Orientador: Gabriela Castellano / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Fisica Gleb Wataghin. / Made available in DSpace on 2018-08-12T12:39:12Z (GMT). No. of bitstreams: 1 CuellarBaena_SandraPatricia_M.pdf: 2971806 bytes, checksum: 993051b37ed11a93fc4c48a83e24003d (MD5) Previous issue date: 2008 / Resumo: Este trabalho visou o estudo e validação de técnicas de pre-processamento e quantificação de dados provenientes da técnica de Espectroscopia por Ressonância Magnética (MRS, do inglês Magnetic Resonance Spectroscopy), obtidos do cérebro humano in vivo, para a extração de informação que fosse clinicamente relevante para o estudo e diagnostico de tumores cerebrais. Para isso, foi feito o estudo da técnica com base na literatura, incluindo a revisão dos aspectos físicos envolvidos, estudando os métodos computacionais utilizados para o pre-processamento e quantificação dos dados, e os aspectos bioquímicos dos metabólicos de interesse presentes no cérebro humano, passiveis de serem quantificados através da técnica. Especificamente, foi estudado um método de quantificação de dados de MRS, o método. AMARES (Advanced Method for Accurate, Robust and Efficient Spectral fitting of MRS data), aplicado na quantificação de dados de MRS adquiridos de sujeitos controles e pacientes portadores de tumores cerebrais, provenientes de uma base de dados do Laboratório de Neuroimagem (LNI - Hospital das Clinicas - UNICAMP). Isso foi feito utilizando o software de domínio público jMRUI (http://sermn02.uab.es/mrui/)[1], que possui o método AMARES já implementado. Estes resultados foram comparados com resultados provenientes de uma quantificação manual desses mesmos dados, realizada previamente como parte do projeto de doutorado da Dra. Andréia Vasconcellos (atual docente do Depto. de Radiologia da FCM/UNICAMP)[2]. Foi verificada a concordância entre os dois métodos de quantificação, e também a viabilidade de usar os resultados da quantificação com o método automático para alem de diferenciar entre os grupos de pacientes e controles, realizar a separação dos Pacientes com tumores em diferentes grupos. Obteve-se que os resultados obtidos com o método automático foram mais precisos e consistentes que os obtidos com o método manual, e permitiram uma melhor classificação dos tipos de tumores. Adicionalmente, foram incluídos neste trabalho os resultados do estudo de perfis metabólicos ex vivo em tumores cerebrais pediátricos através da técnica HR-MAS (do inglês High Resolution Magic Angle Spinning). Este estudo adicional foi realizado no Laboratório de Imagem Molecular da Faculdade de Medicina da Universidade de Valencia (Espanha) através do Programa Santander de Mobilidade Internacional e financiado através de uma bolsa do Banco Santander-Banespa. / Abstract: The aim of this work was to study and validate techniques for pre-processing and quantificating Magnetic Resonance Spectroscopy data, obtained in vivo from the human brain, in order to get information clinically useful for the study and diagnosis of brain tumors. Therefore, a literature-based study of the technique was made, including a review of the Physics concepts involved, the data acquisition process in the scanner and the computational methods used to pre-process and quantificate the spectral data, as well as the biochemical aspects of the metabolites of interest in the human brain that can be detected by this technique. Special attention was given to the AMARES (Advanced Method for Accurate, Robust and Efficient Spectral fitting of MRS data) method for MRS data quantification, which was studied and applied to the quantification of data from control subjects and patients with brain tumors. The data came from a database of the Neuroimaging Laboratory (LNI - Hospital das Clinicas - UNICAMP). The quantification with AMARES was made through the jMRUI software (http://sermn02.uab.es/mrui/) [1], a public domain software for processing and quantification of MRS data. These results were compared to the results obtained with a manual quantification of the same data, previously done as part of the PhD thesis work of Dr. Andreia Vasconcellos (lecturer from the Radiology Department of the School of Medicine, UNICAMP) [2]. The agreement between the results from both quantification methods was verified, as well as the feasibility of using the automatic quantification results to differentiate among tumor types, besides differentiating between patients and controls. Results obtained by the automatic method were more accurate and consistent than those obtained by the manual method allowing a better classification. Additionally, in this work were included the results of the study of ex vivo and in vivo metabolic profiling in pediatric brain tumors using the HR-MAS (High Resolution Magic Angle Spinning) technique. This study was carried out in the Molecular Imaging Laboratory, School of Medicine at the University of Val¿encia (Spain), within the Santander-Banespa Bank International Exchange Program. / Mestrado / Física / Mestre em Física

