The aim of this project was to explore the role of the tyrosine kinases JAK1 and JAK3 in cytokine signalling, focusing on interleukin-2 signalling in CD8⁺ effector T lymphocytes. Initial experiments compared the effects of the pan JAK1/JAK3 inhibitor tofacitinib, the selective JAK1 inhibitor GSK186, and the selective JAK3 inhibitor GSK192 on IL-2 control of effector CD8+ cytotoxic T cells (CTL). On the basis of these preliminary data, a detailed analysis of the effect of tofacitinib on effector CD8⁺ T lymphocytes was performed. Phosphorylation events regulated by tofacitinib were identified using mass spectrometry analysis of SILAC (stable isotope labelling with amino acids in cell culture) labelled CTL. Tofacitinib regulated a selective number of phosphorylation sites, with less than 1.2% of the CTL phosphoproteome significantly regulated by tofacitinib treatment following 4hrs tofacitinib treatment. Proteins with downregulated phosphorylation sites were enriched in functions related to the Jak-STAT signalling, regulation of gene expression, and MAPK signalling cascades. Proteins with upregulated phosphorylations were also enriched in functions related to regulation of gene transcription. The proteome of tofacitinib treated CTL was defined by label free mass spectrometry. Approximately 4.5% of the CTL proteome was significantly regulated following 24 hours tofacitinib treatment, suggesting tofacitinib regulates the expression of a selective subset of proteins. Tofacitinib treatment resulted in the downregulation of proteins involved in ribosome biosynthesis, steroid biosynthesis, regulation of transcription and the cell cycle; and the upregulation of proteins with hydrolase activity, and with roles in the lysosome and extracellular exosomes. The phosphoproteomic and proteomic data demonstrates that JAK kinase dependent IL-2 signalling regulates essential processes in CTL by controlling a selective number of phosphorylation events and proteins. Validation of proteins identified as regulated following tofacitinib treatment identified new targets of IL-2 signalling in CTL, including the transcription factor NFIL3. NFIL3 was shown to be upregulated in CD8⁺ T lymphocytes following stimulation with IL-2 and regulated perforin and CD62L expression, suggesting a role in the regulation of CTL effector function.

Due to the growing burden of malignancy and chronic infections, manipulating CD8+ T cell responses for clinical use has become an important goal for immunologists. CD8+ T cells have the unique capacity to recognize and kill tumor cells and intracellular pathogens. Thus far, failed or only minimally effective T cell vaccines against chronic pathogen infections and tumors have highlighted basic knowledge gaps for eliciting memory CD8+ T cell protection. Defining the immunological mechanisms that determine protective capacity and longevity in T cells will be critical to both therapeutic and prohylactic vaccine efficacy. My studies focus on strategies to manipulate effector and memory CD8+ T cell responses, including their mechanisms of action. Specifically, I show that dendritic cell (DC) immunization coupled with relatively early (days 1-3) or late (days 4-6) administration of enhanced IL-2 signals drive either effector or memory programs. DC + IL-2c administered 4-6 days post-DC transfer is shown to enhance Ag-specific effector CD8+ T cell responses; this approach is further explored in the context of a cancer immunotherapy, demonstrating effective control of tumor burden in multiple murine models of cancer. Temporal alterations of IL-2 signaling from day 4-6 to day 1-3 post-DC immunization is shown to increase memory potential and memory CD8+ T cell numbers long-term. Additional studies reveal CTLA-4-mediated down-regulation of B7-ligands on DCs after IL-2c treatment, demonstrating that weaker or more transient signaling through the CD28-B7 axis may favor memory CD8+ T cell programs. My work contributes valuable concepts in memory CD8 T cell generation to develop T cell vaccines that are both safe and predictable.

