Immunotherapy is a promising tool in the fight against cancer and aims to recruit patients own immune systems to seek out and destroy malignant cells. Options such as oncolytic viruses (OVs), autologous tumour vaccines and chimeric antigen receptors have shown clinical success to date, yet there remain significant hurdles to overcome. Here, we demonstrate a novel vaccine combining irrCell priming and infected cell boosting dramatically improves the tumour-specific CTL response against CT26 tumours and can be further enhanced using additional immunogenic factors (armed OVs, adjuvants). We also developed a novel fluorescence-based high-throughput screening platform to identify compounds that sensitize resistant solid tumours to killing by CAR-T cells, which ultimately revealed cardiac glycosides as putative tumour sensitizers. Overall, this thesis identifies several novel enhancers of the anticancer immune response, including a heterologous irr:ICV vaccine regimen and the potential ability to identify molecules to overcome resistance to CAR-T therapy.

Fluorescerande protein kan användas som reportrar i genetiska kretsar för att identifiera framgångsrikt modifierade celler. Med hjälp av syntetisk biologi kan genetiska kretsar, som är en sammansättning gener som kodar för protein, skapas. Genetiska kretsar möjliggör modifiering av celler och har haft stor framgång, vid immunterapi av cancer. Chimeric antigen receptor (CAR) T-celler är en genetisk krets där T-celler modifieras till att eliminera tumörceller baserat på en utvald ytmarkör. Med hjälp av fluorescerande proteiner kan olika komponenter i genetiska kretsar märkas in och därmed tydligt följas vid modifieringen av celler, ofta används Blue fluorescent protein (BFP) eller Green fluorescent protein (GFP). För att utveckla mer komplexa genetiska kretsar med flera komponenter krävs fler fluorescerande proteiner som kan kombineras med BFP och GFP, såsom sådana i det röda spektrumet. I denna studie undersöktes rödfluorescerade proteinerna E2Crimson, TagRFP657, mNeptune2.5, mKelly2, mKate2, mCardinal och Katushka2S. Med hjälp av klonade vektorer för respektive protein kan lentivirus produceras för att transducera Jurkat celler. Flödescytometri användes för att identifiera proteinernas fluorescensintensitet i det röda spektrumet, samt deras läckage i BFP och GFP spektrat. Proteinerna med högst fluorescensintensitet i det röda spektrumet samt minst läckage i BFP och GFP spektrat var E2Crimson samt mCardinal. E2Crimson har enligt tidigare studie låg toxicitet och god ljusstyrka samt hade i denna studie högst fluorescensintensitet i det röda spektrumet. E2Crimson anses därför vara optimal att kombinera med BFP och GFP i genetiska kretsar med flera komponenter. / Fluorescent protein can be used as reporters in genetic circuits to identify successfully modified cells. Using synthetic biology, genetic circuits, which are an assembly of genes that code for protein, can be created. Genetic circuits enable the modification of cells and have had great success in immunotherapy of cancer. Chimeric antigen receptor (CAR) T cells are a genetic circuit which modifies T cells to eliminate tumor cells based on a selected surface marker. With the help of the fluorescent protein, various components of genetic circuits can be marked and thus followed during the modification of cells, Blue fluorescent protein (BFP) or Green fluorescent protein (GFP) are often used as reporters. When developing complex genetic circuits with multiple components more fluorescent proteins that can combine with BFP and GFP are required, such as those in the red spectrum. In this study, the far-red fluorescent proteins E2Crimson, TagRFP657, mNeptune2.5, mKelly2, mKate2, mCardinal and Katushka2S were included. Using cloned vectors for each protein, lentiviruses can be produced to transduce Jurkat cells. Flow cytometry was used to identify the proteins fluorescence intensity in the red spectrum, as well as their leakage in the BFP and GFP spectra. The proteins with the highest fluorescence intensity in the red spectrum and the least leakage in the BFP and GFP spectra were E2Crimson and mCardinal. E2Crimson has according to other studies low toxicity and good brightness and in this study E2Crimson showed the highest fluorescence intensity in the red spectrum. E2Crimson is therefore considered optimal to combine with BFP and GFP in multicomponent genetic circuits.

