Peanuts (<i>A. hypogaea</i>) are an economically important crop, a source of food allergies and a member of the South American genus <i>Arachis</i>. The eighty species of genus <i>Arachis</i> have been divided into nine sections. The largest, section <i>Arachis</i>, has been further subdivided into three genome groups. The current intuitive understanding of the evolutionary relationships among <i>Arachis</i> is based on morphological, geographic and cytogenetic data, but a comprehensive phylogenetic study for the genus is lacking. A total of 48 species representing all nine sections were used to reconstruct a phylogeny based on sequence information from plastid <i>trn</i>T-<i>trn</i>F and nuclear ITS genomic regions. Phylogenetic analysis resolved section <i>Extranervosae</i> at the base, followed by sections <i>Triseminatae</i> and <i>Caulorrhizae</i>. Two major terminal lineages were recovered. One is comprised of sections <i>Erectoides</i>, <i>Heteranthae</i>, <i>Procumbentes</i>, Rhizomatosae</i>, and <i>Trierectoides</i>, referred to here as group erectoides. The other is comprised of two major clades, arachis I (B genome, D genome, and aneuploid species) and arachis II (A genome species). The phylogenetic trees show that sequence data partially agrees with the relationships described in the monograph; however, some further investigation is necessary to clarify relationships within and among species of the two terminal lineages. In addition, the major allergen Ara h 2 from 12 wild species from across the genus was analyzed for mutations that could potentially produce a hypoallergenic ortholog. It was found that the evolution of the allergen mostly reflected the species phylogenies based on ITS and combined. The majority of substitutions and length variations were concentrated in the loop connecting helices H2 and H3. Section <i>Arachis</i> species tended to have larger H2-H3 loops, while those from other sections had shorter loops. The immunodominant epitopes #6 and #7, located within this loop, tended to contain mutations or were truncated among species outside of section <i>Arachis</i>. Dot immunoblots showed reduced IgE-binding to peptides representing portions of the H2-H3 loop from <i>A. guarantica</i> and <i>A. triseminata</i>. Orthologs from wild species have demonstrated that they could potentially contain variations of the allergen Ara h 2 that could be utilized to develop a safer peanut cultivar. / Ph. D.

acase@tulane.edu

Il existe plusieurs méthodes de radiomarquage de protéines avec du fluor-18 utilisées en imagerie de tomographie par émission de positons (TEP). Nous travaillons sur le développement d'une méthode de radiomarquage d'Ara h 2, une protéine d'arachide responsable de réactions anaphylactiques. Actuellement, il n’existe pas de méthodes permettant l’évaluation in vivo de la pharmacocinétique des allergènes alimentaires après ingestion. Nous avons utilisé un module automatisé avec cassettes pour préparer le groupement prosthétique radiomarqué ([18F]F-Py-TFP). Le précurseur réagit avec les ions fluorure (18F-) dans une cartouche de méthylammonium quaternaire (QMA). Le [18F]FPy-TFP est élué via QMA et une cartouche d'échange de cations en mode mixte (MCX) avec de l'éther. Après l’évaporation de l'éther, le [18F]FPy-TFP réagit avec de l'albumine sérique bovine (BSA) (ou Ara h 2) dans du DMSO, de l'eau et un tampon borate. Dans une autre méthode, la BSA lyophilisée (ou Ara h 2) est dissoute avec [18F]F-Py-TFP dans le DMSO en présence de DIPEA. Dans les deux protocoles, la réaction est chauffée pendant 15 minutes à 40 °C. Le produit est analysé par chromatographie liquide de haute performance. Ces réactions en deux étapes ont été optimisées séparément pour améliorer le rendement radiochimique (RCY). Le [18F]FPy-BSA a été préparé avec succès avec un RCY de 24 % ± 4 (décroissance corrigée) et une activité molaire de 1 196 Ci/mmol. Après plusieurs tentatives, nous n'avons pas réussi le radiomarquage de l’Ara h 2, mais d’autres études de radiofluoration sont en cours de développement. / There are several methods for radiolabeling proteins with fluorine-18 for positron emission tomography (PET) imaging. We are working to develop a clinically relevant reproducible method of radiolabeling Ara h 2, an allergenic peanut protein producing anaphylactic reactions. There is currently no accurate method to assess in vivo the pharmacokinetics of food allergens after ingestion. In an automated cassette-based module, the precursor was added and reacts with dried reactive 18F within a quaternary methyl ammonium (QMA) cartridge. Fluoropyridine-tetrafluorophenyl ester ([18F]FPy-TFP) is then eluted through the QMA and a mixed-mode cation exchange (MCX) cartridges with ether. Following ether evaporation, dissolved [18F]FPy-TFP in DMSO is added to bovine serum albumin (BSA) (or Ara h 2) in a borate buffer and water. In another method, lyophilized BSA (or lyophilized Ara h 2) is dissolved with dried [18F]FPy-TFP in DMSO and DIPEA. In both procedures, the reaction mixture is heated for 15 minutes at 40°C and analyzed by analytical high performance liquid chromatography. [18F]FPy-BSA was successfully produced from [18F]FPy-TFP in a radiochemical yield (RCY) up to 24% ± 4 (decay corrected) and in molar activity higher than 1 196Ci/mmol. These two-step reactions were separately optimized to maximize the RCY and provide low toxicity in future PET imaging studies. Following similar approaches as BSA, Ara h 2 radiolabeling was unsuccessful with both methods. Work is currently in progress for labeling Ara h 2.

Molecular modeling is a term referring to the study of proteins, nucleic acids, lipids, and other bio- or macro- or small molecules at the atomistic level using a combination of computational methods, physico-chemical principles, and mathematical functions. It can be generally sub-divided into two areas: molecular mechanics, which is the treatment of atoms and bonds as Newtonian particles and springs, and quantum mechanics, which models electronic behaviors using the Schrödinger equation and wavefunctions. Each technique is a powerful tool that, when used alone or in combination with wet lab experiments, can yield useful results, the products of which have broad applications in studying human disease models, oxidative damage, and other biomolecular processes that are otherwise not easily observed by experiment alone. Within this document, we study seven different such systems. This includes the mode of inhibitor binding to the enzyme monoamine oxidase B, the active site mechanism of that same enzyme, the dynamics of the unstructured p53 C-terminal domain in complex with globular, structured proteins, the process of the viral protein B2 unbinding from double-stranded RNA, and a focus on the dynamics of a variable loop in the antigenic peanut protein Ara h 2. In addition to those conventional molecular modeling studies, several of which were done in tandem with wet lab experiment, we also discuss the validation of charges and charge group parameters for small molecules used in molecular mechanics, and the development of software for the analysis of lipid bilayer systems in molecular mechanics simulations. As computational resources continue to evolve, and as more structural information becomes available, these methods are becoming an integral part of the study of biomolecules in the context of disease. / Ph. D.