Infection with Plasmodium species leads to nearly 400,000 deaths a year despite widespread use of mosquito bed nets, insecticides, and anti-malarial drugs. To date, there is not a licensed vaccine capable of providing complete protection from Plasmodium infection to vaccinees. Whole parasite vaccination of humans and rodents can achieve complete protection in vaccines, but the dose of sporozoites, number of administrations, and production concerns in generating these types of vaccines will likely prevent these approaches from achieving worldwide use. However, the protective immunological responses against Plasmodium parasites engendered by these vaccination approaches can be studied and aid in the development of advanced subunit vaccines against Plasmodium. Using rodent models of malaria to elucidate the features of protective immunity engendered by whole parasite vaccination, it has been repeatedly shown that CD8 T cell responses directed against liver-stage parasite antigens can provide complete protection with some contribution by CD4 T cells and antibody responses depending on the model system studied. However, the quantatitive and qualitative requirements for CD8 T cell immunity against Plasmodium remains largely undefined. To enhance our understanding of how to generate protective immunity against Plasmodium, I have utilized rodent models of malaria to study the superior protection afforded from single-dose vaccination with virulent sporozoites administered under prophylatic chloroquine-cover, referred to as chemoprophylaxis sporozoites (CPS) vaccination, compared to the well-studied approach of administering radiation-attenuated Plasmodium sporozoites (RAS). RAS vaccination has long been considered the “gold standard” in vaccination due the ability of RAS vaccination to engender complete protection following sporozoite challenge of vaccinated humans and rodents. However, CPS vaccination is arguably a superior vaccination approach since it can achieve protection through less vaccine administrations relative to RAS vaccination, but the immunological basis of this enhanced CPS vaccine-induced immune response was unclear. In my study, I utilized a stringent host/parasite model to find that C57Bl/6 mice administered CPS vaccination with P. yoelii sporozoites elicit substantially higher parasite-specific CD8 T cell responses than RAS vaccination, but CPS-induced CD8 T cells were not necessary for protection following liver-stage sporozoite or blood-stage parasite challenge. CPS vaccination resulted in a low grade, transient parasitemia shortly following cessation of chloroquine treatment, which lead to the generation of potent antibody responses to blood-stage parasites; this blood-stage parasite-specific antibody response correlated with sterilizing protection in sporozoite challenged CPS-vaccinated mice. Therefore, my data provide a mechanistic basis for enhanced protective immunity elicited by single-dose CPS vaccination in a rodent model that is independent of CD8 T cells. The other portion of my work examines how CD8 T cell specificity impacts protective capacity against Plasmodium. I show that robust CD8 T cell responses of similar phenotype are mounted following prime-boost immunization against three novel Plasmodium berghei protein-derived epitopes in addition to a previously described protective, immunodominant epitope. I show that only CD8 T cells specific to sporozoite surface-expressed protein-derived epitopes, but not the intracellular protein-derived epitopes, are efficiently recognized by sporozoite-infected hepatocytes in vitro. These results suggest that antigenic targets must be efficiently presented by infected hepatocytes for CD8 T cells to eliminate liver-stage Plasmodium infection and proteins expressed on the surface of sporozoites may be good target antigens for protective CD8 T cells. Collectively, my work highlights the ability to generate protective CD8 T cell independent and dependent immunity against Plasmodium infections, whether achieved through potent blood-stage-specific antibody responses, or via numerically large monospecific CD8 T cell responses that target parasite antigens that are efficiently presented during liver-stage infection. These studies are relevant in understanding how to efficiency engender protective immunity against Plasmodium, and could aid in the advancement of subunit vaccination approaches that generate immunity through the priming of responses from multiple arms of the immune response, targeting both the liver- and blood-stages of Plasmodium.

The increasing threat of epidemic and pandemic influenza underscore the need to better-understand the immune response to influenza virus infections and to better understand the factors that contribute to the clearance of virus without complications of immunopathology. A hallmark of the adaptive immune response to primary influenza virus infections is the induction of influenza-specific CD8+ T cell responses. These T cells target and kill influenza-infected epithelial cells in the airway, thereby clearing the virus and allowing recovery of the infected host. Recent reports demonstrated that CD8+ T cells express TNF-related apoptosis-inducing ligand (TRAIL) after influenza virus infection. While roles for perforin/granzyme and Fas:FasL interactions in clearing influenza virus infections had been established, little was known about the role of TRAIL in the CD8+ T cell responses to influenza virus infection. We hypothesized that influenza-specific CD8+ T cells would express TRAIL after influenza infection and could utilize TRAIL to induce the apoptosis of virally-infected cells. We discovered that CD8+ T cells do express TRAIL after influenza infection, and that this expression occurs in an influenza-specific fashion. Further, we demonstrated that these influenza-specific CD8+ T cells utilize this TRAIL to kill virally infected cells and protect the host from death, while T cells lacking TRAIL were unable to kill targets as efficiently and provided reduced protection. These data supported our hypothesis that CD8+ T cells utilize TRAIL to kill infected cells. Unexpectedly, when we increased the initial viral inoculum, the pulmonary cytotoxicity of T cells in TRAIL-/- mice was increased compared to those in TRAIL+/+ mice. Investigation of this phenomenon revealed that changes in cytotoxicity correlated not with changes in effector molecule expression on the T cells, but with increased recruitment of T cells to the lung. T cell recruitment to the lungs of TRAIL-/- mice was dependent on CCR5 and CXCR3, and likely the result of aberrant expression of MIG and MIP-1α in the lungs. Together, these data suggest that TRAIL expression contributes not only to T cell cytotoxicity, but also to the regulation of chemokine expression and associated cell recruitment after influenza virus infections. To confirm the relevance of our animal model to the study of human disease, we examined the potential role for TRAIL in the human immune response to infection. We determined that in vitro influenza infection stimulates upregulation of functional TRAIL on the surface of CD3+, CD14+, CD19+, and CD56+ PBMC populations. This expression was not caused by infection of the cells, but by interferon produced as a result of the infection. Infected (TRAIL-expressing) PBMCs killed influenza-infected lung epithelial cells, revealing that influenza infection sensitizes epithelial cells to TRAIL-induced apoptosis. Surprisingly, blocking TRAIL signaling, but not FasL signaling, was able to abrogate this killing of infected epithelial cells. Together, these data support a role for TRAIL in the human immune response to influenza virus infections. Considered as a whole, the data from these studies suggest an additional, previously-unappreciated mechanism by which CD8+ T cells can kill virally infected cells, TRAIL. They also suggest additional, previously-unappreciated roles for TRAIL in immune responses: in helping clear virally infected cells after infections and in helping control cytokine/chemokine expression, and thus the immune response, after virus infection.

