Charakterisierung der Latenz des murinen Cytomegalovirus in vivo

Die primäre, produktive Cytomegalovirus (CMV)-Infektion wird im immunkompetenten Patienten effizient durch antivirale CD8+ T-Zellen kontrolliert. Das virale Genom besitzt jedoch die Fähigkeit, in einem nicht replikativen, Latenz genannten Zustand, in gewissen Zelltypen zu persistieren, ohne dass infektiöse Nachkommenviren produziert werden. Die molekularen Mechanismen, welche der Etablierung und Aufrechterhaltung der Latenz zugrundeliegen, sind noch weitestgehend unbekannt. Es gibt Hinweise darauf, dass zelluläre Verteidigungsmechanismen die Zirkularisierung und Chromatinisierung viraler Genome hervorrufen und dadurch die virale Genexpression größtenteils verhindert wird (Marks & Spector, 1984; Reeves et al., 2006).rnAllerdings liegen die Genome nicht in einem komplett inaktiven Zustand vor. Vielmehr konnte für das murine CMV (mCMV) bereits die sporadische Transkription der Gene ie1 und ie2 während der Latenz nachgewiesen werden (Kurz et al., 1999; Grzimek et al., 2001).rnIn der vorliegenden Arbeit wurde zum ersten Mal eine umfassende in vivo Latenz-Analyse zur Charakterisierung der viralen Transkription in einer Kinetik anhand der alle drei kinetischen Klassen repräsentierenden Transkripte IE1, IE3, E1, m164, M105 und M86 vorgenommen.rnNach Latenz-Etablierung, verifiziert durch Abwesenheit von infektiösem Virus, konnten alle getesteten Transkripte in der Lunge quantifiziert werden. Interessanterweise war die transkriptionelle Aktivität zu keinem Analyse-Zeitpunkt mit der klassischen IE-E-L-Kinetik der produktiven Infektion kompatibel. Stattdessen lag eine stochastische Transkript-Expression vor, deren Aktivität mit voranschreitender Zeit immer weiter abnahm.rnWährend der Latenz exprimierte Transkripte, die für antigene Peptide kodieren, können infizierte Zellen für das Immunsystem sichtbar machen, was zu einer fortwährenden Restimulation des memory T-Zell-pools führen würde. Durch zeitgleiche Analyse der Transkript-Expression, sowie der Frequenzen Epitop-spezifischer CD8+ T-Zellen während der Latenz (IE1, m164, M105), wurde eine möglicher Zusammenhang zwischen der transkriptionellen Aktivität und der Expansion des memory T-Zell-pools untersucht. Die weitere Charakterisierung von Subpopulationen der Epitop-spezifischen CD8+ T-Zellen identifizierte die SLECs (short-lived-effector cells; CD127low CD62Llow KLRG1high) als die dominante Population in Lunge und Milz während der mCMV-Latenz.rnIn einem weiteren Teil der Arbeit sollte untersucht werden, ob IE-Genexpression zur Etablierung von Latenz notwendig ist. Mit Hilfe der Rekombinanten mCMV-Δie2-DTR, die die Gensequenz des Diphtherietoxin-Rezeptors (DTR) anstelle des Gens ie2 trägt, konnten infizierte, DTR exprimierende Zellen durch eine DT-Applikation konditional depletiert werden.rnIm latent infizierbaren Zelltyp der Leber, den LSECs (liver sinusoidal endothelial cells) wurde die virale Load durch 90-stündige DT–Applikation nach mCMV-Δie2-DTR Infektion auf das Level latent infizierter LSECs reduziert. Diese Daten sprechen für die Hypothese eines von Beginn an inaktiven Genoms, das keine IE-Genexpression zur Latenz-Etablierung benötigt. Zusätzlich stellt dieser Ansatz ein neues Tier-Modell zur Latenz-Etablierung dar. Verringerte Wartezeiten bis zur vollständigen Latenz-Etablierung, im Vergleich zum bisherigen Knochenmarktransplantations-Modell, könnten anfallende Tierhaltungskosten erheblich reduzieren und das Voranschreiten der Forschung beschleunigen.

Einfluss subviraler Partikel des humanen Cytomegalovirus auf die Induktion der antiviralen Immunantwort

