京都大学 / 0048 / 新制・論文博士 / 博士(医学) / 乙第13277号 / 論医博第2185号 / 新制||医||1039(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 生田 宏一, 教授 河本 宏, 教授 杉田 昌彦 / 学位規則第4条第2項該当 / Doctor of Medical Science / Kyoto University / DFAM

Signals via the T cell receptor (TCR) are critical for the development of T cells in the thymus, maintenance of a self-tolerant peripheral T cell repertoire and the activation of T cells in secondary lymphoid organs. A dynamic balance between tyrosine phosphorylation and dephosphorylation is essential for the maintenance of homeostasis and proper regulation of the immune system. The cytoplasmic phosphatase, PTPN22 (protein tyrosine phosphatase non-receptor type 22) is involved in negative modulation of signal transduction through the TCR and plays a central role in regulating lymphocyte homeostasis. Genome wide association studies reveal that point mutations in PTPN22 confer an increased risk of developing multiple autoimmune diseases in humans. The precise function of PTPN22 and how mutations contribute to autoimmunity is controversial. Loss-of-function mutations in PTPN22 are associated with elevated T effector cell expansion and autoreactive B cells in both humans and mice. A thorough dissection of the molecular involvement of PTPN22 and its allelic variant R619W is important to delineate its role in autoimmunity, to this end we utilised the Ptpn22-/- mice generated in our laboratory. In order to address whether R619W allelic variant is a gain- or loss-of-function mutation, we expressed both PTPN22WT and PTPN22R619W constructs in primary activated Ptpn22-/- T lymphocytes using lentiviral transduction. Surprisingly expression of either wild type or variant phosphatase showed no affect on cytokine production. Preliminary results from bone marrow chimeras generated by retroviral expression of PTPN22WT and PTPN22R619W in Ptpn22 deficient mice showed reduced T cell activation compared to Ptpn22-/- T cells. PTPN22WT appeared to be more suppressive of T cell responses than variant PTPN22R619W. Consistent with studies conducted in comparable knock-in mouse models, our data point to the variant PTPN22R619W as being a partial loss of function allele. To elucidate the mechanism of PTPN22 action in context of an autoimmune disease, we investigated the effect of Ptpn22 deficiency on the phenotype of SKG mice. The SKG mouse harbours a point mutation (W163C) within the carboxyl terminal SH2- domain of ZAP-70, which results in decreased TCR signalling and impaired thymocyte development with defective positive and negative selection. These mice are prone to developing CD4+ T cell mediated autoimmune arthritis that closely resembles rheumatoid arthritis in humans. We found that thymus differentiation was partially restored in SKG Ptpn22-/- thymocytes and Ptpn22 deficiency enhanced TCR mediated signalling in SKG Ptpn22-/- thymocytes relative to SKG thymocytes. Consistent with increased signalling observed in the thymocytes, there was improved in vitro proliferation and IL-2 production of CD4+ T lymphocytes from SKG Ptpn22-/- mice compared to SKG mice. By contrast to SKG mice, SKG Ptpn22-/- mice developed less severe mannan-induced arthritis and showed decreased proportions of Th17 and higher numbers of regulatory T cells. These results show that removal of PTPN22 can compensate, at least partially, for the deficient ZAP-70 activity in the SKG mouse, thus linking PTPN22 and ZAP-70 to the same signalling pathway. This study advances our understanding of how manipulating TCR signals impacts on downstream T cell functions, suggesting PTPN22 may be a valuable target for the treatment of autoimmune diseases. Further studies to determine physiological role of the phosphatase and its disease-associated variants could provide insight into mechanism of immune activation, tolerance and autoimmunity.