We have recently demonstrated in a model of influenza A virus (IAV) infection that the absence of specific pulmonary DC subsets, including plasmacytoid DC (pDC) and CD8a+ DC, from the lungs leads to a significant decrease in the number of virus-specific CD8 T cells. Reconstitution of the lungs with physiologic numbers of pDC or CD8a+ DC is able to restore the pulmonary IAV-specific CD8 T cell response to near normal levels via a mechanism that is dependent upon direct DC:T cell interactions, DC-expressed MHC I and the presence of viral antigen. Interestingly, however, this rescue is DC subset specific, as reconstitution with purified alveolar and airway DC or alveolar macrophages was unable to rescue the virus-specific CD8 T cell response. Following IAV infection there is an abundance of IAV antigen and MHC I expressing cells present in the lungs, including infected epithelial cells. Given this fact and the inability of all DC subsets to rescue the virus-specific CD8 T cell response, it suggested that there were additional, undefined requirements for pDC- and CD8a+ DC-mediated rescue of the T cell response in the lungs. Further, although it was known that the reduction in virus-specific CD8 T cells in the lungs was a result of increased T cell apoptosis, it remained unclear what pathways of apoptosis were contributing to the increased cell death, and what mechanism pulmonary DC subsets were utilizing to rescue this defect. Here, we demonstrate that in the absence of lung-resident DC subsets, virus-specific CD8 T cells undergo significantly increased levels of apoptosis via both extrinsic activation induced cell death and intrinsic activated cell-autonomous death pathways. Reconstitution of aDC depleted lungs with pulmonary pDC and CD8a+ DC promotes increased T cell expression of the pro-survival molecule Bcl-2 and hence, increased T cell survival and accumulation in the lungs. Our studies herein demonstrate that pulmonary DC subsets utilize a variety of mechanisms to promote the rescue of virus-specific CD8 T cells in the lungs. Blockade of the costimulatory molecules CD70, and in some cases, 4-1BBL and OX40L, ablates the pulmonary DC mediated rescue of CD8 T cell numbers in the lungs, suggesting that late costimulation is one essential mechanism that pulmonary DC use to regulate CD8 T cell immunity following IAV infection. Further, we demonstrate that the absence of DC following IAV infection results in significantly reduced levels of IL-15 in the lungs and that pulmonary DC-mediated rescue of virus-specific CD8 T cell responses in the lungs requires the trans-presentation of IL-15 via DC-expressed IL-15Ra. In addition to the role of pulmonary DC mediated costimulation and IL-15 trans-presentation, we further demonstrate a previously unrecognized role for viral antigen in regulating the accumulation of both pulmonary DC and virus-specific CD8 T cells in the lungs, suggesting that viral load can dictate the nature of the inflammatory environment in the lungs and thus, regulate the character of the ensuing IAV-specific immune response. Collectively, the results detailed here demonstrate a previously unrecognized role for pulmonary DC in regulating primary IAV-specific CD8 T cell immunity, and hence, promoting enhanced viral clearance and recovery from disease.