Das humane Cytomegalovirus (HCMV) ist ein opportunistischer Krankheitserreger, der insbesondere bei Patienten mit unreifem oder geschwächtem Immunsystem schwere, teilweise lebensbedrohliche Erkrankungen verursacht. Aufgrund der klinischen Relevanz wird die Entwicklung einer Impfung gegen HCMV mit großem Nachdruck verfolgt. Subvirale Partikel des HCMV, sogenannte Dense Bodies (DB), stellen eine vielversprechende Impfstoff-Grundlage dar. Die innere Struktur der Partikel besteht aus viralen Proteinen, die als dominante Antigene der zellulären Immunantwort gegen HCMV identifiziert wurden. Die äußere Hülle der Partikel entspricht der Virushülle, sie enthält die viralen Oberflächenproteine als Zielantigene der neutralisierenden Antikörper (NTAk)-Antwort in ihrer natürlichen Konformation. Die für ein Totantigen außergewöhnlich hohe Immunogenität der Partikel wurde bereits in Vorarbeiten dokumentiert. Ein Ziel dieser Arbeit war es, den molekularen Hintergrund für die herausragenden, immunogenen Eigenschaften von DB aufzuklären. Im ersten Teil der Arbeit wurde daher die Hypothese geprüft, dass DB geeignet sind, die Ausreifung und Aktivierung von dendritischen Zellen (DC) zu vermitteln und damit deren Fähigkeit zur Antigenpräsentation zu stimulieren. Derart aktivierten DC kommt eine wichtige Rolle beim Priming der T-lymphozytären Immunantwort zu. In der Tat konnte gezeigt werden, dass die Behandlung von unreifen dendritischen Zellen (iDC) mit DB zu verstärkter Expression von solchen Molekülen auf der DC-Oberfläche führt, die mit Ausreifung der Zellen verknüpft sind. Der Nachweis der verstärkten Freisetzung proinflammatorischer Zytokine belegte die Aktivierung der Zellen im Sinne einer entzündlichen Reaktion. Die erfolgreiche Stimulation von CD4 und CD8 T-Lymphozyten durch DB-behandelte DC belegte schließlich die funktionelle Relevanz der Ergebnisse. Zusammengefasst konnten in diesem Abschnitt der Arbeit die molekularen Grundlagen der adjuvanten Wirkung von DB aufgeklärt werden. rnIn einem zweiten Abschnitt wurde die NTAk-Antwort nach DB-Immunisierung näher untersucht. Der humoralen Immunantwort kommt eine entscheidende Bedeutung bei der Prävention der HCMV-Übertragung zu. Hier galt es zu prüfen, welchen Einfluss stammspezifische Unterschiede in der Expression viraler Oberflächenproteine auf die Induktion der NTAk-Antwort nach DB-Immunisierung nehmen. Im Fokus stand dabei die variable Expression des pentameren Proteinkomplexes aus den viralen Proteinen gH/gL/pUL128-UL131A. Dieser Komplex wird nur von kliniknahen HCMV-Stämmen (HCMVKlin) exprimiert und ist für deren breiten Zelltropismus verantwortlich. Der pentamere Komplex fehlte in allen bisherigen Analysen der DB-Immunogenität, die auf der Grundlage von Laborstämmen des HCMV (HCMVLab) durchgeführt worden waren. Ein erster Versuchsansatz zeigte, dass die NTAk-Antwort, die durch DB von HCMVLab (DBLab) induziert wird, auch gegen die Infektion mit HCMVKlin einen gewissen Schutz vermittelt. Dies war ein überraschender Befund, da Antikörpern gegen den pentameren Komplex eine entscheidende Rolle bei der Neutralisation von HCMVKlin zugeschrieben wurde. Die Ergebnisse zeigten jedoch, dass Antikörper gegen andere Zielstrukturen zur Neutralisation von HCMVKlin beitragen. Hieraus resultierte unmittelbar die Frage, inwieweit eine Aufnahme des pentameren Komplexes in einen DB-basierten Impfstoff überhaupt notwendig war. Um dies zu beantworten war es notwendig, DB herzustellen, die den pentameren Komplex enthielten. Hierzu wurde ein HCMVLab durch Mutagenese des 235 kpb Genoms so modifiziert, dass von dem resultierenden Stamm der pentamere Komplex exprimiert wurde. In gereinigten DB dieses Stammes konnten die Komponenten des pentameren Komplexes nachgewiesen werden. Die Seren von Tieren, die mit DB dieses neuen Stammes immunisiert wurden, zeigten in der Tat eine deutlich gesteigerte Kapazität zur Neutralisation von HCMVKlin auf verschiedenen Zielzellen. Diese Ergebnisse unterstreichen schlussendlich, dass die Expression des pentameren Komplexes einen Vorteil bei der Induktion der antiviralen NTAk-Antwort erbringt. Zusammengefasst liefern die Erkenntnisse aus dieser Arbeit einen wichtigen Beitrag zum Verständnis der immunogenen Wirkung von DB. Auf dieser Grundlage war es nunmehr möglich, ein Projekt zur Austestung der DB-Vakzine in einer ersten klinischen Studie zu initiieren.

Generation and characterisation of acute myeloid leukaemia-reactive CD4 + T cells and their potential to eliminate human leukaemia blasts in NSG mice