Incident ¹H⁺ and ⁴He⁺ ions at 0.3-2.6 MeV and ¹⁹F^q⁺ ions at 25, 27 and 35 MeV were used to study the M-shell x-ray production cross sections of Au, Pb, Bi and U. For the incident fluorine ions, projectile charge state dependence of the cross sections were extracted from measurements made with varying target thicknesses ( ~1 to ~300 μg/cm²). The efficiency of the Si(Li) detector was determined by measuring the K-shell x-ray production of various low Z elements and comparing these values to the prediction of the CPSS theory. The experimental results are compared to the prediction of first Born approximation for direct ionization to the continuum and to the OBK of Nikolaev for the electron capture to the K-, L-, M-...shells of the incident ion. Comparison is also made with the ECPSSR theory that accounts for the energy loss, Coulomb deflection, and relativistic effects in the perturbed stationary state theory.

Le nombre de cibles potentielles pour la conception de nouvelles molécules à activité thérapeutique ne cesse de croître. Pour chaque cible, il est nécessaire d’identifier des molécules actives et de les optimiser afin d’atteindre l’affinité et la sélectivité recherchées. Ces nouveaux défis accentuent la nécessité d’améliorer notre compréhension des facteurs qui mènent à la reconnaissance moléculaire entre une drogue potentielle et une macromolécule biologique, et particulièrement des facteurs énergétiques à la base de la stabilisation d’un complexe d’interaction. Dans le cadre de ce travail, nous nous sommes intéressés à l’effet que pouvaient avoir les équilibres de protonation/déprotonation des résidus ionisables d’une protéine sur les paramètres thermodynamiques caractérisant la complexation d’un ligand. Dasns ce but, nous avons étudié l’interaction entre l’α-chymotrypsine et un de ses inhibiteurs compétitifs, la proflavine. Cette protéine est représentative d’un nombre important d’enzymes présentant le même mécanisme catalytique. La compréhension des facteurs qui régissent les équilibres de protonation/déprotonation des résidus ionisables présents dans son site actif ainsi que de l’effet sur ceux-ci de l’interaction avec des ligands est d’une importance primordiale pour le développement d’inhibiteurs plus sélectifs de ces protéases.<p>Cette étude s’est essentiellement composée de trois volets. (i) La réalisation d’un modèle du complexe d’interaction afin de confronter des données structurales aux données expérimentales recueillies. (ii) L’étude de l’interaction entre l’α-chymotrypsine et la proflavine par spectroscopie de Résonance Magnétique Nucléaire (RMN) afin de mettre en évidence les résidus ionisables dont les équilibres de protonation/déprotonation sont influencés par la complexation du ligand. (iii) L’étude de la thermodynamique de l’interaction par Calorimétrie à Titrage Isotherme (ITC) et spectroscopie d’absorption en fonction de l’état d’ionisation des résidus identifiés par l’étude RMN.<p>Le modèle du complexe d’interaction entre l’α-chymotrypsine et la proflavine a été réalisé sur base de la structure cristallographique du complexe entre cet inhibiteur et une protéase apparentée à la chymotrypsine, la thrombine. Il ressort de l’analyse du modèle obtenu que la proflavine est profondément enfouie dans le subsite S1 de l’enzyme et présente une très grande complémentarité de surface avec cette poche hydrophobe. Nous avons également pu constater la présence de plusieurs molécules d’eau immobilisées au sein du complexe, et d’une molécule en particulier faisant office de relais de liens-H.