Peripheral blood mononuclear cells (PBMCs) are the only source of human lymphoid cells routinely available for immunologic research and for immunomonitoring of T-cell responses to microbial and tumor-associated antigens. However the large majority of human T-cells resides in tissues, especially in lymphatic organs, while only 1 % of the body’s T-cells circulate in the blood stream. Previous work in mice and humans had indicated that CD4 T-cells transiently lose antigen sensitivity when cellular contacts are lost, e.g. by leaving lymphoid organs such as lymph nodes (LNs) and entering the circulation. In this study, these findings were extended to CD8 T-cells. Thus, CD8 T-cell responses of the human tonsil show a significant drop in sensitivity to viral antigens if tissue-exit was simulated by keeping cells in dispersed culture at body temperature for two hours. Conversely, tissue-like functionality in blood-derived CD8 T-cells was restored by applying the simple and robust RESTORE protocol. Indeed, application of the RESTORE protocol, i.e. pre-culturing PBMCs for two days at a high cell density before initiation of antigenic stimulation, demonstrated that CD8 T-cell responses to a broad range of viral and to tumor-associated antigens are greatly underestimated, and sometimes even remain undetected if conventional, unprocessed PBMC cultures are used. The latter finding is particularly striking with regard to the appearance of Wilms tumor 1 (WT1)-specific CD8 T-cell responses in leukemia patients after allogeneic bone marrow transplantation. My studies on the mechanism of the RESTORE protocol show that HD preculture of PBMCs does not involve antigen-or cytokine-driven clonal expansion of T-cells. Moreover, the gain in antigen sensitivity cannot be explained by a decreased activity of regulatory T-cells during the preculture step. The increased antigen sensitivity of CD8 T-cells from HD precultures of PBMCs is associated with tonic T-cell receptor signaling as indicated by enhanced tyrosine phosphorylation of the CD3 ζ chains and the tyrosine kinase Lck, thereby preparing T-cells for full responses. The upregulation of genes involved in aerobic glycolysis in “restored” CD8 memory T-cells relative to fresh cells might be an essential requirement for increased T-cell functionality including the regulation of IFN-γ production. Taken together, the RESTORE protocol, which was initially described for the CD4 T-cell response to the antibody TGN1412 permits a more meaningful monitoring of CD8 T-cell responses to viral infections and tumors. Furthermore, when generating T-cell lines for adoptive T-cell therapy, the RESTORE protocol allows the generation of CD8 T-cell lines with an improved representation of clones responding to low antigen concentrations. / Mononukleäre Zellen des peripheren Blutes (PBMCs: peripheral blood mononuclear cells) stellen die einzige routinemäßig zugängliche Quelle für humane Lymphozyten dar, welche für die immunologische Forschung und das „Immunomonitoring“ von T-Zellantworten gegen mikrobielle und Tumor-assoziierte Antigene verwendet werden. Jedoch befindet sich der Großteil der T-Zellen des Menschen in Geweben, insbesondere den lymphatischen Organen, wohingegen sich nur 1 % der T-Zellen im Blut aufhalten. Frühere Studien, die sowohl mit murinen als auch mit humanen Zellen durchgeführt wurden, zeigten, dass CD4 T-Zellen ihre Sensitivität gegenüber Antigenen zeitweise verlieren sobald zelluläre Kontakte unterbrochen werden. Dies erfolgt beispielsweise beim Verlassen der T-Zellen von Geweben und dem Eintreten in die Blutzirkulation. In dieser Arbeit wurden diese Beobachtungen auf CD8 T-Zellen ausgeweitet. So weisen humane tonsilläre CD8 T-Zellen eine signifikant niedrigere Sensitivität gegenüber viralen Antigenen auf, wenn diese in Dispersion bei Körpertemperatur für zwei Stunden gehalten werden, um das Verlassen von Geweben und somit den Verlust von zellulären Kontakten zu simulieren. Im Gegenzug konnte eine gewebeähnliche T-Zellfunktionalität bei Blutzellen durch Anwendung des RESTORE Protokolls wiederhergestellt werden. In der Tat zeigte die Anwendung des RESTORE Protokolls, welches eine Vorkultur von PBMCs für zwei Tage bei hoher Zelldichte vor antigenspezifischer T-Zellstimulation einschließt, dass CD8 T-Zellantworten gegen eine Vielzahl viraler und Tumor-assoziierter Antigene deutlich unterschätzt werden, wenn herkömmliche Stimulationsansätze verwendet werden. Teilweise können diese so gemessenen T-Zellantworten bei Verwendung herkömmliche Stimulationsansätze auch gar nicht nachgewiesen werden. Dieser Effekt war bei der Detektion von Wilms Tumor 1 (WT1)-spezifischen CD8 T-Zellantworten bei Leukämiepatienten nach allogener Stammzelltransplantation besonders deutlich zu beobachten. Meine