INTRODUCTION: Multiple myeloma (MM) is a disease that results in the production of ineffective immunoglobulins and monoclonal proteins in the blood and urine, leading to insufficient organ function or death. Currently, there is a 5-year survival rate of 47% for patients diagnosed with MM, with a proportion of patients ultimately succumbing to the disease. The current standard of care for MM includes toxic combinations of chemotherapy. The evolution of chimeric antigen receptor (CAR) T-cell therapy for hematologic cancers such as lymphoma, leukemia, and now myeloma has provided another effective treatment option for patients who have relapsed after standard treatments for MM. Idecabtagene Vicleucel (ide-cel), was approved in March 2021 for patients with relapsed and refractory MM. While CAR T-cell treatment appears to be far less toxic than standard chemotherapy, this therapy comes with its own associated toxicities, mainly cytokine release syndrome (CRS) and neurotoxicity (NT). In clinical trials, ide-cel demonstrated to be an effective treatment in some patients, leading to the FDA approval for patients who have exhausted multiple other lines of therapy. Currently, it is unclear why patients respond differently to CAR T-cell treatment and why some patients present with more severe toxicity than others. Therefore, this study aims to examine patient factors such as demographics, age, and treatment history to determine if such characteristics may influence the CAR T-cell response; also, we assess the efficacy of ide-cel in a real-world experience outside of a clinical trial. METHODS: In this study, 14 patients’ medical records were reviewed after receiving commercial CAR T-cell therapy between August 2021 and January 2022. Eligible patients for the therapy were determined by strict inclusion criteria, including having a confirmed diagnosis of MM and exhausting at least four prior lines of therapy, as well as exclusion criteria, such as excluding individuals who have received CAR T-cells prior in a clinical trial setting. Approximately one month before preparation lymphodepletion chemotherapy, eligible patients underwent leukapheresis and had their blood sent to a laboratory to extract T-cells and genetically modify them to express the CAR for reinfusion. On 3 and 5 days prior to CAR T-cell infusion, patients underwent lymphodepletion using fludarabine and cyclophosphamide. Patients remained in the hospital for approximately one week following infusion, pending adverse reactions. After discharge, patients returned to the hospital for routine follow-ups. Data analysis was then performed on collected clinical readouts such as: prior treatments, bone marrow biopsies, response rates, laboratory values from blood samples, and pre- and post-infusion scans of various tissues within the body. RESULTS: At a median follow-up time of 15 weeks, six patients (43%) achieved a complete response (CR), three patients demonstrated a partial response (PR, 21%), and four patients showed disease progression (PD, 28%). Post-infusion scans were not available for one subject (7%) as they were still in the hospital. These results are similar to the phase I and phase II trials in which 45% and 33% of patients demonstrated a CR post-infusion, respectively. As for associated toxicities, 10 patients (71%) experienced CRS and one patient (7%) presented with ICANS. All patients that achieved a CR experienced ide-cel related toxicities, compared with only 38% of those with less favorable or unknown outcomes, which indicates that systemic immune system activation which causes CRS may be required to achieve a CR but CRS is not always linked with a CR outcome. There were 28 different chemotherapy regimens used as the standard of care treatment prior to ide-cel therapy. We assessed the most recent chemotherapeutic regimen in each patient to assess whether there is an association with most recent treatment and response. Of the six patients that achieved a CR to ide-cel, all were previously treated with RVD or CyBorD regimens, compared to the four patients who had disease progression who were mainly treated with salvage DCEP chemotherapy. Four patients (29%) received DCEP as their final chemotherapy regimen, and 3 of these 4 (75%) demonstrated progressive disease after ide-cel. Two patients received Belantamab-Mafodotin prior to ide-cel treatment, with one patient presenting with disease progression and the other patient achieving CR. 71% of patients experienced CRS following ide-cel infusion, which is resembles the phase II trial of ide-cel in which 84% of patients demonstrated CRS. In this study, only 7% of patients experienced neurological toxicity, which is comparable to the 18% of patients that demonstrated to have ICANS in the phase II study. CONCLUSIONS: We found similar performance of the ide-cel CAR-T therapy in the real world setting as in the clinical trial. Also, the complete responses were achieved by subjects with an array of characteristics, including varying recent chemotherapeutic treatments, IgG, IgA, and light-chain only subtypes of MM, and diverse demographics and other characteristics. The characteristic that demonstrated the most predictability and somewhat unique to subjects with CR was the associated toxicities from ide-cel. Development of these associated toxicities may attest that substantial immune activation, of CAR T-cells and other immune cells, leads to the efficacy of the product in eliminating cancer cells. Further analysis will need to be completed as more individuals enroll in this study to be able to determine if there are significant associations between demographics and prior lines of treatment with response to ide-cel CAR-T therapy. Lastly, future studies should assess the immune cell effector functions that are generated in CR patients that will help to specify the association between ide-cel activation, experienced associated toxicities, and its efficacy.

CAR T-cellbehandling är en immunterapi som har visat lovande resultat vid behandling av cancer. Trots det riktade immunsvaret som kan uppnås, betonar komplexiteten i tillverkningsprocessen och behandlingsproceduren det utrymme somm finns för förbättringar. Omprogrammerade T-celler har illustrerat en hög persistens hos patienter, som utsätter dem för risken för systemisk toxicitet. In-vivo transienta CAR T-celler som använder självförstärkande mRNA leverade genom affinitetsproteinbelagda LNP, föreslås som ett standardiserat alternativ som möjligör dosering av terapin vid behov.  Med hjälp av fagdisplay utfördes ett urval av affibody molekyler mot de tre immunonkologiska målproteinerna CD5, CD8 och CD19, i fyra cykler. Monoklonal fag-ELISA och DNA-sekvensering identifierade sju förmodade kandidater mot CD5, en förmodad kandidat mot CD8 och tre mot CD19. SPR analys visade specifik binding från CD5 kandidaterna, medan binding till målprotein inte kunde påvisas för CD8- och CD19 kandidaterna. De identifierade CD5-bindarna kan konjugeras till LNP för T-cell inriktad leverans av själv-amplififerande mRNA, med genetisk kod för en valfri CAR. / Chimeric antigen receptor (CAR) T-cell therapy is an immunotherapy which has shown promising results in treating patients suffering from oncological malignancies. Despite the targeted immune response that can be achieved, elaborate manufacturing and procedure processes emphasise room for improvement. Engineered T-cells have illustrated a high persistence in patients, exposing them to the risk of systemic toxicity. In-vivo transient CAR T-cells using self-amplifying mRNA by delivery through affinity protein coated lipid nanoparticles (LNP) is proposed as a standardised and reversible alternative, allowing for dosing when needed. Using phage display technology, selection of affibody molecules toward the three immune oncology proteins CD5, CD8 and CD19 was performed in four cycles. Monoclonal phage enzyme-linked immunosorbent assay (ELISA) and DNA sequencing identified seven putative candidates toward CD5, one putative candidate was isolated toward CD8, and three toward CD19. Surface plasmon resonance analysis (SPR) showed specific target binding of the CD5 candidate binders, while target binding could not be demonstrated for the CD8 and CD19 candidates. The identified CD5 binders could be conjugated to LNP for T-cell targeted delivery of self-amplifying mRNA encoding any CAR of interest.

