüència) del domini de fosforilació de bicarbonat. A més, hem introduït 18 mutacions de canvi de sentit associades a CPS1D, les quals estan localitzades en un domini no catalític, central i d'elevada eloqüència clínica. Aquestos resultats han demostrat la naturalesa patogènica d'aquestes mutacions, ja que, en la majoria dels casos produïxen un mal plegament o/i desestabilització de l'enzim. Pel fet que aquestos resultats han posat de manifest l'important paper d'aquest domini en la integració estructural de la proteïna multidomini CPS1, l'hem anomenat Domini Integrador. Finalment, hem examinat els efectes de huit mutacions associades a CPS1D, un polimorfisme trivial i cinc mutacions no clíniques, totes elles localitzades en el domini C-terminal de l'enzim, on s'unix NAG. A més, hem reanalitzat resultats anteriors amb altres quatre mutacions clíniques i cinc no clíniques que afecten aquest domini. Hem confirmat el caràcter patogènic de les mutacions clíniques, les quals predominantment causen una disminució en l'activitat enzimàtica, en molts casos pel fet que la unió de NAG es troba obstaculitzada. Unes poques mutacions van mostrar efectes negatius substancials en l'estabilitat/plegament de CPS1. Les nostres anàlisis revelen que l'activació de NAG comença amb un moviment de la part final del bucle ß4-¿4 del lloc de NAG. La transmissió del senyal activadora als dominis de fosforilació involucra l'hèlix ¿4 d'aquest domini i es transmet, possiblement, a través dels bucles homòlegs 1313-1332 i 778-787 (numeració dels residus), pertanyents, respectivament, als dominis de fosforilació de carbamato i bicarbonat. Per això, hem anomenat a ambd / Díez Fernández, C. (2015). USING RECOMBINANT HUMAN CARBAMOYL PHOSPHATE SYNTHETASE 1 (CPS1) FOR STUDYING THIS ENZYME'S FUNCTION, REGULATION, PATHOLOGY AND STRUCTURE [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/52855 / TESIS