mechanistischen Studien zeigten, dass die Vorkultur von PBMCs bei hoher Zelldichte selbst nicht zu einer Antigen- oder Zytokin-getriebenen T-Zell Expansion führt. Des Weiteren wurde gezeigt, dass der RESTORE Effekt durch den Zugewinn an CD8 T-Zellsensitivität erklärt werden kann und nicht auf eine verringerte CD8 T-Zellsuppression durch regulatorische T-Zellen während der Vorkultur zurückzuführen ist. Die erhöhte Antigensensitivität von vorkultivierten CD8 T-Zellen steht im Zusammenhang mit tonischer T-Zell Signalweiterleitung, welche anhand von erhöhter Tyrosin Phosphorylierung der CD3 ζ Ketten des T-Zell-Rezeptors und der Tyrosinkinase Lck nachgewiesen werden kann. Diese tonischen T-Zellsignale bereiten CD8 T-Zellen darauf vor, bereits auf kleine Mengen Antigen effektiv zu reagieren. Auch die Hochregulierung von Genen, welche der aeroben Glykolyse zuzuordnen sind in vorkultivierten CD8 Gedächtniszellen im Vergleich zu CD8 Gedächtniszellen, welche direkt aus dem Blut isoliert wurden, trägt zu einer erhöhten T-Zellfunktionalität bei, welche die Regulation der IFN-γ Produktion einschließt. Zusammenfassend lässt sich sagen, dass die Anwendung des RESTORE Protokolls, welches ursprünglich zum Nachweis von CD4 T-Zellantworten gegen den Antikörper TGN1412 entwickelt wurde, eine verlässliche Methode zum Nachweis von CD8 T-Zellantworten gegen virale Infektionen und Tumore darstellt. Des Weiteren kann das RESTORE Protokoll zur Generierung von T-Zelllinien in der adoptive T-Zelltherapie eingesetzt werden. Die Anwendung des Protokolls erlaubt das Generieren von Zelllinien, welche auch T-Zellklone beinhalten, die durch Immunantworten auf geringe Antigenkonzentrationen entstanden sind.

CD8 wird von thymisch gereiften a/b T Zellen als CD8ab-Heterodimer exprimiert und dient als Korezeptor bei der MHC Klasse I (MHC I) restringierten Antigenerkennung. In dieser Funktion stabilisiert CD8 die Bindung des T Zellrezeptors (TCR) an seinen Liganden (MHC I mit antigenem Peptid) und vermittelt darüber hinaus kostimulatorische Signale an die T Zelle. Die Antigene der g/d T Zellen sind bis auf einige Ausnahmen unbekannt, aber g/d T Zellen von Maus und Mensch weisen im allgemeinen höchstwahrscheinlich keine MHC-restringierte Antigenerkennung auf. Dementsprechend findet man auf den meisten g/d T Zellen dieser Spezies keinen der MHC-spezifischen Korezeptoren CD4 oder CD8. In auffälligem Gegensatz dazu exprimieren in der Milz der Ratte bis zu 80 Prozent der g/d T Zellen das CD8ab-Heterodimer. Um die mögliche Funktion von CD8ab auf g/d T Zellen besser zu verstehen, wurden zunächst die CD8-vermittelten Signale untersucht. Eine Schlüsselrolle bei der Initiierung der TCR-Signalkaskade kommt der mit CD8a assoziierten Proteintyrosinkinase p56lck zu. Diese ist, wie durch Kopräzipitationsexperimente gezeigt wurde, bei a/b und g/d T Zellen gleichermaßen mit CD8 assoziiert und zeigt dieselbe Kinaseaktivität. Weiterhin konnte in vitro ein kostimulatorisches Potential von CD8-spezifischen Antikörpern für a/b und g/d T Zellen gezeigt werden. Demnach besitzt CD8 auf a/b und g/d T Zellen prinzipiell eine vergleichbare Fähigkeit zur Übertragung kostimulatorischer Signale. Es war folglich naheliegend, einen Zusammenhang zwischen der CD8ab-Expression der g/d T Zellen mit einer möglichen MHC I estringierten Antigenerkennung zu prüfen. Da bisher keine Antigene für g/d T Zellen der Ratte bekannt sind, wurden die Antigen bindenden Regionen des g/d TCR analysiert. Bei Antigenrezeptoren von Mensch und Maus bestehen nämlich charakteristische Längenunterschiede zwischen den mit dem Antigen interagierenden CDR3-Proteinschleifen (englisch: complementarity determining regions), die in der Verknüfungsregion der V-, D- und J-Gensegmente codiert sind. Bei Maus und Mensch sind die CDR3-Längen der TCRd-Kette ähnlich variabel und etwa so lang wie bei der schweren Kette des B Zellrezeptors. Dagegen sind die CDR3-Regionen der TCRb-Kette sehr kurz, was wahrscheinlich durch die notwendige Interaktion des MHC-Peptid-Komplexes mit allen drei CDR-Schleifen bedingt wird. Daraus wurde geschlossen, daß die Bestimmung der CDR3d-Längen Aussagen über eine mögliche MHC-Restriktion von g/d T Zellen der Ratte erlauben sollte. Zur Durchführung der CDR3d-Längenanalysen wurde die Methode des Spektratyping etabliert. Hierfür wurden in der Milz häufig vorkommende TCRd V-Gensegmente (DV105 und fünf Mitglieder der ADV7-Familie) partiell kloniert und sequenziert. In CD8 positiven und negativen g/d T Zellen wurden sowohl DV105 als auch Mitglieder der ADV7-Familie jeweils etwa gleich häufig exprimiert. Spektratyp-Analysen ergaben in g/d T Zellen der Ratte für beide V-Genfamilien etwas längere CDR3d-Schleifen als bei der Maus; CD8ab positive, CD8aa positive und CD8 negative g/d T Zellen