Acute myeloid leukaemia (AML) is a cancer of the haematopoietic system, which can in many cases only be cured by haematopoietic stem cell transplantation (HSCT) and donor lymphocyte infusion (DLI) (Burnett et al., 2011). This therapy is associated with the beneficial graft-versus-leukaemia (GvL) effect mediated by transplanted donor T and NK cells that either recognise mismatch HLA molecules or polymorphic peptides, so-called minor histocompatibility antigens, leukaemia-associated or leukaemia-specific antigens in the patient and thus eliminate remaining leukaemic blasts. Nevertheless, the mature donor-derived cells often trigger graft-versus-host disease (GvHD), leading to severe damages in patients’ epithelial tissue, mainly skin, liver and intestine (Bleakley & Riddell, 2004). Therefore, approaches for the selective mediation of strong GvL effects are needed, also in order to prevent relapse after transplantation. One promising opportunity is the in vitro generation of AML-reactive CD4+ T cells for adoptive transfer. CD4+ T cells are advantageous compared to CD8+ T cells, as HLA class II molecules are under non-inflammatory conditions only expressed on haematopoietic cells; a fact that would minimise GvHD (Klein & Sato, 2000). In this study, naive CD4+ T cells were isolated from healthy donors and were successfully stimulated against primary AML blasts in mini-mixed lymphocyte/leukaemia cell cultures (mini-MLLC) in eight patient/donor pairs. After three to seven weekly restimulations, T cells were shown to produce TH1 type cytokines and to be partially of monoclonal origin according to their TCR Vβ chain usage. Furthermore, they exhibited lytic activity towards AML blasts, which was mediated by the release of granzymes A and B and perforin. The patient/donor pairs used in this study were fully HLA-class I matched, except for one pair, and also matched for HLA-DR and -DQ, whereas -DP was mismatched in one or both alleles, reflecting the actual donor selection procedure in the clinic (Begovich et al., 1992). Antibody blocking experiments suggested that the generated CD4+ T cells were directed against the HLA-DP mismatches, which could be confirmed by the recognition of donor-derived lymphoblastoid cell lines (LCLs) electroporated with the mismatched DP alleles. Under non-inflammatory conditions primary fibroblasts did not express HLA-DP and were thus not recognised, supporting the idea of a safer application of CD4+ T cells regarding induction of GvHD. For the assessment of the biological significance of these T cells, they were adoptively transferred into NSG mice engrafted with human AML blasts, where they migrated to the bone marrow and lymphoid tissue and succeeded in eliminating the leukaemic burden after only one week. Therefore, AML-reactive CD4+ T cells expanded from the naive compartment by in vitro stimulation with primary leukaemia blasts appear to be a potent tool for DLI in HSCT patients and promise to mediate specific GvL effects without causing GvHD.

Safety and therapeutic efficacy of adoptive p53-specific T cell antigen receptor (TCR) gene transfer

Immunotherapy with T cells genetically modified by retroviral transfer of tumor-associated antigen (TAA)-specific T cell receptors (TCR) is a promising approach in targeting cancer. Therefore, using a universal TAA to target different tumor entities by only one therapeutic approach was the main criteria for our TAA-specific TCR. Here, an optimized (opt) αβ-chain p53(264-272)-specific and an opt single chain (sc) p53(264-272)-specific TCR were designed, to reduce mispairing reactions of endogenous and introduced TCR α and TCR β-chains, which might lead to off-target autoimmune reactions, similar to Graft-versus-host disease (GvHD). rnIn this study we evaluated the safety issues, which rise by the risk of p53TCR gene transfer-associated on/off-target toxicities as well as the anti-tumor response in vivo in a syngeneic HLA-A*0201 transgenic mouse model. We could successfully demonstrate that opt sc p53-specific TCR-redirected T cells prevent TCR mispairing-mediated lethal off-target autoimmunity in contrast to the parental opt αβ-chain p53-specific TCR. Since the sc p53-specific TCR proofed to be safe, all further studies were performed using sc p53-specific TCR redirected T cells only. Infusion of p53-specific TCR-redirected T cells in Human p53 knock-in (Hupki) mice after lymphodepletion-preconditioning regimen with either sublethal body irradiation (5Gy) or chemotherapy (fludarabine and cyclophosphamide) in combination with vaccination (anti-CD40, CpG1668 and p53(257-282) peptide) did not result in a depletion of hematopoietic cells. Moreover, adoptive transfer of high numbers of p53-specific TCR-redirected T cells in combination with Interleukin 2 (IL-2) also did not lead to toxic on-target reactions. The absence of host tissue damage was confirmed by histology and flow cytometry analysis. Furthermore, p53-specific TCR-redirected T cells were able to lyse p53+A2.1+ tumor cells in vitro. However, in vivo studies revealed the potent suppressive effect of the tumor microenvironment (TME) mediated by tumor-infiltrating myeloid-derived suppressor cells (MDSC). Accordingly, we could improve an insufficient anti-tumor response in vivo after injection of the sc p53-specific TCR-redirected T cells by additional depletion of immunosuppressive cells of the myeloid lineage.rnTogether, these data suggest that the optimized sc p53(264-272)-specific TCR may represent a safe and efficient approach for TCR-based gene therapy. However, combinations of immunotherapeutic strategies are needed to enhance the efficacy of adoptive cell therapy (ACT)-mediated anti-tumor responses.

Synthese von MUC1-Glycopeptid-Konjugaten mit fluorierten Analoga des Thomsen-Friedenreich-Antigens

