Schizophrenia is a chronic psychotic illness affecting ~21 million people globally. Currently available antipsychotic agents act through a dopamine D2 receptor mechanism, and produce extrapyramidal or metabolic side effects. Hence, there is a need for novel targets and agents. The mGlu2/5-HT2A receptor heteromer has been implicated in the action of antipsychotic agents, and represents a novel and attractive therapeutic target for the treatment of schizophrenia. A long-term goal of this project is to synthesize bivalent ligands where a 5-HT2A receptor antagonist is tethered to an mGlu2 PAM via a linker. The goals of the investigation were to study the SAR of risperidone (an atypical antipsychotic agent) at 5-HT2A receptors using a “deconstruction-reconstruction-elaboration” approach to determine the minimal structural features of risperidone that contribute to its 5-HT2A receptor affinity and antagonism, and to determine where on the “minimized risperidone” structure an mGlu2 PAM can be introduced. Additional goals included studying the binding modes of various mGlu2 PAMs and identifying where on an mGlu2 PAM a risperidone “partial” structure could be introduced. Biological studies of deconstructed/elaborated analogs of risperidone suggest that the entire structure of risperidone is not necessary for 5-HT2A receptor affinity and antagonism, and that a fluoro group contributes to 5-HT2A binding. 6-Fluoro-3-(4-piperidinyl)-1,2-benz[d]isoxazole that has only half the structural features of risperidone retains 5-HT2A receptor affinity and antagonist activity, and represents the “minimized risperidone” structure with the piperidine nitrogen atom representing a potential linker site for eventual construction of bivalent ligands. Molecular modeling studies at 5-HT2A receptors suggest that risperidone and its analogs have more than one binding mode. Modeling studies to evaluate binding modes of various PAMs at mGlu2 receptors, coupled with known SAR information, were used to identify a PAM (JNJ-40411813), and the pyridone nitrogen atom of JNJ-40411813 as a potential linker site. Additionally, potential synthetic routes for JNJ-40411813 were explored that might be of value in the synthesis of bivalent ligands. Based on the structural features of 6-fluoro-3-(4-piperidinyl)-1,2-benz[d]isoxazole, a new pharmacophore for 5-HT2A receptor antagonists, consisting of one aromatic region, a basic protonated amine and hydrogen bond acceptors, has been proposed.

Sulfated low molecular weight lignins (LMWLs), CDSO3 and FDSO3, designed recently as macromolecular mimetics of heparin, were found to exhibit potent anticoagulant activity. Small molecules based on the same scaffold, SBD and SBT, showed promising thrombin inhibition. We were able to address the mechanism of the inhibition using Michaelis-Menten kinetics. All the molecules were found to be allosterically impairing thrombin activity using either noncompetitive or uncompetitive mechanism. Absence of competition with hirugen, an exosite 1 ligand, and competition with polymeric heparin points to exosite 2 as the site of interaction for these inhibitors. Yet mixed competition results with other exosite 2 ligands indicated that the molecules utilize different sub-sites within exosite 2 for interaction. Site-directed mutagenesis was used to pin point the key residues important for inhibition. All of all positively charged exosite 2 residues were mutated one at a time to alanine to abolish its charge. The data showed that Arg93 and Arg175 are the major residues involved in CDSO3 binding. FDSO3 showed a progressively greater defect in inhibition with double point mutations, the triple mutant Arg93,97,101Ala displayed a 50 fold drop in inhibition. A single mutant, Arg173Ala, displayed 22-fold reduction in IC50 of SBD, while Arg233Ala was the only mutation that impaired SBT inhibition. This proves the fact that inspite of the structural similarity between the two polymers and the two small molecules, thtey do not share the same binding space in exosite 2. To understand the types of interactions involved in thrombin interaction with the polymers, we resorted to salt-dependence studies. This showed that CDSO3 had fewer ionic contacts with thrombin, with most of its binding energy derived from non-ionic interactions. FDSO3 on the other hand had a balanced contribution of ionic and non-ionic forces. Thermodynamic studies showed that both polymers have a positive ΔCp of binding, which proves the involvement of electrostatic forces and signals the burial of the polar residues on thrombin exosite 2. These molecules offer a rare chance to study thrombin allostery. Little is known about the interplay between exosite 2, active site and sodium binding site. The allosteric nature of inhibition indicated that, for the first time, a link is proven to exist between exosite 2 and the active site that could be used to inhibit the enzyme. The presence of sodium was found to enhance the binding of FDSO3 at exosite 2, which establish the energetic coupling between exosite 2 and sodium binding site. The results identify novel binding sub-sites within exosite 2 that are energetically coupled to thrombin’s catalytic function and linked to the sodium binding site. The design of high affinity small molecules based on LMWLs scaffold presents major opportunities for developing clinically relevant, allosteric modulators of thrombin.