Rapid developments in the field of CAR T cells offer important new opportunities while at the same time increasing numbers of patients pose major challenges. This review is summarizing on the one hand the state of the art in CAR T cell trials with a unique perspective on the role that Europe is playing. On the other hand, an overview of reproducible processing techniques is presented, from manual or semi-automated up to fully automated manufacturing of clinical-grade CAR T cells. Besides regulatory requirements, an outlook is given in the direction of digitally controlled automated manufacturing in order to lower cost and complexity and to address CAR T cell products for a greater number of patients and a variety of malignant diseases.

Glioblastoma (GBM) is the most frequent and lethal malignant brain tumor in adults. It emerges with an incidence of 3.2 per 100.000 in the US and 3.91 in Europe. Today, standard treatment after diagnosis consists of surgical removal of tumor tissue, followed by radiation therapy and adjuvant chemotherapy using temozolomide. Even after this rigorous therapy, patients show a median overall survival of only 15.6 months or 20.5 months when the tumor is additionally treated with so-called tumor treating fields. GBM is characterized by molecular heterogeneity within the same patient but also between different patients, which impedes development of novel therapeutics. During the last decades various immunotherapies including (multi-epitope) peptide vaccines, oncolytic viruses or immune checkpoint inhibitors against GBM were tested in small clinical studies, but failed to show a benefit in large studies. A novel kind of immunotherapies that showed great success in hematological tumors so far, is based on chimeric antigen receptors (CAR). These synthetic receptors can be introduced into immune cells to retarget their function towards tumor cells, independently of the major histocompatibility complex (MHC) that is often down regulated by tumors for immune evasion. A large hurdle for treatment of GBM using immunotherapies such as CAR-T cells, is antigen heterogeneity that limits the effect of therapies against single targets and renders the need for discovery of novel targets to enable treatment of a wide variety of patients with high success. Analyzing publicly available data and performing RT-qPCR experiments with RNA isolated from GBM tissue of a local cohort of patients, overexpression of two candidate GBM antigens, namely TMEM158 and PTPRZ1 were observed. Overexpression of both antigens in GBM in comparison to normal brain tissue and low-grade gliomas (only TMEM158) was revealed. In addition, a negative correlation between expression and patient survival was detected, as well as a correlation between TMEM158 and CD44 expression, the latter being a marker for GBM stem cells and the mesenchymal GBM subtype. Induction of chemoresistance by TMEM158 seems likely for GBM, since this was already discovered for several other tumor entities. Protein expression of TMEM158 was confirmed by Western blot analysis of different GBM cell lines. Since cell surface expression of a target protein is a prerequisite for targeting by a CAR-therapy, the expression of TMEM158 on cells from GBM cell lines was analyzed by flow cytometry. For this analysis a fluorescence-labeled peptide, based on sequence information of a known naturally occurring TMEM158 ligand (BINP) was designed. Binding to T98G and U-87 MG was observed, while only very low binding to the neuroblastoma cell line SH-SY5Y was seen in flow cytometry. Partial knockdown of TMEM158 was achieved using DsiRNAs, followed by Western blot (antibody staining) and flow cytometry (peptide staining), confirming the specificity of binding detectable by both methods. A recombinant fusion protein, consisting of the extracellular part of TMEM158 and a human Fc-antibody fragment was produced in 293T cells by transient transfection of an expression vector. The expected size of the protein produced was confirmed by Western blot. Furthermore, binding of the BINP-peptide to the recombinant protein was analyzed and compared to a scrambled BINP-peptide. In these experiments specific binding of the BINP-peptide was observed, also indicating the functionality of the recombinant protein. Next, CAR-constructs were designed using the original sequence information from BINP as binding domain and additional variants with amino acid exchanges at different positions. Significant cytotoxicity of all BINP-CAR-T cells was observed against T98G, which showed highest binding of BINP when analyzed by flow cytometry. A BINP-CAR version in which phenylalanine 11 was exchanged with alanine (BINP-F11A-CAR) showed significantly higher cytotoxicity against T98G than the BINP-CAR containing the original BINP sequence (BINP-WT-CAR). Against the U-87 MG cell line, only a version of the BINP-CAR containing an RGD- (arginine-glycine-aspartic acid) motif showed significant cytotoxicity. RGD-motifs are known to bind integrins like αVβ3, which was abundantly present on this cell line, as it was confirmed by flow cytometry within this work. Using this BINP-RGD-CAR version, targeting of both antigens at the same time seems possible. No significant cytotoxicity of the different CAR versions was observed against the TMEM158- and αVβ3-low cell line SH-SY5Y. In conclusion, overexpression of TMEM158 and PTPRZ1 and their negative influence on survival of patients, as found in recent literature, was confirmed for glioblastoma. Significantly higher expression of TMEM158 in GBM in comparison to low-grade gliomas as well as the correlation with CD44 hint at an association of TMEM158 with the aggressive phenotype of GBM. For all of these reasons, targeting of TMEM158 appears to be very feasible. Cytotoxicity of the produced BINP-CAR-T cells, which are the first CAR-T cells targeting TMEM158 so far, was demonstrated against GBM cells. Additional to cytotoxicity of the CAR-T cells, other in vitro assays and in vivo models should be utilized to determine more aspects of CAR-T cell function, in the future. For example, proliferation, cytokine release, invasion of tumor tissue, and inactivation of CAR-T cells by the tumor milieu should be quantified. To estimate how many patients could benefit from a therapy against it, percentage of patients and distribution within the tumors should be determined.