Krebserkrankungen gehen oft mit der Überexpression von mucinartigen Glycoproteinen auf der Zelloberfläche einher. In vielen Krebserkrankungen wird aufgrund der fehlerhaften Expression verschiedener Glycosyltransferasen das transmembranständige Glycoprotein MUC1, mit verkürzten Glycanstrukturen, überexprimiert. Das Auftreten der verschiedenen tumor-assoziierten Antigene (TACA) korreliert meist mit dem Fortschreiten des Krebs und der Metastasierung. Daher stellen TACAs interessante Zielmoleküle für die Entwicklung einer aktiven Tumorimmuntherapie zur spezifischen Behandlung von Adenokarzinomen dar. In dieser Arbeit galt das Interesse dem epithelialen Mucin MUC1, auf Basis dessen ein synthetischer Zugang zu einheitlichen Antitumorvakzinen, welche aus mucinanalogen Glyco-peptid¬konjugaten des MUC1 und Carrierproteinen bestehen, hergestellt werden sollten.rnUm eine tumorspezifische Immunantwort zu erhalten, müssen die selbst schwach immunogenen MUC1-Antigene über einen nicht-immunogenen Spacer mit einem geeigneten Trägerprotein, wie Tetanus Toxoid oder Rinderserumalbumin (BSA), verbunden werden. rnDa ein Einsatz von Glycokonjugaten in Impfstoffen durch die metabolische Labilität der O-glycosidischen Bindungen eingeschränkt ist, wurden hierzu erstmals fluorierte Vetreter von MUC1-analogen Glycopeptiden verwendet, in denen das Kohlenhydrat-Epitop durch den strategischen Einbau von Fluor¬atomen gegenüber einem raschen Abbau durch Glycosidasen geschützt werden soll. Dazu wurden auf Basis des literaturbekannten Thomsen-Friedenreich-Antigens Synthesestrategien zur Herstellung eines 2’F- und eines 2’,6’-bisfluorierten-Analogons erarbeitet. rnSchlüsselschritte in der Synthese stellten neben der elektrophilen Fluorierung eines Galactalvorläufers auch die -selektive 3-Galactosylierung des TN-Antigen-Bausteins zum 2’F- und 2’,6’-bisfluorierten-Analogons des TF-Disaccharids dar. Durch entsprechende Schutzgruppentransformationen wurden die beiden Derivate in entsprechende Glycosyl¬amino-säure-Bausteine für die Festphasensynthese überführt.rnNeben den beiden Analoga des TF-Antigens wurde auch erstmals ein 2F-Analogon des 2,6-Sialyl-T-Antigens hergestellt. Dazu wurde der entsprechende 2’F-TF-Baustein mit Sialinsäure-xanthogenat nach bereits bekannten Syntheseprotokollen umgesetzt. Aufgrund von Substanzmangel konnte die Verbindung nicht zur Synthese eines MUC1-Glycopeptid-Analogons herangezogen werden.rnDer Einbau der hergestellten Glycosylaminosäure-Bausteine erfolgte in die aus 20 Amino-säuren bestehende vollständige Wiederholungseinheit aus der tandem repeat-Sequenz des MUC1, wobei die entsprechenden Glycanseitenketten stets in Position 6 eingeführt wurden. Um die erhaltenen Glycopeptide für immunologische Studien an Carrier-Proteine anbinden zu können und so ggf. zu funktionsfähigen Impfstoff-Konjugaten zu gelangen, wurden diese stets N-terminal mit einem nicht-immunogenen Triethylenglycol-Spacer verknüpft. Die anschließende Funktionalisierung mit Quadratsäurediethylester erlaubte die spätere chemoselektive Konjugation an Trägerproteine, wie Tetanus Toxoid oder BSA.rnIn ersten immunologischen Bindungsstudien wurden die synthetisierten BSA-Glycopeptid-Konjugate mit Serum-Antikörpern aus Vakzinierungsstudien von MUC1-Tetanus Toxoid-Konjugaten, die (i) eine natürliche TF-Antigenstruktur und (ii) ein entsprechendes TF-Antigenderivat mit Fluorsubstituenten an C-6 des Galactosamin-Bausteins und C-6’ des Galactoserests tragen, untersucht.rn

Functional characterization of the placenta specific protein PLAC1 and its use for cancer immunotherapy

Summary Antibody-based cancer therapies have been successfully introduced into the clinic and have emerged as the most promising therapeutics in oncology. The limiting factor regarding the development of therapeutical antibody vaccines is the identification of tumor-associated antigens. PLAC1, the placenta-specific protein 1, was categorized for the first time by the group of Prof. Sahin as such a tumor-specific antigen. Within this work PLAC1 was characterized using a variety of biochemical methods. The protein expression profile, the cellular localization, the conformational state and especially the interacting partners of PLAC1 and its functionality in cancer were analyzed. Analysis of the protein expression profile of PLAC1 in normal human tissue confirms the published RT-PCR data. Except for placenta no PLAC1 expression was detectable in any other normal human tissue. Beyond, an increased PLAC1 expression was detected in several cancer cell lines derived of trophoblastic, breast and pancreatic lineage emphasizing its properties as tumor-specific antigen. rnThe cellular localization of PLAC1 revealed that PLAC1 contains a functional signal peptide which conducts the propeptide to the endoplasmic reticulum (ER) and results in the secretion of PLAC1 by the secretory pathway. Although PLAC1 did not exhibit a distinct transmembrane domain, no unbound protein was detectable in the cell culture supernatant of overexpressing cells. But by selective isolation of different cellular compartments PLAC1 was clearly enriched within the membrane fraction. Using size exclusion chromatography PLAC1 was characterized as a highly aggregating protein that forms a network of high molecular multimers, consisting of a mixture of non-covalent as well as covalent interactions. Those interactions were formed by PLAC1 with itself and probably other cellular components and proteins. Consequently, PLAC1 localize outside the cell, where it is associated to the membrane forming a stable extracellular coat-like structure.rnThe first mechanistic hint how PLAC1 promote cancer cell proliferation was achieved identifying the fibroblast growth factor FGF7 as a specific interacting partner of PLAC1. Moreover, it was clearly shown that PLAC1 as well as FGF7 bind to heparin, a glycosaminoglycan of the ECM that is also involved in FGF-signaling. The participation of PLAC1 within this pathway was approved after co-localizing PLAC1, FGF7 and the FGF7 specific receptor (FGFR2IIIb) and identifying the formation of a trimeric complex (PLAC1, FGF7 and the specific receptor FGFR2IIIb). Especially this trimeric complex revealed the role of PLAC1. Binding of PLAC1 together with FGF7 leads to the activation of the intracellular tyrosine kinase of the FGFR2IIIb-receptor and mediate the direct phosphorylation of the AKT-kinase. In the absence of PLAC1, no FGF7 mediated phosphorylation of AKT was observed. Consequently the function of PLAC1 was clarified: PLAC1 acts as a co-factor by stimulating proliferation by of the FGF7-FGFR2 signaling pathway.rnAll together, these novel biochemical findings underline that the placenta specific protein PLAC1 could be a new target for cancer immunotherapy, especially considering its potential applicability for antibody therapy in tumor patients.