Les moteurs moléculaires sont des protéines capables de produire une force : elles transforment l'énergie chimique de l'hydrolyse de l'ATP en énergie mécanique. Cette thèse se focalise sur l'étude d'une famille de moteurs moléculaires, les myosines, qui se déplacent le long des filaments d'actine et assurent d'importantes fonctions cellulaires.La myosine VI est une myosine très particulière car elle est la seule à se déplacer vers l'extrémité négative des filaments d'actine. Elle est produite dans la cellule sous forme auto-inhibée, inactive. Dans la cellule, son activité est également régulée par plusieurs protéines interagissant avec la queue C-terminale de la myosine VI. Ces protéines, présentes à des endroits précis de la cellule, recrutent la myosine VI et dictent l'action qu'elle doit effectuer. Des analyses de SAXS, de dispersion de la lumière, de microscopie, d'interaction et de mutagénèse ont permis de mieux comprendre le mécanisme régulant l'adoption de l'état auto-inhibé, ainsi que son activation par le calcium. L'interaction avec différents partenaires a été caractérisée. GIPC1, le partenaire le plus étudié, promeut de façon indirecte la dimérisation et l'activation de la myosine VI.Pendant ma thèse, j'ai également été impliqué dans deux autres projets qui s'inscrivent dans la logique du projet de thèse et qui ont mené à la publication de quatre articles. Deux chapitres, plus brefs, sont donc dédiés à ces projets. Le deuxième chapitre porte sur la régulation de l'activité de la myosine VII par ses partenaires cellulaires. Finalement, le troisième chapitre est dédié à l'étude de la modification allostérique de l'activité des myosines par des petites molécules. / Molecular motors are essential agents of force production in the cells: they convert the chemical energy released by the hydrolysis of ATP into mechanical work. This thesis focuses on myosins, a family of molecular motors responsible for actin-based motility. Myosin VI is unique among all of the myosin superfamily members in that it moves in the opposite direction of all other known myosins. Previous work revealed myosin VI tail ability to fold back, constituting a potential auto-inhibited state that prevents motor activity. Several myosin VI partners, binding to the C-terminal tail of the myosin, have been identified and shown to recruit the motor for different functions. In the first chapter of this thesis, the mechanism allowing the regulation of myosin VI activity has been studied using biochemical and biophysical analysis (SAXS, light scattering, microscopy, binding assays and mutagenesis). GIPC1, the most studied myosin VI partners, promotes myosin dimerization and activation. During my PhD, I have been also involved in two other projects, in line with my thesis project, that have led to the publication of four articles. Two shorter chapters are therefore devoted to these projects. The second chapter of my thesis explores myosin VII activity regulation by its cellular partners. Finally, the third chapter is devoted to the allosteric regulation of myosins activity by small molecules.