Recombinant oncolytic virus (OV) vaccines that encode tumour-associated antigens are potent boosting agents for adoptive transfer of tumor-specific T cells (adoptive T cell therapy or ACT). Current strategies to exploit boosting vaccines for ACT rely on a priori knowledge of targetable tumour epitopes and isolation of matched epitope-specific T cells. Therefore, booster vaccines must be developed on a patient-by-patient basis, which severely limits clinical feasibility. To overcome the requirement for individualized pairing of vaccines and T cells, we propose a “universal” strategy for boosting tumor-specific T cells where the boost is provided through a synthetic receptor that can be engineered into any T cell and a matched vaccine. To this end, we are employing chimeric antigen receptors (CARs), which confer MHC-independent antigen specificity to engineered T cells, and a paired OV vaccine that encodes the CAR target. As proof-of-concept, we have developed and evaluated a model where murine TCR transgenic T cells are engineered with boosting CARs against a surrogate antigen for studies in immunocompetent hosts. In chapter 3, I optimized a murine CAR-T cell manufacturing protocol that allows for generation of highly-transduced T cells that maintain a predominantly central memory (Tcm) phenotype. This protocol leads to generation of highly functional CAR-T cell products that can be cryopreserved at the end of ex vivo culture for future use in adoptive transfer and vaccination studies. In chapter 4, I evaluated the in vivo boosting potential of our dual-specific CAR-T cells with paired OV vaccines. Adoptive transfer of these CAR-engineered tumor-specific T cells followed by vaccination with paired oncolytic vesicular stomatitis virus (VSV) vaccine leads to robust, but variable and transient, CAR-mediated expansion of tumour-specific CD8+ T-cells, resulting in delayed tumour progression in aggressive syngeneic tumour models. In chapter 5, I investigated the role of OV-induced type I interferon (IFN-I) responses on CAR-T cell boosting. I found that CAR-T cell expansion and anti-tumour function following OV vaccination is limited by the IFN-I response and can be further enhanced by blocking interferon alpha and beta receptor subunit 1 (IFNAR1). This IFN-I-mediated T cell suppression was found to be T cell-extrinsic and related to premature termination of OV infection and antigen expression in vivo. In chapter 6, I investigated the role of CD4+ T cell help in vaccine-mediated T cell boosting and evaluated different genetic engineering strategies to integrate pro-survival STAT5 signaling into the CAR-T cell product in an effort to improve persistence and long-term anti-tumour efficacy. The work presented herein describes a novel and clinically feasible approach to enhancing adoptive T cell therapies and contributes to the basic understanding of T cell biology in the context of CAR-engineering and cancer vaccination. / Thesis / Doctor of Philosophy (PhD) / Despite recent advances in cancer prevention, detection, and treatment, 2 in 5 Canadians are expected to be diagnosed with cancer in their lifetime and approximately 1 in 4 will succumb to their disease. New, more specific therapies are needed to improve responses to treatment and reduce therapy-related side effects. Cell therapy is a new way to treat cancer that uses the patient’s own immune cells as a living drug. The immune cells are taken from a patient’s blood or tumour, trained to attack cancer in the laboratory, and infused back into the patient where they will find and kill cancer cells. A major challenge with this strategy is that the trained immune cells do not always survive in the patient for long enough to get rid of the tumour. To “boost” the immune cells, we are developing a new strategy where the immune cells are genetically modified and combined with a vaccine to enhance their anti-tumor activity. Just like a vaccine against a bacteria or virus, this vaccine will tell the modified immune cells to turn on, make more of themselves, and to find and kill the cancer cells. By delivering this “go” signal through a vaccine, we think that the immune cells will be better able to survive and generate a stronger, longer-lasting immune response against the cancer. This thesis tests this approach in relevant mouse models of cancer and aims to understand how we can best design the immune cells and vaccine to work together in their tumour-killing activities.