Development of a highly potent bispecific antibody format targeting the novel tumor-specific antigen CLDN18.2

Due to multiple immune evasion mechanisms of cancer cells, novel therapy approaches are required to overcome the limitations of existing immunotherapies. Bispecific antibodies are potent anti-cancer drugs, which redirect effector T cells for specific tumor cell lysis, thus enabling the patient’s immune system to fight cancer cells. The antibody format used in this proof of concept study–bispecific ideal monoclonal antibodies termed BiMAB–is a tailor-made recombinant protein, which consists of two fused scFv antibodies recognizing different antigens. Both are arranged in tandem on a single peptide chain and the individual variable binding domains are separated by special non-immunogenic linkers. The format is comprised of a scFv targeting CLDN18.2–a gastric cancer tumor associated antigen (TAA) –while the second specificity binds the CD3 epsilon (CD3ε) subunit of the T cell receptor (TCR) on T cells. For the first time, we compared in our IMAB362-based BiMAB setting, four different anti-CD3-scFvs, respectively derived from the mAbs TR66, CLB-T3, as well as the humanized and the murine variant of UCHT1. In addition, we investigated the impact of an N- versus a C-terminal location of the IMAB362-derived scFv and the anti-CD3-scFvs. Thus, nine CLDN18.2 specific BiMAB proteins were generated, of which all showed a remarkably high cytotoxicity towards CLDN18.2-positive tumor cells. Because of its promising effectiveness, 1BiMAB emerged as the BiMAB prototype. The selectivity of 1BiMAB for its TAA and CD3ε, with affinities in the nanomolar range, has been confirmed by in vitro assays. Its dual binding depends on the design of an N-terminally positioned IMAB362 scFv and the consecutive C-terminally positioned TR66 scFv. 1BiMAB provoked a concentration and target cell dependent T cell activation, proliferation, and upregulation of the cytolytic protein Granzyme B, as well as the consequent elimination of target cells. Our results demonstrate that 1BiMAB is able to activate T cells independent of elements that are usually involved in the T cell recognition program, like antigen presentation, MHC restriction, and co-stimulatory effector molecules. In the first in vivo studies using a subcutaneous xenogeneic tumor mouse model in immune incompetent NSG mice, we could prove a significant therapeutic effect of 1BiMAB with partial or complete tumor elimination. The initial in vitro RIBOMAB experiments correspondingly showed encouraging results. The electroporation of 1BiMAB IVT-RNA into target or effector cells was feasible, while the functionality of translated 1BiMAB was proven by induced T cell activation and target cell lysis. Accordingly, we could show that the in vitro RIBOMAB approach was applicable for all nine BiMABs, which proves the RIBOMAB concept. Thus, the CLDN18.2-BiMAB strategy offers great potential for the treatment of cancer. In the future, administered either as protein or as IVT-RNA, the BiMAB format will contribute towards finding solutions to raise and sustain tumor-specific cellular responses elicited by engaged and activated endogenous T cells. This will potentially enable us to overcome immune evasion mechanisms of tumor cells, consequently supporting current solid gastric cancer therapies.

Synthese und Charakterisierung von Polymer-Wirkstoff-Antikörper-Konjugaten zur Anwendung in der Krebsimmuntherapie