<p>L’étude de l’interaction entre l’α-chymotrypsine et la proflavine par RMN du 1H a été précédée par une étude de l’effet du degré de maturité de l’enzyme sur les interactions liant les différents résidus composant la triade catalytique (Asp102, His57 et Ser195). Lors de l’activation du précurseur inactif de l’enzyme, le chymotrypsinogène, vers la forme mature, l’α-chymotrypsine, il semble en effet que le lien-H entre le NH&949;2 de l’His57 et le Oγ de la Ser195 soit affaibli, contrairement à celui qui relie le NHδ1 de cette même histidine au Oδ1 de l’Asp102. Nous rapportons pour la première fois l’observation de l’influence de la protonation de l’Asp102 sur les déplacements chimiques des protons NHδ1 et NH&949;2 de l’His57. L’étude de l’interaction entre l’α-chymotrypsine et la proflavine par RMN, nous a permis de mettre en évidence l’effet de la complexation du ligand sur l’état d’ionisation des résidus His57 et Asp102 de la triade catalytique, les pKa de ces résidus dans l’enzyme libre valant respectivement 7 et approximativement 4.<p>Les paramètres thermodynamiques de l’interaction α-chymotrypsine - proflavine et des différents équilibres de protonation/déprotonation qui y sont liés ont été obtenus par spectroscopie d’absorption et ITC. Cette dernière technique constitue un outil précieux pour l’étude d’interactions moléculaires car il s’agit de la seule technique expérimentale permettant la mesure directe de l’enthalpie d’interaction. Lorsque des équilibres de protonation/déprotonation sont thermodynamiquement liés à l’interaction, il s’agit également de la seule technique permettant la quantification de ces effets. En mesurant la constante d’affinité et l’enthalpie d’interaction observées à différents pH et dans différents tampons, nous avons pu, sur base du modèle obtenu par RMN, déterminer les paramètres thermodynamiques intrinsèques des différents équilibres.<p>La corrélation entre les données thermodynamiques obtenues par ITC et spectroscopie d’absorption et les données structurales obtenues par RMN et sur base de l’analyse du modèle du complexe d’interaction, nous a permis de rationaliser les facteurs à la base de l’interaction préférentielle de l’inhibiteur avec une des formes de l’enzyme. L’interaction entre l’α-chymotrypsine et la proflavine est la plus favorable lorsqu’à la fois l’His57 et l’Asp102 sont déprotonnés. Cette interaction est caractérisée par un terme enthalpique favorable et un terme entropique légèrement défavorable. Ce dernier terme s’expliquerait en partie par l’immobilisation dans le site d’interaction de plusieurs molécules d’eau. L’affinité entre l’α-chymotrypsine et la proflavine diminue lorsque l’His57 se protonne. La répulsion électrostatique entre les charges positives de la proflavine et de l’His57 est vraisemblablement un des facteurs permettant d’expliquer cette diminution de la constante d’affinité. Nous n’avons pu mettre en évidence d’interaction entre ces deux molécules dès lors que l’Asp102 est protonné, malgré que ce résidu soit situé relativement loin de la proflavine dans le complexe. Il s’agit donc d’un effet indirect, probablement relayé par l’His57. Tant que l’Asp102 est déprotonné, sa charge négative compenserait la charge positive de l’His57 et réduirait la répulsion électrostatique avec la proflavine, ce qui n’est plus le cas lorsque l’aspartate se protonne. / Doctorat en sciences appliquées / info:eu-repo/semantics/nonPublished

Abstract The subject of this thesis is the characterization of chemical reactions and adsorption by means of remote detection (RD) method of nuclear magnetic resonance (NMR). The thesis consists of three related topics: In the first one, novel RD NMR based methods for characterizing chemical reactions were presented. In the second topic RD NMR methods were used to study the performance of new kind of microfluidic reactors. The third project concentrated on the development of a novel way to quantify the adsorption of flowing gas mixtures in porous materials. Even though all the topics cover quite different areas of research, they have few common nominators: remote detection NMR, microfluidics and method development. Microfluidic devices are of interest for many areas of science (such as molecular biology, disease diagnosis, chemistry) as they offer great promises for future technologies. Small dimensions enable, among many other things, the benefits of small sample volumes, large surface to volume ratio, efficient heat exchange and precise control of flow features and chemical reactions. The efficient evolution of microfluidic processes requires also the development of new innovative ways to characterize the performance of microfluidic devices. In this work, remote detection NMR is utilized for the purpose. RD is a method where the encoding and detection of information are separated physically. In many cases, the encoding and detection are performed with two separate RF coils while a fluid is passing through the studied system. In the first part of the thesis work, we introduced the concept of remote