der Ratte besaßen identische Längenspektren. Demnach kann man die CD8ab-Expression von g/d T Zellen der Ratte nicht als Hinweis auf eine der klassischen MHC-Restriktion ähnliche Antigenerkennung werten. Als spezifische Eigenschaft von g/d T Zellen der Ratte wurde eine Modulation der CD8b-Genexpression nachgewiesen, die mit der Stärke des Aktivierungssignals zunahm. Nach Aktivierung in vitro exprimierte ein großer Teil zuvor CD8ab positiver g/d T Zellen nur noch das CD8aa-Homodimer, das auf MHC I restringierten a/b T Zellen deutlich schlechtere Korezeptoreigenschaften aufweist als das CD8ab-Heterodimer. Die Modulation war irreversibel, fand prätranslational statt und zeigte keine Einflüsse auf Effektorfunktionen der g/d T Zellen wie Interferon-g-Produktion oder zytotoxische Fähigkeiten. Über die physiologische Bedeutung dieser Modulation kann nur spekuliert werden, doch könnte sie ein Mechanismus sein, um den Schwellenwert für Signale einer erneuten T Zellaktivierung zu erhöhen. Darüber hinaus limitiert sie den Nutzen der CD8ab-Expression als Marker für die thymisch gereifte Linie von g/d T Zellen. / CD8 is expressed on thymus derived a/b T cells and serves as the co-receptor for MHC class I (MHC I) restricted antigen recognition. In this role CD8 stabilises the binding of the T cell receptor (TCR) to its ligand (MHC with antigenic peptide) and moreover transduces co-stimulatory signals to the T cell. Aside from few examples the antigens of g/d T cells are unknown, but mouse and human g/d T cells most probably do not possess a MHC restricted way of antigen recognition. Correspondingly most g/d T cells of these species do not express one of the MHC specific co-receptors CD4 or CD8. In striking contrast up to 80 per cent of splenic rat g/d T cells express the CD8ab heterodimer. To understand the putative function of CD8ab expressed by rat g/d T cells, CD8 signalling was investigated first. The protein tyrosine kinase p56lck plays a key role at the initiation of the TCR signalling cascade. This kinase is equally associated with CD8 on a/b and g/d T cells as found by co-precipitations, and it shows the same kinase activity there. Furthermore, in vitro a co-stimulatory potential of CD8 specific antibodies could be demonstrated for a/b and g/d T cells likewise. Therefore in principle, CD8 possesses similar co-stimulatory properties on rat a/b and g/d T cells. Consequently a correlation between CD8ab expression and a possible MHC I restriction of rat g/d T cells was tested. Because no antigens have been found for rat g/d T cells so far, the antigen binding regions of the g/d TCR were investigated. It is known that the complementarity determining region 3 (CDR3) protein loops, which are encoded by the region of V, D, and J gene segment joining, show characteristic length differences: CDR3 lengths of mouse and human TCRd chains show the same variability and are as long as CDR3 of the B cell receptor heavy chain. In contrast, the CDR3 regions of TCRb chains are very short, what probably results from the necessity of interaction between MHC and all three CDR loops. In consequence, determining the CDR3d lengths should allow predictions about a possible MHC restriction of rat g/d T cells. For CDR3d length analysis a method called spectratyping was established. First of all, common TCRd V segments from rat spleen (DV105 and five members of the ADV7 gene family) were cloned and sequenced partially. Both gene families were expressed in about the same frequency by CD8 negative and CD8 positive g/d T cells, respectively. For both V segment families the spectratyping analyses revealed somewhat longer CDR3d loops in rat than in mouse g/d T cells. CD8ab or CD8aa positive and CD8 negative g/d T cells showed exactly the same CDR3d length spectra. Therefore the CD8ab expression of rat g/d T cells is no indication for a MHC restricted antigen recognition. As a specific property of rat g/d T cells an activation dependent down modulation of the CD8b chain could be demonstrated, that increased with the strength of the activating signal. Following activation in vitro, a major percentage of g/d T cells which previously expressed CD8ab, solely expressed the CD8aa homodimer. On MHC I restricted a/b T cells CD8aa is known to show poor co-stimulatory potential compared to CD8ab. The modulation was irreversible, occurred on a pre-translational level and did not influence g/d T cell effector functions like interferon-g production or cytotoxicity. A physiological role of the modulation is speculative so but the CD8b modulation might be a mechanism to increase the signalling threshold for a following T cell activation. Moreover, the modulation limits the use of CD8ab expression as a marker for the thymically derived g/d T cell lineage.