Polymere Wirkstoffsysteme gewinnen im Bereich der biomedizinischen Forschung immer größeres Interesse. Vielversprechende Systeme für die Entwicklung von neuartigen Krebs-immun¬therapien stellen insbesondere Polymer-Konjugate dar. Das ideale Polymer-Konjugat besitzt eine Größe zwischen 10 nm und 100 nm, ist nicht zytotoxisch und zeigt keine Aggregation in humanem Blutserum. In der vorliegenden Arbeit wurde die Synthese und Charakterisierung von Polymer-Wirkstoff-Konjugaten zur Anwendung in der Krebsimmuntherapie behandelt. Erstes Ziel der Arbeit war es, geeignete polymere Trägersysteme für die in vivo Anwendung zu finden. Hierzu wurde zunächst die Stabilität verschiedener potentieller polymerer Träger-systeme (Nanohydrogele, Succinyliertes-Poly-L-Lysin (Bürste), ELP-Bürsten und Poly(2-oxazolin)bürsten) in humanem Serum untersucht. Weiterhin wurde die unspezifische Zellaufnahme in murinen dendritischen Zellen (DCs) analysiert. Es konnte gezeigt werden, dass vor allem neutrale bzw. zwitterionische Partikel eine hohe Serumstabilität sowie keine unspezifische Zellaufnahme zeigen. Um eine gezielte Adressierung der DCs des Immunsystems zu erreichen und dadurch eine Immunantwort gegenüber einem bestimmten Krebs Zelltyp zu induzieren, wurden Biokonjugate - auf Basis der Succinylierten-Poly-L-Lysin-(Bürste) sowie der Azid-funktionalisierten Poly(2-oxazolin)bürste (POx) – entwickelt, da diese Polymerbürsten keine bzw. kaum unspezifische Aufnahme in DCs zeigen. Hierbei diente der Antikörper aDEC205 der gezielten Adressierung von CD8+ DCs. Die weiteren bioaktiven Komponenten waren das tumorassoziierte Antigen (TAA) mit der Kernsequenz SIINFEKL zur Induktion einer spezifischen Immunantwort sowie der immunaktivierende TLR9 Ligand, CpG1826. Die Komponenten wurden nacheinander an die Fluoreszenz-markierten Polymere kon¬jugiert. Die Konjugation des Antikörpers erfolgte nach vorangegangener DIBO-Modifizierung über kupferfreie Click-Chemie. Mit einer optimierten Aufarbeitungsmethodik gelang es, aggregat-freie, unimere DIBO-modifizierte aDEC205 Antikörper zu isolieren. Für die succinylierten Poly-L-Lysine konnten keine eindeutigen sowie reproduzierbaren Ergebnisse erhalten werden, sodass sich im weiteren Verlauf der Arbeit auf die POx konzentriert wurde. Die Konjugation von aDEC205 an POx wurde mittels verschiedener physiko-chemischer Methoden (UV-VIS, SDS-PAGE, FCS, GPC, CLSM und FACS) gezeigt. Mit Hilfe von „Specific-Hybridization-Internalisation-Sensor“ Experimenten konnte eine spezifische Aufnahme des Konjugats in CD8+ DCs nachgewiesen werden. rnDie Konjugation von Antigen und CpG erfolgte ebenso nach entsprechender Modifizierung über kupferfreie Click-Chemie. SDS-PAGE, UV-VIS und FCS bestätigten eine erfolgreiche Kopplung. T-Zell-Proliferationsversuche ergaben für Antigen enthaltende Polymer-Konjugate eine CD8+ T-Zell-Aktivierung. Des Weiteren zeigten die POx keine bemerkenswerte Toxizität und deren Konjugate keine Aggregation in humanem Serum. rnrnDarüber hinaus wurde der Einfluss verschiedener Polymertopologien auf ihre Biodistribution sowie Blutzirkulation untersucht. Für die nach GPC-Fraktionierung erhaltenen verschiedenen Polymerfraktionen - hochmolekulare wurmartige Polymerbürsten, ellipsoidartige Polymer-bürsten und niedermolekulare kugelförmige Moleküle - konnten vielversprechende Ergebnisse erhalten werden.

Improving adoptive T cell therapies of cancer by ectopic expression of miR181a to repress inhibitory phosphatases

Adoptive T cell therapy using antigen-specific T lymphocytes is a powerful immunotherapeutic approach against cancer. Nevertheless, many T cells against tumor-antigens exhibit only weak anti-tumoral response. To overcome this barrier it is necessary to improve the potency and anti-tumoral efficacy of these T cells. Activation and activity of T cells are tightly controlled to inhibit unwanted T cell responses and to reduce the risk of autoimmunity. Both are regulated by extrinsic signals and intrinsic mechanisms which suppress T cell activation. The intrinsic mechanisms include the expression of phosphatases that counteract the activation-inducing kinases. Modifying the expression of these phosphatases allows the targeted modulation of T cell reactivity. MicroRNAs (miRNAs) are regulatory small noncoding RNA molecules that control gene expression by targeting messenger RNAs in a sequence specific manner. Gene-specific silencing plays a key role in diverse biological processes, such as development, differentiation, and functionality. miR181a has been shown to be highly expressed in immature T cells that recognize low-affinity antigens.rnThe present study successfully shows that ectopic expression of miR181a is able to enhance the sensitivity of both murine and human T cells. In CD4+ T helper cells as well as in CD8+ cytotoxic T cells the overexpression of miR181a leads to downregulation of multiple phosphatases involved in the T cell receptor signaling pathway. Overexpression of miR181a in human T cells achieves a co-stimulatory independent activation and has an anti-apoptotic effect on CD4+ T helper cells. Additionally, increasing the amount of miR181a enhances the cytolytic activity of murine CD8+ TCRtg T cells in an antigen-specific manner.rnTo test miR181a overexpressing T cells in vivo, a mouse tumor model using a B cell lymphoma cell line (A20-HA) expressing the Influenza hemagglutinin (Infl.-HA) antigen was established. The expression of model antigens in tumor cell lines enables targeted elimination of tumors using TCRtg T cells. The transfer of miR181a overexpressing Infl.-HA TCRtg CD8+ T cells alone has no positive effect neither on tumor control nor on survival of A20-HA tumor-bearing mice. In contrast, the co-transfer of miR181a overexpressing Infl.-HA TCRtg CD8+ and CD4+ T cells leads to improved tumor control and prolongs survival of A20-HA tumor-bearing mice. This effect is characterized by higher amounts of effector T cells and the expansion of Infl.-HA TCRtg CD8+ T cells.rnAll effects were achieved by changes in expression of several genes including molecules involved in T cell differentiation, activation, and regulation, cytotoxic effector molecules, and receptors important for the homing process of T cells in miR181a overexpressing T cells. The present study demonstrates that miR181a is able to enhance the anti-tumoral response of antigen-specific T cells and is a promising candidate for improving adoptive cell therapy.