detection exchange (RD-EXSY) NMR spectroscopy. We demonstrated that the RD-EXSY method can provide unique chemical information. Furthermore, the time-of-flight (TOF) information, which is a natural side product of the experimental setup used, can be converted into indirect spatial information, showing the active reaction regions in a microfluidic device. Additionally, we demonstrated that by applying the principles of Hadamard spectroscopy in the encoding of the indirect spectral dimension we were able to produce with high efficiency RD-EXSY TOF images with direct spatial information. This allows even more accurate characterization of the active regions. The second topic concentrates on the development of microfluidc hydrogenation reactors. In the project atomic layer depositon (ALD) method was used for the first time to deposit both catalyst nanoparticles and support material on the surface of wall-coated microreactors. As a model reaction continuous flow propene hydrogenation into propane was studied by means of remote detection NMR. Reaction yield, mass transport phenomena and the activity of the catalyst surface were determined from the RD NMR data. Thirdly we presented a novel method for gas adsorption measurements in porous materials using RD TOF NMR. Traditional adsorption measurements are carried out at static conditions for a single gas component, as multi-component adsorption measurements are challenging and time-consuming. We investigated adsorption of continuously flowing propane and propene gases as well as their mixture in packed beds of mesoporous materials. The unique time-of-flight information obtained using the RD NMR method was utilized in the determination of flow velocity, which was then converted into amount of adsorbed gas.

Abstract This thesis presents computational first-principles investigations of nuclear magnetic resonance (NMR) parameters in metal-containing nanosystems. Special attention is paid to the relativistic effects observed in the vicinity of heavy elements. Small transition metal complexes are used to assess the feasibility of a quasirelativistic density functional theory (DFT) approach for calculating nuclear magnetic shielding tensors of increasingly heavy metal nuclei, followed by applications of the concept to larger systems. Nuclear magnetic shielding constants, shielding anisotropies, and chemical shifts with respect to metal ions are calculated in dimethyl and water complexes of the group-12 transition metals 67Zn, 111/113Cd, and 199/201Hg, using Hartree–Fock and DFT methods with relativistic corrections from the Breit–Pauli Perturbation Theory (BPPT). Four-component relativistic Dirac–Hartree–Fock and correlated, nonrelativistic ab initio calculations are used to benchmark the BPPT and DFT methods, respectively. The DFT/BPPT approach, combined with Monte Carlo simulations at finite temperatures, is subsequently used to calculate the chemical shift of a guest 129Xe inside a tetrahedral, iron-based cage. Complementing experiments, the encapsulation of xenon is verified, and empirically elusive details are revealed about the guest dynamics. Finally, the full shielding tensors of 31P and 195Pt and the indirect spin–spin coupling constants between the two nuclei are studied in five crystalline platinum(II) dialkyldithiophosphato complexes, concentrating on the solid-state chemical shift anisotropy and asymmetry parameters of phosphorus and platinum. The NMR parameters are calculated using DFT and the two-component zerothorder regular approximation (ZORA) for relativistic effects, combining molecular and solid-state models to incorporate indispensable contributions due to spin–orbit and crystal lattice corrections for the shielding tensors. Four-component matrix-Dirac–Kohn–Sham shielding calculations are used to benchmark the ZORA method. Qualitative, in cases nearly quantitative agreement is obtained with experiments, allowing the validation of the X-ray structures of the complexes, as well as a deeper analysis of the differences between them, including the major contributions to the NMR parameters. The results presented here demonstrate that computational NMR, a branch of relativistic quantum chemistry, is applicable and useful in studying nanoscale systems containing heavy elements, such as transition metals. Approximations are necessary to enable the treatment of large and complex targets, but sufficient accuracy is achieved for supplementing experiments with reliable and useful data that provides additional insight and analysis possibilities.