<p>The neurosteroids constitute a group of powerful hormones synthesized and acting in the central nervous system. They participate in a number of important central processes, such as memory and learning, mood and neuroprotection. Their effects emerge from rapid interactions with membrane bound receptors, such as the N-methyl-D-aspartate (NMDA) receptor, the gamma-amino-butyric acid receptor and the sigma 1 receptor. The mechanisms of action are separate from classical genomic interactions. </p><p>The aims of this thesis were to identify and characterize the molecular mechanisms underlying the effects of nanomolar concentrations of neurosteroids at the NMDA receptor. </p><p>The results show that the neurosteroids pregnenolone sulfate (PS) and pregnanolone sulfate 3α5βS) differently modulate the NMDA receptor, changing the kinetics for the NMDA receptor antagonist ifenprodil, through unique and separate targets at the NR2B subunit. The effects that appear to be temperature independent were further confirmed in a calcium imagining functional assay. A second functional study demonstrated that PS and 3α5βS affect glutamate-stimulated neurite outgrowth in NG108-15 cells. </p><p>Misuse of anabolic androgenic steroids (AAS) has powerful effects on emotional states. Since neurosteroids regulate processes involved in mood it can be hypothesised that AAS can interact with the action of neurosteroids in the brain. However, chronic administration of the AAS nandrolone decanoate did not alter the allosteric effects of PS or 3α5βS at the NMDA receptor, but changed the affinity for PS, 3α5βS and dehydroepiandrosterone sulfate to the sigma 1 receptor. The results also showed that the neurosteroids displace ³H-ifenprodil from the sigma 1 and 2 receptors without directly sharing the binding site for ³H-ifenprodil at the sigma 1 receptor. The decreased affinity for the neurosteroids at the sigma 1 receptor may be involved in the depressive symptoms associated with AAS misuse.</p><p>The NMDA receptor system is deeply involved in neurodegeneration and the NMDA receptor antagonist ifenprodil exert neuroprotective actions. The findings that neurosteroids interact with ifenprodil at the NMDA receptor may be an opportunity to obtain synergistic effects in neuroprotective treatment.</p>

The neurosteroids constitute a group of powerful hormones synthesized and acting in the central nervous system. They participate in a number of important central processes, such as memory and learning, mood and neuroprotection. Their effects emerge from rapid interactions with membrane bound receptors, such as the N-methyl-D-aspartate (NMDA) receptor, the gamma-amino-butyric acid receptor and the sigma 1 receptor. The mechanisms of action are separate from classical genomic interactions. The aims of this thesis were to identify and characterize the molecular mechanisms underlying the effects of nanomolar concentrations of neurosteroids at the NMDA receptor. The results show that the neurosteroids pregnenolone sulfate (PS) and pregnanolone sulfate 3α5βS) differently modulate the NMDA receptor, changing the kinetics for the NMDA receptor antagonist ifenprodil, through unique and separate targets at the NR2B subunit. The effects that appear to be temperature independent were further confirmed in a calcium imagining functional assay. A second functional study demonstrated that PS and 3α5βS affect glutamate-stimulated neurite outgrowth in NG108-15 cells. Misuse of anabolic androgenic steroids (AAS) has powerful effects on emotional states. Since neurosteroids regulate processes involved in mood it can be hypothesised that AAS can interact with the action of neurosteroids in the brain. However, chronic administration of the AAS nandrolone decanoate did not alter the allosteric effects of PS or 3α5βS at the NMDA receptor, but changed the affinity for PS, 3α5βS and dehydroepiandrosterone sulfate to the sigma 1 receptor. The results also showed that the neurosteroids displace 3H-ifenprodil from the sigma 1 and 2 receptors without directly sharing the binding site for 3H-ifenprodil at the sigma 1 receptor. The decreased affinity for the neurosteroids at the sigma 1 receptor may be involved in the depressive symptoms associated with AAS misuse. The NMDA receptor system is deeply involved in neurodegeneration and the NMDA receptor antagonist ifenprodil exert neuroprotective actions. The findings that neurosteroids interact with ifenprodil at the NMDA receptor may be an opportunity to obtain synergistic effects in neuroprotective treatment.