Die vergangenen Jahre haben eindrucksvoll gezeigt, dass CD19 und BCMA chimäre Antigen-Rezeptor (CAR)-T-Zell-Therapien nicht nur das Potential haben, die progres¬si-ons¬freie Überlebenszeit von Patienten mit hämatologischen Neoplasien zu verlängern, sondern auch diese erkrankten Menschen in einigen Fällen dauerhaft zu heilen. Trotz der noch immer bestehenden großen Herausforderung CAR-T-Zell-Therapien für mehr Tumorentitäten verfügbar zu machen, insbesondere was die Bekämpfung solider Tumore betrifft, haben die Entdeckung neuer Tumorantigene und Checkpoint-Inhibitoren in den letzten Jahren dazu beigetragen, dass sich immuntherapeutische Ansätze, insbesondere die der adoptiven Immunzelltherapie stetig weiterentwickeln konnten. Akute myeloische Leukämie (AML) ist eine heterogene maligne Bluterkrankung mit schlechten Überlebens¬prognosen und die am häufigsten auftretende Leukämieform bei Erwachsenen. AML ist somit eine der hämatologischen Krebserkrankungen, die noch immer dringend medizini¬sche Innovation benötigt. CAR-T-Zell-Therapien gelten als attraktive neue Strategie, um AML zu bekämpfen, da die leukämischen Blasten empfindlich für T-Zell-vermittelte Lyse sind. In dieser Studie stellen wir Sialic acid-binding immunoglobulin-like lectin (Siglec) 6 als neues Zielmolekül für CAR-T-Zell-Therapien in AML vor. Wir entwickelten einen Siglec 6-spezifischen CAR mit einer vom humanen monoklonalen Antikörper JML 1 abgeleiteten Antigen-bindenden Domäne. Unsere Untersuchungen ergaben, dass Siglec 6 auf AML-Zelllinien als auch primären AML-Blasten sowie leukämischen Stamm¬zellen (LSCs) detektierbar ist. Die Behandlung mit Siglec 6 CAR-T-Zellen vermittelte eine anti-leukämische Reaktivität, die mit der Siglec 6-Expressionsstärke in präklinischen Modellen korrelierte. So bewirkten Siglec 6 CAR-T-Zellen eine komplette Remission in einem Xenograft-AML-Modell mit immundefizienten Mäusen und Tumorzellen mit hoher Siglec 6-Expression (NSG/U937). Darüber hinaus konnten wir beobachten, dass Siglec 6 nicht auf gesunden hämatopoetischen Stamm(vor)läuferzellen (HSC/Ps) exprimiert wird. Unsere Daten legen damit nahe, dass die Siglec 6 CAR-T-Zell-Therapie zur AML-Behandlung eingesetzt werden kann, ohne dass eine nachfolgende hämatopoetische Stammzelltransplantation notwendig wird. In gesunden Blutzellen detektierten wir Siglec 6 auf einer Fraktion naiver B-Zellen und B-Gedächtniszellen sowie basophiler Gra¬nulozyten, sodass insgesamt mit einer limitierten On-Target Off-Tumor-Reaktivität ge¬rechnet werden kann. Die fehlende Siglec 6-Expression auf gesunden HSC/Ps ist ein sig¬nifikantes Unterscheidungsmerkmal zur Expression anderer zuvor untersuchter AML-CAR-Antigene wie CD33 oder CD123. Auch CD70 wird von HSC/Ps nicht exprimiert und wurde in dieser Studie als möglicher Kombinationspartner für duales Targeting in AML evaluiert. Dabei konnten wir feststellen, dass mehrheitlich mindestens eines der beiden CAR-Antigene, Siglec 6 oder CD70, von AML-Zelllinien und primären AML-Zellen expri¬miert wird. Ebenso ließ sich experimentell bestätigen, dass Siglec 6 CAR-T-Zellen mit JML 1-single chain variable fragment (scFv) und jeweils optimierten CAR-Spacersequen¬zen, keine signifikanten Funktionalitätsunter¬schiede aufweisen. Die Vorbehandlung von AML-Zellen mit Histondeacetylase (HDAC)-Inhibitoren konnte die zytolytische Aktivität der Siglec 6 CAR-T-Zellen durch Steigerung der Siglec 6-Oberflächenexpression auf den Targetzellen erhöhen. Sowohl Siglec 6 als auch CD70 CAR-T-Zellen konnten erfolgreich mit virusfreiem Gentransfer hergestellt werden, wobei die getesteten CD70 CAR-Kon¬strukte aufgrund der CD70-Expression aktivierter T-Zellen zu Fratricide führten und die CAR-Antigen-abhängige als auch unabhängige Proliferation der CD70 CAR-T-Zellen limitierte. Zusammengefasst zeigt diese Studie, dass sowohl Siglec 6 als auch CD70 hilf¬reiche Targetantigene zur CAR-T-Zellbehandlung der AML sein können und dass eine klinische Überprüfung, insbesondere von Siglec 6 CAR-T-Zellprodukten, gewährleistet ist. / In recent years, it could be impressively demonstrated that CD19 and BCMA chimeric antigen receptor (CAR) T-cell therapies not only have the potential of prolonging progression-free and overall survival of patients with hematological malignancies, but in many cases are even able to cure them. Despite major challenges still hampering the broad applicability of CAR T-cells therapies, especially in the combat against solid tumors, the discoveries of novel tumor antigens and checkpoint inhibitors in past years have contributed to the steady development of immunotherapeutic approaches particularly those of adoptive immune cell therapy. Acute myeloid leukemia (AML) is a heterogeneous malignant blood disorder with poor survival prognosis that accounts for the most common form of leukemia in adults. Hence, AML is still one of the hematologic cancers with high unmet