Abstract This thesis presents computational studies of both the static and dynamic parameters of the nuclear magnetic resonance (NMR) spectroscopy of gaseous xenon. First, state-of-the-art static magnetic resonance parameters are computed in small xenon clusters by using methods of quantum chemistry, and second, time-dependent relaxation phenomena are investigated via molecular dynamics simulations at different experimental conditions. Based on the underlying quantum and classical mechanics concepts, computational methods represent a procedure complementary to experiments for investigating the properties of atoms, molecules, clusters and solids. Static NMR spectral parameters, chemical shift, shielding anisotropy and asymmetry parameter, nuclear quadrupole coupling, and spin-rotation coupling, are calculated using different electronic structure methods ranging from the uncorrelated Hartree-Fock method to correlated second-order Møller-Plesset many-body perturbation, complete/restricted active space multiconfiguration self-consistent field, and to coupled-cluster approaches. The bond length dependence of these properties is investigated in the xenon dimer (Xe2). A well-characterized property in experimental NMR, the second virial coefficient of nuclear shielding, is theoretically calculated by a variety of methods and convincingly verified against experimental findings. Here, it is mandatory to include effects from special relativity as well as electron correlation. As a side result, a purely theoretical potential energy curve for Xe2, comparable to best experimental ones, is calculated. A pairwise additive scheme is established to approximate the NMR properties in differently coordinated sites of xenon clusters Xen (n = 2 - 12). Especially the pairwise additive chemical shift values are found to be in close agreement with quantum-chemical results and only a small scaling factor close to unity is needed for the correct behavior. Finally, a dynamical magnetic resonance property, the experimental nuclear spin-lattice relaxation rate R1 of monoatomic Xe gas due to the chemical shift anisotropy (CSA) mechanism is validated from first principles. This approach is based on molecular dynamics simulations over a large range of temperatures and densities, combined with the pairwise additive approximation for the shielding tensor. Therein, the shielding time correlation function is seen to reflect the characteristic time scales related to both interatomic collisions and cluster formation. For the first time, the physics of gaseous xenon is detailed in full in the context of CSA relaxation.

A series of chromone-3-carbaldehydes have been prepared using Vilsmeier-Haack methodology while a corresponding series of chromone-2-carbaldeydes have been synthesized via the Kostanecki-Robinson reaction. Baylis-Hillman reactions have been conducted on both series of chromone carbaldehydes using three different catalysts, viz., 1,4-diazabicyclo(2.2.2]octane (DABCO), 1,8-diazabicyclo[5.4.0]undec- 7-ene (DBU) and 3-hydroxyquinuclidine (3HQ), and acrylonitrile, methyl acrylate and methyl vinyl ketone as the activated alkenes. These reactions have typically (but not always!) afforded both normal Baylis-Hillman and dimeric products. Attention has also been given to the use of 1-methyl-2-pyrrolidine (1-NMP), an ionic liquid, to replace normal organic solvents, and it has been found that, in the presence of DABCO, chromone-3-carbaldehydes afford the dimeric products alone. Reactions of chromone-3-carbaldehydes with methyl vinyl ketone have yielded unexpected, novel adducts, which appear to arise from preferential attack at C(2) in the chromone nucleus. Research on chromone-2-carbaldeydes under Baylis-Hillman conditions has also resulted in the formation of some interesting products instead of the expected Baylis-Hillman adducts. The Baylis-Hillman products have been explored as substrates for aza-Michael reactions using various amino derivatives including protected amino acids in the presence of the tetrabutylammonium bromide (TBAB) and the ionic liquid, 3-butyl-1- methylimidazoleboranetetrafluoride (BmimBF₄), as catalysts. The aza-Michael products have been targeted as truncated ritonavir analogues for investigation as potential HIV -1 protease inhibitors, and representative compounds have been subjected to enzyme inhibition assays to explore the extent and type of inhibition. Lineweaver-Burk and Dixon plots have indicated competitive inhibition in one case as well as non-competitive inhibition in another, and the inhibition constants (Ki) have been compared with that of the ritonavir. Computer modelling studies have also been conducted on selected chromonecontaining derivatives, using