The glycine receptor (GlyR) is a ligand-gated ion channel member of the cys-loop receptor superfamily, responsible for inhibitory neurotransmission in the brain and spinal cord. Both glycine and the partial agonist taurine act as endogenous ligands of the GlyR. Taurine-activated GlyR may have a role in the rewarding effects of drugs of abuse, such as ethanol. As a partial agonist, taurine has a decreased efficacy relative to glycine, resulting in a decreased maximum response. We investigated the effects of ethanol, anesthetics, inhalants, and zinc to determine if these allosteric modulators could increase the efficacy of the taurine-activated GlyR. Whole cell recordings of wild type GlyR revealed that each of the allosteric modulators potentiated currents generated by saturating concentrations of taurine but not glycine, implying an increase in efficacy. Zinc is found at GlyR-potentiating concentrations throughout the nervous system, so we examined the combinatorial effects of these allosteric modulators with zinc to mimic in vivo conditions. Whole cell recordings revealed that zinc potentiation of saturating taurine-generated currents decreased further potentiation by another allosteric modulator, indicating no synergistic effects on efficacy. We next investigated the actions of ethanol and isoflurane on the taurine-activated GlyR at the single channel level, finding that both allosteric modulators stabilized the channel open state, increasing the efficacy of the taurine-activated GlyR. We previously identified a mutation in the ligand-binding domain of the GlyR (D97R) that produces spontaneously activating channels, on which taurine has increased efficacy. We identified a residue, R131, as a possible binding partner of D97 in forming an electrostatic interaction that holds the channel in the closed state. We found that disruption of this interaction results in greatly increased taurine efficacy, indicating that efficacy for partial agonists may be determined by agonist ability to break this bond early in the activation process following binding. Thus we find differential mechanisms of allosteric modulation and efficacy determinations for the GlyR when activated by taurine vs. glycine. / text

The glycine receptor (GlyR) is a ligand-gated ion channel member of the cys-loop receptor superfamily, responsible for inhibitory neurotransmission in the brain and spinal cord. Both glycine and the partial agonist taurine act as endogenous ligands of the GlyR. Taurine-activated GlyR may have a role in the rewarding effects of drugs of abuse, such as ethanol. As a partial agonist, taurine has a decreased efficacy relative to glycine, resulting in a decreased maximum response. We investigated the effects of ethanol, anesthetics, inhalants, and zinc to determine if these allosteric modulators could increase the efficacy of the taurine-activated GlyR. Whole cell recordings of wild type GlyR revealed that each of the allosteric modulators potentiated currents generated by saturating concentrations of taurine but not glycine, implying an increase in efficacy. Zinc is found at GlyR-potentiating concentrations throughout the nervous system, so we examined the combinatorial effects of these allosteric modulators with zinc to mimic in vivo conditions. Whole cell recordings revealed that zinc potentiation of saturating taurine-generated currents decreased further potentiation by another allosteric modulator, indicating no synergistic effects on efficacy. We next investigated the actions of ethanol and isoflurane on the taurine-activated GlyR at the single channel level, finding that both allosteric modulators stabilized the channel open state, increasing the efficacy of the taurine-activated GlyR. We previously identified a mutation in the ligand-binding domain of the GlyR (D97R) that produces spontaneously activating channels, on which taurine has increased efficacy. We identified a residue, R131, as a possible binding partner of D97 in forming an electrostatic interaction that holds the channel in the closed state. We found that disruption of this interaction results in greatly increased taurine efficacy, indicating that efficacy for partial agonists may be determined by agonist ability to break this bond early in the activation process following binding. Thus we find differential mechanisms of allosteric modulation and efficacy determinations for the GlyR when activated by taurine vs. glycine. / text