In the present studies, the essential role of inducer cells for the induction of soluble anti-viral activity against feline immunodeficiency virus (FIV) was investigated. Induction of suppression of FIV replication was found to not strictly require autologous cells and was probably not FIV specific. Suppression was maximum when the inducer cells and the effector CD8+ T cells were in contact with each other, suggesting a potential role for membrane antigen interactions and/or cytokines in the induction process. Additionally, flow cytometry analysis demonstrated a significant increase in the percentage of CD8+ B7-1+ T cells in the peripheral blood of chronically FIV infected cats as compared with uninfected cats. Examination of the FIV V3-V4 envelope sequences from PBMC, lymph nodes and spleen from six cats chronically infected from three to six years with the molecular clone of FIV-PPR did not demonstrate viral variants specific for the tissues examined, emphasizing the critical role of the initial diversity and virulence of the infecting virus inoculum. Additionally, in vitro CD8+ T cell antiviral activity demonstrated by four of the six cats could have led to the control of virus replication in vivo, resulting in the uniform viral variants observed. Infection of specific pathogen free cats with FIV-TX53, an FIV isolate that belongs to an emerging subtype more closely related to FIV clade B, demonstrated an acute stage infection characterized by lymphoadenopathy and a viral dose dependent decline of CD4+/CD8+ T cell ratios below 1 by 11 weeks post infection. Interestingly, an expansion of CD8 low population of CD8+ T cells was observed in the infected cats. The soluble antiviral activity generated from inducer T cell stimulated CD8+ T cells from FIV-A-PPR infected cats also suppressed in vitro replication of the emerging FIV-TX53 and FIV-TX078 isolates. This is the first report demonstrating that the CD8+ T cell antiviral activity is inter-clade effective among FIV strains. As the success of a FIV vaccine could be hampered by occurrence of highly divergent viral variants in the fields, the exploitation of this innate, soluble anti-FIV activity could contribute to the design of novel, safe and complementary anti-FIV therapeutic strategies.