the ACCELRYS Cerius² platform. Interactive docking of the chromone-containing ligands into the HIV -1 protease receptor site, using the Ligandfit module, has indicated the importance of hydrogen-bonding interactions mediated by bridging water molecules situated in the receptor cavity. NMR spectroscopy has been used to elucidate complex and competing mechanistic pathways involved in the Baylis-Hillman reactions of selected 2-nitrobenzaldehydes with MVK in the presence of DABCO - reactions which afford the normal BaylisHillman product, the MVK dimer and syn- and anti-Baylis-Hillman type diadducts. The kinetic data confirm the concomitant operation of two pathways and reveal that, in the initial stage of the reaction, the product distribution is kinetically controlled, whereas in the latter stage, thermodynamic control results in the consumption of the normal Baylis-Hillman product and predominance of the anti-diadduct.

Motivées par l'observation mondiale de l'augmentation de la morbidité et de la mortalité associées à des pathologies liées à l'obésité, dont la stéatose, les études pré-cliniques et cliniques s'intéressent à la recherche de nouveaux biomarqueurs pour le diagnostic de la stéatose. Actuellement, la stéatose est diagnostiquée et gradée par des analyses histologiques à partir d'une biopsie du foie. Dans l'intérêt du patient, et afin de permettre un suivi de la stéatose lors d'un régime ou d'un traitement, il est apparu important de se tourner vers des modalités de diagnostic moins invasives. Dans ce cadre, la spectroscopie par résonance magnétique (SRM), non-invasive et non-ionisante, est une méthode de choix pour le diagnostic de la stéatose par la mesure de la fraction lipidique hépatique. De plus, à partir des informations observables sur un spectre de SRM acquis au niveau hépatique, il est possible d'envisager une quantification de la composition en acides gras (AG) des lipides hépatiques, potentiel biomarqueur pour le suivi d'une stéatose. Les travaux de cette thèse ont été réalisés à partir d'objets-tests, et dans le cadre d'études pré-cliniques (4,7 T) et cliniques (3,0 T). Une étude du protocole d'acquisition de spectres de SRM pour la quantification de la composition en AG des lipides a été réalisée, avec notamment un questionnement quant à la nécessité de l'utilisation d'un module de suppression du signal de l'eau. Un état de l'art des algorithmes de quantification de la composition en AG des lipides a été effectué, et des tests de validations de ces algorithmes ont été réalisés afin de déterminer le plus approprié à la problématique hépatique, dans nos conditions expérimentales. Enfin, toujours dans l'objectif de déterminer des nouveaux biomarqueurs de la stéatose, une méthode de mesure par SRM in vivo du T1 de l'eau et de la résonance majeure des lipides hépatiques (LOREEDE pour LOngitudinal RElaxation time Evaluation from Dynamic Equilibrium) a été développée, et validée au cours d'une étude préliminaire sur des objets-tests et in vivo sur modèles murins. / In recent years, there has been an unprecedented increase in the morbidity and mortality associated with diseases such as the steatosis, linked to obesity. In this context, pre-clinical and clinical studies are of interest in the search for new biomarkers allowing the diagnosis of steatosis. Currently, steatosis is diagnosed and graded by histological analyzes from a liver biopsy. On the other hand, it is advantageous to use non-invasive diagnostic modalities, especially in longitudinal studies. In this context, magnetic resonance spectroscopy (MRS), as a non-invasive and non-ionizing approach, is an attractive alternative method for the diagnosis of steatosis by measuring the hepatic fat fraction. Moreover, from the MRS spectrum acquired in the liver, it is possible to quantify the fatty acids (FA) composition of the hepatic lipids, which could be a potential biomarker for the follow-up of steatosis. The work of this thesis has been performed in vitro and in vivo, in the context of pre-clinical (4.7 T) and clinical (3.0 T) studies. An investigation of the optimal MRS acquisition protocol for the quantification of FA was carried out, with particular attention to the role of the water signal suppression module. Different quantification algorithms of the lipid composition were studied and validation of these algorithms was carried out in vitro and in vivo. Finally, still with the objective of determining new biomarkers of steatosis, a method (LOREEDE: LOngitudinal RElaxation time Evaluation from Dynamic Equilibrium) for the measurement in vivo of the T1 of the water resonance and the major lipid resonance, by MRS, was developped and validated in a preliminary study.