In 1996, scientists discovered a connection between the gene for the human protein frataxin (FXN) and the neurodegenerative disease Friedreich’s ataxia (FRDA). Decreased FXN levels result in a variety of aberrant phenotypes including loss of activity for iron-sulfur containing enzymes, mitochondrial iron accumulation, and susceptibility to oxidative stress. These symptoms are the primary focus of current therapeutic efforts. In contrast our group is pursuing an alternate strategy of first defining FXN function at a molecular level then using this information to identify small molecule functional replacements. Recently, our group has discovered that FXN functions as an allosteric activator for the human Fe-S cluster assembly complex. The work presented here helps to further define molecular details of FXN activation and explain how FRDA missense mutants are functionally compromised. First, the FRDA missense mutants L182H and L182F were investigated. Unlike other characterized FRDA missense mutants, the L182F variant was not compromised in its ability to bind and activate the Fe-S assembly complex. The L182H variant exhibited an altered circular dichroism signature; suggesting a change in secondary structure relative to native FXN, and rapidly degraded. Together these studies suggest that L182 variants are less stable than native FXN and are likely prone to degradation in FRDA patients. Second, as a regulatory role of FXN suggests that its function is likely controlled by environmental stimuli, different maturation forms of FXN as well as post-translational modification mimics were tested as mechanisms to control FXN regulation. Here experiments were designed to test if a larger polypeptide form of FXN represents a functional form of the protein. Kinetic and analytical ultracentrifugation studies revealed a complex heterogeneous mixture of species some of which can activate the Fe-S assembly complex. A previously identified acetylation site was also tested using mutants that mimic acetylation. These mutants had little effect on the ability of FXN to bind and activate the assembly complex. Third, mutagenesis experiments were designed in which the FXN surface residues were replaced with alanine and the resulting variants were tested in binding and activity assays. These experiments revealed a localized “hot-spot” on the surface of FXN that suggests small cyclic peptide mimics might be able to replace FXN and function as FRDA therapeutics. Unexpectedly, one of the FXN variants exhibited significantly tighter binding and could have relevance for therapeutic development.

L’approche Module X a été créée dans le but de concevoir de petits peptides modulateurs ayant des propriétés allostériques. Module X reproduit de petites parties des portions extracellulaires flexibles des récepteurs. Ces petits peptides vont interagir en s’interposant entre deux sous unités ou entre deux régions de la même sous-unité qui interagissent par des liens hydrogènes, des ponts salins ou des liens disulfure. Ces régions sont spécialement choisies à l’extérieur du domaine de liaison du ligand orthostérique et sont situées dans les régions inter domaines, la portion juxta membranaire ou dans les boucles. Étant donné que les boucles sont exposées durant les changements de conformation, une séquence peptidique reproduisant certaines régions de ces boucles pourrait s’insérer à un endroit approprié dans la structure où se lier à son partenaire de signalisation dans le complexe protéique, ce qui aurait comme effet de déplacer l’équilibre de l’ensemble vers un état particulier et modulerait ainsi la signalisation. De cette façon, certaines voies de signalisation pourraient être partiellement inhibées tandis que d’autres voies ne seraient pas touchées puisque le ligand orthostérique pourrait toujours se lier au récepteur. Dans une première étude, nous avons conçu des peptides inhibiteurs du récepteur de l’interleukine 1 (IL-1R/IL-1RAcP) plus précisément en reproduisant des régions flexibles de la protéine accessoire, sous-unité signalisatrice du récepteur. IL-1 est un médiateur majeur de l’inflammation, mais le seul antagoniste disponible est l’analogue naturel de IL-1, IL-1Ra qui compétitionne avec IL-1 pour le site de liaison sur le récepteur. Nous avons conçu plusieurs peptides à partir des boucles de la protéine accessoire. Un de ces peptides, rytvela (101.10) a démontré des propriétés de non-compétitivité et de sélectivité fonctionnelle caractéristiques des modulateurs allostériques. 101.10 bloque la prolifération des thymocytes et la synthèse de PGE2 avec un IC50 de 1 nM mais une efficacité de 100 % et 45 % respectivement et ne déplace pas IL-1 radioactif dans des essais de radioliaisons. De plus, 101.10 n’a qu’un effet minime sur l’affinité de IL-1 pour son récepteur. 