Bien que la plupart des études se focalisent sur les lymphocytes T CD4+ régulateurs, il a été observé que les lymphocytes T CD8+ régulateurs peuvent jouer un rôle important dans l’induction et le maintien de la tolérance immunitaire. Le transfert adoptif chez la souris et l’induction chez l’homme de lymphocytes T CD8+ régulateurs peuvent inhiber le rejet d’allogreffes et le développement de pathologies autoimmunes. Ces observations suggèrent que l’induction de ces populations peut avoir un potentiel thérapeutique. Des études supplémentaires sont encore nécessaires pour définir les conditions optimales de leur induction. Tant les nouveaux nés humains que murins ont une plus forte capacité que l’adulte à développer des lymphocytes T CD4+ régulateurs induits par une reconnaissance antigénique. La période néonatale serait donc particulièrement appropriée à l’induction de circuits régulateurs. Dans le cadre de ce travail, nous avons étudié le rôle des lymphocytes T CD8+, induits à la naissance, dans le contrôle de la réponse des lymphocytes T CD4+ de type Th2. Des souris BALB/c sont immunisées à la naissance à l’aide de cellules spléniques semi-allogéniques hybrides F1 (AJAX x BALB/c). Ces cellules persistent dans l’animal, au sein des organes lymphoïdes et stimulent ainsi de manière chronique les lymphocytes T CD4+ et T CD8+ du receveur et induisent une réponse de type Th2. Suite à l’injection des cellules spléniques semi-allogéniques au nouveau né de souris, nous avons observé l’expansion d’une population de lymphocytes T CD8+CD25+, dont le phénotype se caractérise par l’expression de Foxp3 et la production conjointe d’IFN- et l’IL-10. Nous avons pu observer que ces cellules sont capables d’inhiber la production de cytokines Th2 produites par les lymphocytes T CD4+ allospécifiques activés. Par contre, ces cellules régulatrices aggravent des réponses Th2 non apparentées. En effet, suite à une sensibilisation à l’ovalbumine, à l’âge adulte, ces souris développent de plus fortes réponses asthmatiques. D’autre part, les nouveaux nés de souris BALB/c ont été immunisés à la naissance à l’aide de cellules dendritiques semi-allogéniques hybrides F1 (AJAX x BALB/c) qui activent de manière aigüe leurs lymphocytes T. Ces souris présentent une forte réponse Th1 et Tc1/Tc2 spécifique de l’alloantigène et sont protégées contre le développement d’un asthme induit. Il a aussi été montré dans ce travail que suite à l’immunisation néonatale à l’aide de cellules dendritiques semi-allogéniques, le nombre de lymphocytes T CD8+CD44high, CD8+CD62Lhigh et CD8+CD25+ producteurs d’IFN- augmente significativement. L’IFN- produit par ces cellules contrôle l’hyperproduction des IgE spécifiques de l’allergène et la synthèse d’ARNm de l’IL-4 et de l’IL-13 induite dans les poumons des souris allergiques. Des études supplémentaires sont nécessaires pour mieux comprendre comment des lymphocytes T CD8+ régulateurs induits lors de stimulations antigéniques aigües ou chroniques en début de vie peuvent être capables de contrôler l’apparition de pathologies allergiques. De plus, le développement de vaccins capables de cibler les lymphocytes T CD8+ chez le jeune enfant pourrait contribuer à la recherche de nouveaux traitements des allergies et de l’asthme.

Cancer vaccines can fail despite the induction of large numbers of CD8+ T cells. Two categories of memory CD8+ T cells have been defined; central memory (TCM, IL-7RαhighCD44highCD62Lhigh) and effector memory (TEM, IL-7RαhighCD44highCD62Llow). It is clear that the memory phenotype of CD8+ T cells can affect vaccine potential; however methods to augment a beneficial phenotype are not clear. I have compared three vaccine delivery systems: Listeria monocytogenes, Salmonella enterica serovar Typhimurium and the particulate liposomal adjuvant, archaeosomes, for their efficacy to protect against murine melanoma. My study revealed that the anti-tumour response is strongly influenced by the kinetics, phenotype, and lymph node homing potential of CD8+ T cells. Listeria monocytogenes-ovalbumin (LM-OVA) induced TCM cells were adept at long lasting protection against B16-OVA melanoma due to their increased homeostatic and antigen-induced proliferation, interleukin-2 production, and ability to extravasate into tumour draining lymph nodes. Conversely, although Salmonella Typhimurium-ovalbumin (ST-OVA) induced TEM, produced IFN-γ, and killed target cells, this was insufficient for long-term tumour protection. Selectin-ligand engagements of TCM cells influenced their homing potential and efficacy against murine melanoma. Fucosyltransferase deficient (FtDKO) mice, lacking functional selectin ligands, were vaccinated with LM-OVA; despite the activation of cytotoxic CD8+ T cells, there was a reduced protection against murine melanoma compared to wild-type. FtDKO CD8+ T cells exhibited reduced extravasation into FtDKO lymph nodes compared to wild-type. Additionally, fewer FtDKO CD8+ T cells compared to wild-type migrated into tumour sites. Archaeosome vaccination was used to compare the influence of CD8+ T cell quantity versus phenotype. Single or multiple therapeutic vaccinations with archaeosome-OVA yielded transient melanoma tumour protection, despite an increased frequency of circulating and tumour infiltrating CD8+ T cells. This correlated with increased expression of Program death receptor-1 (PD-1) on CD8+ T cells and induction of regulatory T cells. Prophylactic archaeosome-OVA vaccination resulted in a maximal frequency of antigen-specific CD8+ T cells of ~50-60 % with just three injections, and ~50 % of the mice were of mice were afforded long-term tumour protection (> 90 days). Overall, my study shows that the choice of vaccine adjuvant and/or vector can profoundly influence CD8+ T cell quality and cancer vaccine efficacy.