Many proteins are unstable in solution and must be formulated in the solid state. This has led to an increase in the use of lyophilized dosage forms. Lyophilization is a complicated processing method consisting of three major steps: freezing, primary drying, and secondary drying. This can lead to several formulation stability challenges including changes in ionization within the matrix, phase separation of the protein drug from added stabilizers, sufficient mobility within the system for movement of reactive species and protein side chains, and crystallization of excipients upon storage. Solid-State Nuclear Magnetic Resonance Spectroscopy (SSNMR) is used to characterize many important properties of lyophilized formulations including crystalline vs. amorphous content, polymorphic form, ionization profile, interaction between formulation components with domain sizes, and mobility within the cake matrix. In order to study ionization changes in lyophilized solids, SSNMR and UV/Vis Diffuse Reflectance spectroscopy were used. 13C-labeled fumaric, succinic, and butyric acids were added to formulations at various pH levels, and were used to quantify change in the ionization of the matrix by monitoring the ionization ratios of the carboxylic acid peaks using SSNMR. pH indicators were also added to the formulations and their ionization ratio was determined using UV/Visible Diffuse Reflectance Spectroscopy. The ionization profile in the solid state was compared with that in solution before lyophilization. A rank ordering of ionization shift was made in pharmaceutically relevant buffers. SSNMR proton relaxation times (1H T1 and 1H T1rho) for each formulation component can be compared to determine homogeneity within the lyophilized matrix. The concept of spin diffusion is used in order to determine the length scale on which the components are either homogeneous or phase separated. The domain size is typically 20-50 nm or 2-10 nm for 1H T1 and 1H T1rho, respectively. PVP and dextran polymers were phase separated on both domains for physical mixtures and lyophilized mixtures. BSA and lysozyme were both lyophilized with formulations containing sucrose, trehalose, or mannitol as the stabilizer. Mannitol crystallized, and the relaxation times showed phase separation. Sucrose and trehalose both formed homogeneous systems at both length scales when formulated in a 1:1 ratio with BSA or lysozyme. Aspartame was shown to be phase separated from trehalose. The SSNMR proton relaxation times were also used to measure the local mobility in the lyophilized matrix, as a timescale of picoseconds to nanoseconds is associated with the 1H T1 relaxation time. Mobility was monitored in formulations containing a fixed amount of sucrose and mannitol, but with a variable amount of an IgG2 protein. The 1H T1 relaxation times decreased as protein content increased. The formulations with the highest relaxation time (lowest mobility), was the most stable in accelerated temperature conditions as monitored by size exclusion chromatography and capillary isoelectric focusing. This method can be used to rank order the most stable formulations at time-zero. Anti-plasticization was also studied by formulating sorbitol in various ratios with trehalose. The 1H T1 relaxation times increased with increasing sorbitol content, while the glass transition temperature decreased. Sorbitol and trehalose glasses were also exposed to different temperature storage conditions. Sorbitol appears to promote aging, as the formulations with higher sorbitol content showed larger increases in proton relaxation time.