medical need. CAR T-cell therapies are considered an attractive new strategy to tackle AML as leukemic blasts are susceptible to T-cell-mediated lysis. In this study, we introduce sialic acid-binding immunoglobulin-like lectin (Siglec)-6 as a novel target for CAR T-cell treatment in AML. We developed a Siglec-6-specific CAR with an antigen-binding domain derived from the human monoclonal antibody JML-1. Our studies revealed that Siglec-6 is detectable on AML cell lines as well as primary AML blasts and leukemic stem cells (LSCs). Siglec-6 CAR T-cell treatment conferred anti-leukemic reactivity that correlated with Siglec-6 expression levels in preclinical models. Thus, Siglec-6 CAR T-cells induced complete remission in a xenograft AML model with immunodeficient mice and engrafted tumor cells with high Siglec-6 expression (NSG/U937). Furthermore, we observed that Siglec-6 is not expressed on healthy hematopoietic stem/precursor cells (HSC/Ps). Our data suggest that Siglec-6 CAR T-cell therapy can be effectively used for AML treatment without the need for subsequent hematopoietic stem cell transplantation. In healthy blood cells, we detected Siglec-6 on a fraction of naïve and memory B cells as well as on basophilic granulocytes, therefore overall on-target off-tumor reactivity may be expected to be limited. The lack of Siglec-6 expression on healthy HSC/Ps is a significant distinctive feature from the expression of other previously studied AML CAR antigens e.g. CD33 or CD123. Also, CD70 is not expressed by HSC/Ps and was evaluated in this study as potential combination partner for dual targeting in AML. We found that the majority of AML cell lines and primary AML cells are expressing at least one of the two antigens, either Siglec-6 or CD70. We also confirmed that Siglec-6 CAR T-cells with JML-1-single chain variable fragment (scFv) and two optimized CAR spacer sequences did not show significant differences in function. Pretreatment of AML cells with histone deacetylase (HDAC) inhibitors could augment cytolytic activity of Siglec-6 CAR T-cells by increasing Siglec-6 surface expression on AML target cells. Both, Siglec-6 and CD70 CAR T-cells were successfully manufactured via non-viral gene transfer, whereby tested CD70 CAR constructs induced fratricide due to CD70 expression of activated T-cells resulting in limited CAR antigen-dependent as well as antigen-independent proliferation of CD70 CAR T-cells. In summary, this study suggests that Siglec-6 as well as CD70 may be useful target antigens for CAR T-cell treatment of AML and warrants clinical investigation of Siglec-6 CAR T-cells in particular.

Background: The CD19 chimeric antigen receptor (CAR) adoptive T-cell therapy for B-cell leukemia is a promising treatment for relapsed or refractory malignities. The overall response rate of CD19 CAR-T cells in clinical trials was greater than 80% for patients with B-cell acute lymphoblastic leukemia (B-ALL) and non-Hodgkin's lymphoma (NHL). However, CAR-T cell therapy of leukemias and solid tumors has been limited by a lot of factors such as antigen loss of tumor escape variants, reduced proliferation, persistence and tumor-infiltration of CAR-T cells in vivo, immunosuppressive tumor environment, absence of ideal antigens and on-target, off-tumor toxicities. Therefore, new strategies improving the safety and efficacy of CAR-T cells, including further T-cell modification to overcome the immune suppression, are tested. Aims: (i) Bispecific CARs designed to express two antigen-binding domains prevent of antigen escape. (ii) T-cells were genetically modified to express CAR along with an inducible IL-21 gene cassette driven by NFAT-responsive promoter. IL-21 directly enhances CAR-T cell activity and anti-tumor effects. (iii) Applying suicide epitope modification in CAR enables significantly increasing the therapeutic safety of CAR-T cells. Methods: CARs were constructed by using molecular biology...

Die Entwicklung von Autoimmunerkrankungen wird durch eine Vielzahl verschiedener Faktoren verursacht, darunter gewisse Umwelteinflüsse, genetische Veranlagungen oder Virusinfektionen. So vielfältig die Ursprünge von Autoimmunerkrankungen sind, so divers sind die daraus resultierenden Erkrankungen, wodurch die Entwicklung zuverlässiger Therapien erschwert wird. Obwohl verschiedene Behandlungsmöglichkeiten existieren, welche die Symptome bei Autoimmunerkrankungen wie neuroinflammatorischer Multipler Sklerose (MS) mildern können, z.B. mit monoklonalen Antikörpern (mAk), wirken die meisten Medikamente breit und unspezifisch. Dies beeinträchtigt die Funktionalität des Immunsystems, was wiederum zu einem höheren Risiko für bakterielle und virale Infektionen, maligne Erkrankungen oder sekundäre Autoimmunität führen kann. Andere Therapieansätze untersuchen daher Möglichkeiten einer Antigen-abhängigen Immuntoleranz-Induktion. Allerdings