G -rich oligodeoxyribonucleotides: Triplexes and transcription

Formation of a triple helix resulting from oligonucleotide binding to the DNA double helix offers new possibilities to control gene expression at the transcriptional level. Purine-motif triplexes can be formed under physiological pH. Nevertheless, this formation was inhibited by certain monovalent cations during the association but not during dissociation. Since triplexes are very stable, it was possible to assemble them in the absence of KCl and have them survive throughout the course of an in vitro transcription reaction. As for the design of a better triplex-forming oligonucleotide, 12 nucleotides in length afforded the highest binding affinity. G/T-rich oligonucleotides can be very polymorphic in solution. The conditions for forming purine-motif triplexes, duplexes or G-quartets were determined. Understanding these parameters will be important for the practical use of G-rich oligonucleotides in the development of DNA aptamers where the structure of the oligonucleotide is paramount in dictating its function. Finally, purine-motif triplexes were demonstrated to significantly inhibit gene transcription in vitro. The optimal effect on this process was dependent on the location of triplexes within the promoter, i.e., whether upstream or proximally downstream of the transcription start site. The mechanism for the inhibition of transcription appeared to be interference with initiation through preventing engagement by RNA polymerase. This finding is revolutionary when compared to the conventional model where triplexes inhibit transcription only by occluding binding by trans-acting proteins. Our findings broaden the utility of triplexes and support a strategy for antigene therapy by triplexes. ^

Detection of competing DNA structures by thermal gradient gel electrophoresis: from self-association to triple helix formation by (G,A)-containing oligonucleotides

Sequence-specific recognition of DNA can be achieved by triple helix-forming oligonucleotides that bind to the major groove of double-helical DNA. These oligonucleotides have been used as sequence-specific DNA ligands for various purposes, including sequence-specific gene regulation in the so-called ‘antigene strategy’. In particular, (G,A)-containing oligonucleotides can form stable triple helices under physiological conditions. However, triplex formation may be in competition with self-association of these oligonucleotides. For biological applications it would be interesting to identify the conditions under which one structure is favoured as compared to the other(s). Here we have directly studied competition between formation of a parallel (G,A) homoduplex and that of a triple helix by a 13 nt (G,A)-containing oligonucleotide. Temperature gradient gel electrophoresis allows simultaneous detection of competition between the two structures, because of their different temperature dependencies and gel electrophoretic mobilities, and characterisation of this competition.

Sequence-specific arrest of primer extension on single-stranded DNA by an oligonucleotide-minor groove binder conjugate.

A minor groove binder (MGB) derivative (N-3-carbamoyl-1,2-dihydro-3H-pyrrolo[3,2-e]indole-7-carboxylate tripeptide; CDPI3) was covalently linked to the 5' or 3' end of several oligodeoxyribonucleotides (ODNs) totally complementary or possessing a single mismatch to M13mp19 single-stranded DNA. Absorption thermal denaturation and slot-blot hybridization studies showed that conjugation of CDPI3 to these ODNs increased both the specificity and the strength with which they hybridized. Primer extension of the same phage DNA by a modified form of phage T7 DNA polymerase (Sequenase) was physically blocked when a complementary 16-mer with a conjugated 5'-CDPI3 moiety was hybridized to a downstream site. Approximately 50% of the replicating complexes were arrested when the blocking ODN was equimolar to the phage DNA. Inhibition was unaffected by 3'-capping of the ODN with a hexanol group or by elimination of a preannealing step. Blockage was abolished when a single mismatch was introduced into the ODN or when the MGB was either removed or replaced by a 5'-acridine group. A 16-mer with a 3'-CDPI3 moiety failed to arrest primer extension, as did an unmodified 32-mer. We attribute the exceptional stability of hybrids formed by ODNs conjugated to a CDPI3 to the tethered tripeptide binding in the minor groove of the hybrid. When that group is linked to the 5' end of a hybridized ODN, it probably blocks DNA synthesis by inhibiting strand displacement. These ODNs conjugated to CDPI3 offer attractive features as diagnostic probes and antigene agents.

Vector-mediated delivery of a polyamide ("peptide") nucleic acid analogue through the blood-brain barrier in vivo.