In vivo proton magnetic resonance spectroscopy (1H MRS) is the only method available to measure small-molecule metabolites in living human tissue, including the brain, without ionizing radiation or invasive medical procedures. Despite its attendant potential for supporting clinical diagnostics in a range of neurological and psychiatric conditions, the metabolite concentration estimates produced by 1H-MRS experiments, and therefore their sensitivity and specificity to any particular biological phenomenon under study, are readily distorted by a number of confounds. These include but are not limited to static and radiofrequency field characteristics, signal relaxation dynamics, macromolecule and lipid contributions to the spectral baseline, spectral fitting artifacts, and other uncontrolled idiosyncrasies of 1H-MRS data acquisition, processing, and quantification. Using 1H-MRS data obtained via 3-Tesla and 7-Tesla magnetic resonance (MR) scanners from healthy controls, individuals with progressive and relapsing-remitting multiple sclerosis (MS), and individuals with post-traumatic stress disorder (PTSD) and/or major depressive disorder (MDD), this work therefore aims to build and apply a framework for quantifying and thereby reducing such confounds introduced to 1H-MRS estimates of in vivo metabolite concentrations at the steps of data processing and quantification, with an ultimate aim to maximizing the potential of 1H MRS for supporting sensitive and specific clinical diagnosis of neurological or psychiatric disease. The steps examined include spectral quantification by linear combination modeling (Chapter 2), absolute quantification by internal concentration referencing (Chapter 3), and cross-sectional statistical analysis of results (Chapters 4 and 5). Chapter 2 designs and implements a graphical user interface (GUI)-supported validation pipeline for measuring how data quality, spectral baseline, and baseline model affect the precision and accuracy of 1H-MR spectral quantification by linear combination modeling. This validation pipeline is then used to show that spectral data quality indices signal to noise ratio (SNR) and full width at half maximum (FWHM) interact with spectral baseline to influence not only the precision but also the accuracy of resultant metabolite concentration estimates, with fit residuals poorly indicative of true fit error and spectral baselines modeled as regularized cubic splines not significantly outperformed by those employing simulated macromolecules. A novel method for extending the commonly used spectral quantification precision estimate Cramér-Rao Lower Bound (CRLB) to incorporate considerations of continuous and piecewise polynomial baseline shapes is therefore presented, tested, and similarly integrated into a GUI-supported toolkit to improve the correspondence between estimated CRLB and metabolite fit error variability when this now empirically justified approach to spectral baseline modeling is used. In Chapter 3, age- and disease-associated differences in transverse (T2) water signal relaxation measured at 7 Tesla in the prefrontal cortex of individuals with progressive (N=21) relative to relapsing-remitting (N=26) or no (N=25) multiple sclerosis are shown to influence absolute quantification of metabolite concentrations by internal referencing to water. In Chapter 4, these findings from Chapters 2 and 3 are used to justify an evidence-based 1H-MR spectral processing and quantification protocol that focuses optimization efforts on baseline modeling approach and references metabolite concentration estimates to internal creatine instead of water. When this protocol is applied to 7-Tesla prefrontal cortex 1H-MR spectra from the aforementioned multiple sclerosis and control cohorts, it supports metabolite concentration estimates that, in the absence of any additional supporting data, inform supervised-learning-enabled identification of progressive multiple sclerosis at nearly 80% held-out validation sensitivity and specificity. Finally, in Chapter 5, the same processing, quantification, and machine-learning pipeline employed in Aim 3 is independently applied to a new set of 7-Tesla prefrontal cortex 1H-MRS raw data from an entirely different cohort of individuals with (N=20) and without (N=18) PTSD and/or comorbid or primary MDD. Here the processing, quantification, and statistics procedures designed using lessons in Chapters 2 and 3 and optimized for classifying multiple sclerosis phenotype in Chapter 4 generalize directly to metabolite-only classification of PTSD and/or MDD with sensitivity and specificity similarly near to or greater than 80%. In both Chapters 4 and 5, supervised learning avoids dimensionally reducing metabolite feature sets in order to pinpoint the specific metabolites most informative for identifying each disease group. Taken together, these findings justify the potential and continued development of 1H MRS, at least as applied in the human brain and especially as supported by multivariate approaches including supervised learning, as an auxiliary or mainstay of clinical diagnostics for neurological or psychiatric disease.