befinden sich diese noch in den frühen Entwicklungsphasen und deren klinische Wirksamkeit muss noch bewiesen werden. Alternativ hat es sich in der onkologischen Immuntherapie als erfolgreich erwiesen, entweder natürliche oder gentechnisch veränderte Immunzellen zu aktivieren. Bisher wurden humane T-Zellen, welche mit Hilfe chimärer Antigenrezeptoren (CAR) mit ausgewählter Spezifität ausgestattet sind, sehr erfolgreich in der Klinik gegen Tumorerkrankungen genutzt. Basierend auf diesen klinischen Erfolgen stellt sich die Frage, ob CAR T-Zellen auch zur Behandlung von Autoimmunerkrankungen eingesetzt werden können. Herkömmlichen CAR T-Zellen mangelt es jedoch sowohl an Flexibilität als auch an Kontrollierbarkeit, da sie mit einem CAR gegen ein einzelnes Antigen ausgestattet sind und in Gegenwart des Zielantigens dauerhaft aktiviert und nicht kontrollierbar sind. Um Limitationen konventioneller CARs zu überwinden, wurden universelle Adapter-CARs (UniCARs, RevCARs) in der Gruppe von Prof. Bachmann entwickelt. Die modulare UniCAR-Plattform besteht aus universell einsetzbaren UniCAR T-Zellen und anpassbaren antigenspezifischen Zielmodulen (TMs). Die Bindungseinheit des UniCAR basiert auf einem mAk mit Spezifität gegen ein Peptidepitop, das Teil des TMs ist und welches spezifisch an Zielantigene auf Tumorzellen bindet. Das TM fungiert als Adaptermolekül, das eine Vernetzung der UniCAR T-Zelle mit der Zielzelle herstellt. Nach der Vernetzung mit der Zielzelle über das TM wird die UniCAR T-Zelle aktiviert, so dass die Zielzelle eliminiert wird. Eine vor Kurzem vorgenommene Modifikation der extrazellulären UniCAR-Domäne führte zu der Reverse CAR (RevCAR)-Plattform, welche die Spezifität und Sicherheit noch weiter erhöhen sowie tonische Signale konventioneller CARs reduzieren soll. Dabei wurde die extrazelluläre mAk-basierte UniCAR-Domäne mit dem Peptidepitop des TM ausgetauscht. Eines der am besten untersuchten Autoantigene bei neuroinflammatorischen Erkrankungen ist das Myelin-Oligodendrozyten-Glykoprotein (MOG). Dieses Protein befindet sich ausschließlich im zentralen Nervensystem an der abaxonalen Membran der nervenschützenden Myelinscheide. Obwohl neuroinflammatorische Erkrankungen wie MS hauptsächlich durch T-Zellen verursacht werden, wurden bei MS-Patienten auch Autoantikörper gegen MOG nachgewiesen, was auf die Beteiligung autoreaktiver B-Zellen an der Verschlimmerung der Krankheit hinweist. Diese Ergebnisse werden durch die wirksame Behandlung von MS-Patienten mit anti-CD20-mAks belegt. In dieser Arbeit wurde MOG als erstes Modell-Zielantigen für das Retargeting von CAR-modifizierten T-Zellen gegen anti-MOG-Ak-exprimierende humane Zellen verwendet. Ziel dieser Arbeit war neue TMs basierend auf der extrazellulären Domäne des MOG Antigens zu entwickeln, die dazu dienen, UniCAR oder RevCAR T-Zellen zur Eliminierung von anti-MOG Ak-exprimierenden Zielzellen zu aktivieren. Zur Bestimmung des optimalen MOG-Antigen TM-Formats wurden für die UniCAR-Plattform ein monovalentes (25 kDa) und ein bivalentes MOG-Antigen TM (50 kDa) entwickelt. Die Wirksamkeit des UniCAR-Systems wurde in vitro demonstriert. Es konnte dabei gezeigt werden, dass beide TMs bereits nach einer kurzen Inkubationszeit von 8 Stunden eine TM-spezifische Lyse Anti-MOG scFv-exprimierender menschlicher Zelllinien durch UniCAR-T-Zellen vermitteln. Darüber hinaus war es möglich, das zytotoxische Potential der UniCAR-T-Zellen durch die Dosierung der TMs zu kontrollieren. Um zu überprüfen, ob der Effekt basierend auf der UniCAR-Platform gegen anti-MOG scFv-exprimierende Zielzellen noch gesteigert werden kann, wurde ein monovalentes MOG-Antigen RevTM für die RevCAR-Plattform entwickelt. In Kombination mit RevCAR-T-Zellen übertraf die Anwendung des RevTM beide UniCAR TMs hinsichtlich Bindungsaffinität und dosisabhängiger Zytotoxizität gegen zwei anti-MOG scFv-exprimierende Zelllinien in vitro. Außerdem war die Freisetzung proinflammatorischer und T-Zellwachstum-fördernder Zytokine im Vergleich zu UniCAR T-Zellen höher und ausgeprägter. Weiterhin wurde die Funktionalität des RevCAR-Systems gegen anti-MOG-scFv-exprimierende Zielzellen in vivo bewiesen. Zusammenfassend kann geschlussfolgert werden, dass die modularen Adapter-CAR T-Zell-Plattformen neben der Krebsimmuntherapie auch bemerkenswertes Potential für die Behandlung von MOG-assoziierten Autoimmunerkrankungen hat. Damit konnte ein erster Grundstein dafür gelegt werden, CAR T-Zellen auf die Anwendung in Autoimmunerkrankungen zu übertragen, um zukünftig gezielt fehlerhafte Immunzellen zu beseitigen, die nachweislich zur Verschlimmerung von Autoimmunerkrankungen beitragen.:Table of contents I List of abbreviations VI 1 Introduction 1 1.1 The human immune system 2 1.2 B cells and antibodies 3 1.2.1 Antibody structure 3 1.2.2 Tolerance induction 4 1.2.3 B cell activation 5 1.2.4 B cell functions in autoimmune