101.10 démontre, de plus, une activité inhibitrice in vivo dans des modèles d’inflammation de l’intestin chez le rat (efficacité supérieure aux corticostéroïdes et à IL-1Ra) et de dermatite chez la souris de même que dans un modèle d’hyperthermie induite par IL-1. La deuxième étude démontre que Module X peut être utilisé pour concevoir des inhibiteurs pour une autre grande famille de récepteurs : les récepteurs couplés aux protéines G. La vasopressine joue un rôle important dans l’équilibre hydro-osmotique et un moindre rôle dans la vasomotricité. Six peptides ont été conçus à partir de régions juxta membranaires du récepteur de la vasopressine V2R. Le peptide le plus actif, VRQ397 (IC50 = 0,69 nM dans un modèle de vasorelaxation du crémastère), a démontré de la sélectivité fonctionnelle en inhibant la synthèse de prostacycline mais sans inhiber l’activation de la protéine Gs et la génération d’ AMP cyclique. Le peptide VRQ397 ne pouvait déplacer le ligand naturel AVP marqué radioactivement; de même VRQ397 radioactif ne se liait que sur V2R et non pas sur d’autres récepteurs de la même famille tel que V1R (récepteur de la vasopressine de type I). Ces études décrivent la caractérisation de petits peptides modulateurs de la signalisation de IL-1R et V2R et présentant des propriétés de modulateurs allostériques. / The Module X approach was conceived to generate small allosteric peptides that do not (by definition) compete with the natural ligand to inhibit or modulate signalling. Orthosteric inhibition blocks the entire signalling pathways while allosteric modulators will bind to another site on the target and show functional selectivity. By reproducing parts of the flexible regions (loops) of two receptors, the IL-1 and vasopressin receptors, we generated small peptides that showed allosteric properties. To prove our concept we started with a pro-inflammatory target: IL-1 receptor. Interleukin (IL)-1 is a major pro-inflammatory cytokine which interacts with the IL-1 receptor I (IL-1RI) complex, composed of IL-1RI and IL-1R accessory protein (IL-1RacP) subunits. Presently, there are no small antagonists of the IL-1RI complex. Given this void, we derived 15 peptides from loops of IL-1RacP, which are putative interactive sites with the IL-1RI subunit. Here we substantiate the merits of one of these peptides, rytvela (we termed, 101.10), as an inhibitor of IL-1R and describe its properties consistent with those of an allosteric negative modulator. 101.10 (IC50  1 nM) blocked human thymocyte proliferation in vitro, and demonstrated robust in vivo effects in models of hyperthermia and inflammatory bowel disease as well as topically in contact dermatitis, superior to corticosteroids and IL-1ra; 101.10 did not bind to IL-1RI deficient cells and was ineffective in vivo in IL-1RI knockout mice. Importantly, characterization of 101.10, revealed non-competitive antagonist actions and functional selectivity by blocking certain IL-1R pathways while not affecting others. The second study involved a representative of the biggest family of membrane proteins: G-protein coupled receptors. Vasopressin type 2 receptor (V2R) exhibits mostly important properties for hydro-osmotic equilibrium and to a lesser extent on vasomotricity. Drugs currently acting on this receptor are analogs of the natural neuropeptide, vasopressin (AVP), and hence are competitive ligands. Six peptides reproducing juxtamembranous regions of V2R were designed and screened; the most effective peptide, CRAVKY (labelled VRQ397), was characterized. VRQ397 was potent (IC50 = 0.69 ± 0.25 nM) and fully effective in inhibiting V2R-dependent physiological function (specifically DDAVP-induced cremasteric vasorelaxation; this physiological functional assay was utilized to avoid overlooking interference of specific signaling events). Dose-response profile revealed non-competitive property of VRQ397; correspondingly, VRQ397 bound specifically to V2R-expressing cells, could not displace its natural ligand, AVP, but modulated AVP binding kinetics (dissociation rate). VRQ397 exhibited pharmacological permissiveness on V2R-induced signals as it inhibited DDAVP-induced PGI2 generation, but not that of cAMP or recruitment of — arrestin2. Consistent with in vitro and ex vivo effects as a V2R antagonist, VRQ397 displayed anticipated in vivo aquaretic efficacy. Findings describe the discovery of potent and specific small (peptide) antagonists of IL-1RI and V2R with properties in line with an allosteric negative modulator.

Thesis (Ph.D.) -- University of Texas Southwestern Medical Center at Dallas, 2006. / Not embargoed. Vita. Bibliography: 158-177.