Polyamide ("peptide") nucleic acids (PNAs) are molecules with antigene and antisense effects that may prove to be effective neuropharmaceuticals if these molecules are enabled to undergo transport through the brain capillary endothelial wall, which makes up the blood-brain barrier in vivo. The model PNA used in the present studies is an 18-mer that is antisense to the rev gene of human immunodeficiency virus type 1 and is biotinylated at the amino terminus and iodinated at a tyrosine residue near the carboxyl terminus. The biotinylated PNA was linked to a conjugate of streptavidin (SA) and the OX26 murine monoclonal antibody to the rat transferrin receptor. The blood-brain barrier is endowed with high transferrin receptor concentrations, enabling the OX26-SA conjugate to deliver the biotinylated PNA to the brain. Although the brain uptake of the free PNA was negligible following intravenous administration, the brain uptake of the PNA was increased at least 28-fold when the PNA was bound to the OX26-SA vector. The brain uptake of the PNA bound to the OX26-SA vector was 0.1% of the injected dose per gram of brain at 60 min after an intravenous injection, approximating the brain uptake of intravenously injected morphine. The PNA bound to the OX26-SA vector retained the ability to bind to synthetic rev mRNA as shown by RNase protection assays. In summary, the present studies show that while the transport of PNAs across the blood-brain barrier is negligible, delivery of these potential neuropharmaceutical drugs to the brain may be achieved by coupling them to vector-mediated peptide-drug delivery systems.

Reformulation of a thermostable broadly protective recombinant vaccine against human papilloma virus

The causal relationship between Human Papilloma Virus (HPV) infection and cervical cancer has motivated the development, and further improvement, of prophylactic vaccines against this virus. 70% of cervical cancers, 80% of which in low-resources countries, are associated to HPV16 and HPV18 infection, with 13 additional HPV types, classified as high-risk, responsible for the remaining 30% of tumors. Current vaccines, Cervarix® (GlaxoSmithKline) and Gardasil®(Merk), are based on virus-like particles (VLP) obtained by self-assembly of the major capsid protein L1. Despite their undisputable immunogenicity and safety, the fact that protection afforded by these vaccines is largely limited to the cognate serotypes included in the vaccine (HPV 16 and 18, plus five additional viral types incorporated into a newly licensed nonavalent vaccine) along with high production costs and reduced thermal stability, are pushing the development of 2nd generation HPV vaccines based on minor capsid protein L2. The increase in protection broadness afforded by the use of L2 cross-neutralizing epitopes, plus a marked reduction of production costs due to bacterial expression of the antigens and a considerable increase in thermal stability could strongly enhance vaccine distribution and usage in low-resource countries. Previous studies from our group identified three tandem repeats of the L2 aa. 20-38 peptide as a strongly immunogenic epitope if exposed on the scaffold protein thioredoxin (Trx). The aim of this thesis work is the improvement of the Trx-L2 vaccine formulation with regard to cross-protection and thermostability, in order to identify an antigen suitable for a phase I clinical trial. By testing Trx from different microorganisms, we selected P. furiosus thioredoxin (PfTrx) as the optimal scaffold because of its sustained peptide epitope constraining capacity and striking thermal stability (24 hours at 100°C). Alternative production systems, such as secretory Trx-L2 expression in the yeast P. pastoris, have also been set-up and evaluated as possible means to further increase production yields, with a concomitant reduction of production costs. Limitations in immune-responsiveness caused by MHC class II polymorphisms –as observed, for example, in different mouse strains- have been overcome by introducing promiscuous T-helper (Th) epitopes, e.g., PADRE (Pan DR Epitope), at both ends of PfTrx. This allowed us to obtain fairly strong immune responses even in mice (C57BL/6) normally unresponsive to the basic Trx-L2 vaccine. Cross-protection was not increased, however. I thus designed, produced and tested a novel multi-epitope formulation consisting of 8 and 11 L2(20-38) epitopes derived from different HPV types, tandemly joined into a single thioredoxin molecule (“concatemers”). To try to further increase immunogenicity, I also fused our 8X and 11X PfTrx-L2 concatemers to the N-terminus of an engineered complement-binding protein (C4bp), capable to spontaneously assemble into ordered hepatmeric structures, previously validated as a molecular adjuvant. Fusion to C4bp indeed improved antigen presentation, with a fairly significant increase in both immunogenicity and cross-protection. Another important issue I addressed, is the reduction of vaccine doses/treatment, which can be achieved by increasing immunogenicity, while also allowing for a delayed release of the antigen. I obtained preliminary, yet quite encouraging results in this direction with the use of a novel, solid-phase vaccine formulation, consisting of the basic PfTrx-L2 vaccine and its C4bp fusion derivative adsorbed to mesoporus silica-rods (MSR).

Chapter 1 : nucleobases with designed patterns of hydrogen bonding

The chemistry used in key bond-forming steps to prepare nucleobases with designed patterns of hydrogen bonding is surveyed. Incorporation of the nucleobases into DNA and RNA oligomers is achieved either chemically using building blocks such as nucleoside phosphoramidites or enzymatically using nucleotide triphosphates. By varying the hydrogen bonding pattern within nucleobases, and by incorporating additional substituents, new structures have been designed that "reach over" so that contacts with both strands in targeted duplex DNA can be made in antigene strategies to control gene expression. Various new base-pairing systems have been evaluated that expand the genetic alphabet beyond Watson-Crick base pairs A.T and G.C. For example, benzo-homologated analogs of the natural DNA bases represent a new genetic set of orthogonal, size-expanded derivatives that have been shown to encode amino acids of a protein in a living organism.