diseases 6 1.3 Multiple sclerosis – an example for neuroinflammatory demyelination diseases 8 1.3.1.1 Disease phenotypes 9 1.3.1.2 Immunopathogenesis 9 1.3.2 Autoantigen myelin oligodendrocyte glycoprotein 11 1.4 Immunotherapy 14 1.4.1 Monoclonal antibody therapy 14 1.4.2 Chimeric antigen receptor therapy 16 1.4.2.1 UniCAR and RevCAR T cell system 18 1.4.2.2 CAR T cells in autoimmune diseases 22 1.5 Objectives 23 2 Materials and Methods 25 2.1 Materials 25 2.1.1 Consumables 25 2.1.2 Devices and software 27 2.1.3 Chemicals and reagents 32 2.1.4 Buffers and solutions 36 2.1.5 Enzymes and enzyme buffers 38 2.1.6 Kit systems 39 2.1.7 Plasmid vectors 39 2.1.8 Oligonucleotides 41 2.1.9 Antibodies 41 2.1.10 Basic media, additives, and recombinant proteins 43 2.1.11 Composition of culture media 44 2.1.12 Bacterial strain 46 2.1.13 Cell lines 46 2.1.14 Mouse strain 47 2.2 Methods 47 2.2.1 Molecular biological and microbiology methods 47 2.2.1.1 DNA digestion with restriction enzymes 47 2.2.1.2 Dephosphorylation of vectors 48 2.2.1.3 Agarose gel electrophoresis 48 2.2.1.4 Isolation and purification of DNA fragments from agarose gels 48 2.2.1.5 Ligation of DNA fragments 49 2.2.1.6 Heat-shock transformation of competent E. colis 49 2.2.1.7 Plasmid mini preparation 49 2.2.1.8 Plasmid midi preparation 50 2.2.1.9 Determination of DNA concentration 50 2.2.1.10 DNA sequencing 50 2.2.2 Cell biology methods 50 2.2.2.1 Cultivation of eukaryotic cells 50 2.2.2.2 Freezing and thawing cultured cells 51 2.2.2.3 Determination of cell number 52 2.2.2.4 Lentiviral transduction of eukaryotic cells 52 2.2.2.5 Immunofluorescence labeling 54 2.2.2.6 Flow cytometry and analysis of flow cytometry data 55 2.2.2.7 Isolation of human peripheral blood mononuclear cells 57 2.2.2.8 Isolation of T cells from PBMCs with magnetic-activated cell sorting 58 2.2.2.9 Stimulation of isolated human T cells 58 2.2.2.10 Engraftment of T cells with chimeric antigen receptors 59 2.2.3 Methods of protein biochemistry 59 2.2.3.1 Isolation of target module constructs 59 2.2.3.2 Dialysis of purified target module constructs 60 2.2.3.3 Discontinuous Sodium dodecyl sulfate polyacrylamide gel electrophoresis 60 2.2.3.4 Determination of concentration and immunochemical detection of target module constructs 62 2.2.3.5 Determination of binding affinities of target modules using enzyme-linked immunosorbent assay 63 2.2.4 In vitro functional studies 64 2.2.4.1 T cell activation and exhaustion assay 65 2.2.4.2 Luciferase assay (cytotoxicity assay) 65 2.2.4.3 Determination of cytokine concentration 65 2.2.5 In vivo functionality studies 66 2.2.5.1 Evaluation of tumor killing in vivo 66 2.2.5.2 Optical imaging of luciferase-expressing tumors in vivo 67 2.2.6 Statistical evaluation 67 3 Results 68 3.1 Design and generation of novel MOG target modules 68 3.1.1 MOG target module constructs 68 3.1.2 Expression of target modules 69 3.2 Establishment of scFv MOG-presenting cell models 72 3.3 Binding properties of MOG target modules 74 3.3.1 Determination of binding affinity between anti-MOG antibody and MOG target modules with enzyme-linked immunosorbent assay 74 3.3.2 Determination of binding affinity between anti-MOG receptor-expressing cell lines and MOG target modules with flow cytometry 75 3.4 Generation of human CAR-expressing T cells and binding of MOG target modules 80 3.4.1.1 Genetic modification of human T cells for UniCAR expression 81 3.4.1.2 Genetic modification of human T cells for RevCAR expression 83 3.5 Redirection of UniCAR T cells in vitro 85 3.5.1 Activation of redirected UniCAR T cells 85 3.5.2 Cytokine profile of redirected UniCAR T cells 87 3.5.3 Elimination of scFv MOG-positive target cells by UniCAR T cells 89 3.5.3.1 Time-dependent retargeting of scFv MOG-positive target cells by UniCAR T cells 89 3.5.3.2 Efficacy of UniCAR target modules 91 3.6 Redirection of RevCAR T cells in vitro 92 3.6.1 Activation of redirected RevCAR T cells 92 3.6.2 Cytokine profile of redirected RevCAR T cells 95 3.6.3 Elimination of scFv MOG-positive target cells by RevCAR T cells 99 3.6.3.1 Time-dependent retargeting of scFv MOG-positive target cells using RevCAR T cells 99 3.6.3.2 Efficacy of RevCAR target module 101 3.7 Investigation of RevCAR T cell-mediated cytotoxicity in vivo 103 4 Discussion 105 4.1 Structure and purification of MOG target modules 106 4.2 Expression and purification of target modules 107 4.3 Binding properties of MOG target modules 107 4.4 In vitro cytotoxic potential of UniCAR and RevCAR T cells redirected by MOG target modules 110 4.4.1 Target module-specific redirection of UniCAR T cells to scFv MOG-expressing target cells 110 4.4.2 Target module-specific redirection of RevCAR T cells to scFv MOG-expressing target cells 112 4.5 Future prospective 116 5 Summary 119 6 Zusammenfassung 121 7 References 124 List of figures 145 List of tables 147 Acknowledgement 148