Didelė dalis širdies susirgimų diagnostinių kriterijų gaunama registruojant ir analizuojant kardiosignalus, kurie atspindi tiek elektrinės širdies veiklos sutrikimus (EKG), tiek ir hemodinaminės bei mechaninės veiklos pokyčius, t.y. impedanskardiograma (IKG) ir seismokardiograma (SKG). Dar daugiau, efektyvus širdies susirgimų diagnostikos problemų sprendimas yra naujų kardiosignalų analizės technologijų kūrimas. Jau kelis dešimtmečius Furjė transformacija taikoma EKG dažnumų analizei, tuo tarpu kai IKG ir SKG dažnio charakteristikų vertinimui šis metodas nebuvo naudojamas. Darbo tikslas buvo pritaikyti Furjė analizę įvertinant bei palyginant tris sinchroniškai užregistruotus kardiosignalus, nes jie atspindi elektrinės širdies, hemodinaminės bei mechaninės širdies veiklos pokyčius geriau nei vienas EKG signalas. Kitas darbo tikslas buvo pritaikyti Furjė analizę įvertinant bei palyginant trijų sinchroniškai užregistruotų signalų - EKG, IKG ir SKG dažnio charakteristikas ir koherenciją bei klasifikuoti dvi grupes - "sveikas" ir "ligonis". Rezultatai rodo, kad koherencijos vertinimas ir spektrinė analizė gali būti naudinga gali būti naudingas širdies kraujagyslių bei plaučių sistemų ligų diagnostikai. / A big part of heart disease diagnostics criteria is collected by registration and analysis of cardio signals that reflect the disturbances of the electric heart activity – electrocardiogram (EСG), changes of hemodynamic - impedance cardiograms (IСG) and mechanic activity - seismocardiogram (SСG). ECG analysis is generally applying in clinic practice, but usually in visual way only. Due to the development of the technologies, the bigger amount of data could be stored and more exact analysis of information could be carried out. Therefore, a solution of problem of effective diagnostics of heart diseases is the creation of new technologies for analysis of cardio signals. Previously Fourier series were applied to frequency analysis of ECG, but this method was not applied for estimation of ICG and SCG frequency characteristics. In this thesis the frequency analysis method was applied to three cardio signals, because they reflect the electrical and mechanical work of the human heart better as entirely ECG signal. The main aim of this work was to adapt Fourier transformation to assessing and comparing some characteristics of hereinbefore signals, such as coherence and classify two searching groups - “healthy” and “sick”. Results showed that rating of coherence and spectral analysis could be useful for rightly analyzing and classifying the searching groups.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第18684号 / 医博第3956号 / 新制||医||1007(附属図書館) / 31617 / 京都大学大学院医学研究科医学専攻 / (主査)教授 生田 宏一, 教授 伊達 洋至, 教授 竹内 理 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

La sécurité des communications sans fil omniprésentes devient, ces dernières années, de plus en plus une exigence incontournable. Bien que la cryptographie symétrique assure largement la confidentialité des données, la difficulté concerne la génération et la distribution de clés secrètes. Récemment, des études indiquent que les caractéristiques inhérentes du canal de propagation peuvent être exploitées afin de consolider la sécurité. En particulier, le canal radio fournit en effet une source d'aléa commune à deux utilisateurs à partir de laquelle des clés secrètes peuvent être générées. Dans la présente dissertation, nous nous intéressons au processus de génération de clés secrètes (SKG), tout en reliant les propriétés du canal radio à la qualité des clés générées. D'abord nous développons un modèle du canal stochastique, traitant la sécurité du point de vue de l'espion, qui montre une mémoire de canal résiduelle bien au-delà d'une distance de quelques longueurs d'onde (scénarios spatialement non-stationnaires). Ensuite, nous exploitons les degrés de liberté (DOF) du canal et analysons leur impact sur la performance de SKG dans différentes conditions, tout en considérant des canaux plus réalistes en environnements extérieur et intérieur (respectivement grâce à des données déterministes simulées et à des mesures). Les résultats montrent que, même pour des bandes modérées (comme standardisées dans la norme IEEE 802.11), le seul DoF de fréquence ou de son association avec le DoF spatial est souvent suffisant pour générer des longues clés, à condition d'utiliser une méthode efficace de quantification des coefficients complexes du canal. / Nowadays, the security of ubiquitous wireless communications becomes more and more a crucial requirement. Even though data is widely protected via symmetric ciphering keys, a well-known difficulty is the generation and distribution of such keys. In the recent years therefore, a set of works have addressed the exploitation of inherent characteristics of the fading propagation channel for security. In particular, secret keys could be generated from the wireless channel, considered as a shared source of randomness, available merely to a pair of communicating entities. ln the present dissertation, we are interested in the approach of secret key generation (SKG) from wireless channels, especially in relating the radio channel properties to the generated keys quality. We first develop a stochastic channel model, focusing on the security with respect to the eavesdropper side, which shows a residual channel memory weil beyond a few wavelengths distance (spatially nonstationary scenarios). Then, we analyze the channel degrees of freedom (DoF) and their impact on the SKG performance in different channel conditions, especially by considering more realistic channels in both outdoor and indoor environments (respectively through simulated ray tracing data and through measurements). The results show that, even for moderately wide band (such as standardized in IEEE 802.11), the sole frequency DOF or its association with the spatial DOF is often enough for generating long keys, provided an efficient quantization method of the complex channel coefficients is used.