i2TRON - PhD Training Program

Individual projects - Work package 2: Immuno-Oncology

Project 2.1. - The role of mitochondrial folate metabolism during macrophage polarization (Mac1L)

Introduction

Today it is well accepted that immune cells play an important role in the context of tumor development. In regard to brain tumors, immune cells of the myeloid compartment are of special importance as they can build up to 20 % of the tumor mass. Moreover, it is well known that within the tumor microenvironment, tumor associated macrophages become reprogrammed towards a cancer cell supportive rather than anti-cancer phenotype. However, the intrinsic and extrinsic factors conferring such reprogramming and especially the role of metabolism need to be specified. Especially, our understanding of the role of mitochondrial folate metabolism in myeloid cells is still rudimentary. Yet, this pathway is a topic of intense research in the field of cancer metabolism and it is targeted in the clinic by the use of antifolates. Moreover, small molecules are being developed to target specific enzymes of mitochondrial folate metabolism.

Aim and Hypothesis

As immune cells residing in the tumor microenvironment can play an important role in the clinical outcome of intervention strategies, it is important to study folate metabolism not only in cancer cells but also in immune cells of the tumor microenvironment. 

Therefore, we are currently developing different in vitro and in vivo tools to study the role of folate mediated 1C metabolism in macrophages/microglia. Using these models, we will be able to study (i) tumor development in vivo using mouse derived cancer cell lines and (ii) study macrophage metabolism ex vivo using bone-marrow derived macrophages (BMDMs).

Methods

Analysis of metabolism using GC-MS and LC-MS, BMDMs, ex vivo brain slice cultures, orthotopic brain tumor models, mouse models for melanoma, in vitro invasion and migration assays, co-culture system (macrophages and cancer cells)

Project 2.2. - Biological and clinical implications of PD-L1 protein polarization at the tumor tissue and cell levels

Introduction

Immune checkpoint blockade is currently one of the most attractive therapeutic approach to treat cancer. Inhibitors of the PD-1/PD-L1 pathway are increasingly used in the clinic and have become part of the standard therapy for multiple malignancies. The most commonly used (FDA-approved) biomarker to predict success of immune check point therapy efficacy is PD-L1 protein in the tumor microenvironment. However, there is growing evidence that patients with low/no PD-L1 expression before treatment might also benefit from immune check point therapy.

Aim and Hypothesis

In positive clinical samples, PD-L1 protein frequently shows strongest expression in tumor areas with prominent infiltration of lymphocytes, suggesting that short distance interaction with lymphocytes induces PD-L1 upregulation in tumor cells (hypothesis 1). Very interestingly, at higher magnification, IHC staining reveals asymmetrical distribution of PD-L1 in individual tumor cells with PD-L1 polarization toward the inflammatory stroma. Such an observation is consistent with our recent in vitro investigations showing that, in resistant tumor cells engaged in an immunological synapse with cytotoxic lymphocytes, PD-L1 molecules cluster in the region of the immunological synapse. We hypothesize that PD-L1 recruitment to the synapse, which is driven by actin cytoskeleton remodeling in tumor cells, provides a potent inhibitory signal to cytotoxic lymphocytes. The present project aims at investigating the significance of PD-L1 polarization at both tissue and cell levels. More specifically, the project will investigate the following questions. 1) Is PDL1 expression in tumor cells induced by lymphocyte interaction, and if so by which mechanism? 2) Is PD-L1 clustering in the immunological synapse involved in tumor immune evasion?

Methods

The project is largely based on tissue and cell imaging methods, including light microscopy (mostly for IHC analyses of tumor tissue), fluorescence (confocal) microscopy, correlative light and electron microscopy and imaging flow cytometry. In vitro models that recapitulate PD-L1 upregulation at the tumor/lymphocyte border will be established (e.g. using 3D tumor spheroids co-cultured with cytotoxic lymphocytes) and used in functional studies. The signal driving PD-L1 upregulation (soluble vs. direct mechanical trigger) and the molecular link(s) between PD-L1 and the actin cytoskeleton (ongoing BioID analysis) will be characterized. Finally, the findings will be validated in both primary and metastatic patient samples and correlated with clinico-epidemiologic data.

Project 2.3. - Immunotherapies and drug screening in improved patient-derived models: a translational approach towards personalized treatments for leukemia

Introduction

Chronic Lymphocytic leukemia (CLL) represents the most frequent leukemia in adults. Despite new targeted therapies, CLL remains an incurable disorder and a high burden for the healthcare system. Therefore alternative therapies should be developed. CLL cells survival, proliferation and resistance to therapies are highly dependent on the microenvironment in lymphoid organs. Interactions with stromal cells and the immune compartment by inducing a strong immuno-suppression and various leukemic promoting immune cells is key to leukemia progression. In recent years, immunotherapy has revolutionized the management of cancer patients. However, only a limited number of patients respond to these new therapies (30-40% response to single-agent therapy) and these are not approved yet for all cancers. It is therefore necessary to further optimize the benefits of these therapies aimed at reactivating the immune system to fight against the tumor. Based on deep immune-phenotyping results, our group recently demonstrated that dual immune checkpoint blockade-based immunotherapy was efficient to treat a murine pre-clinical model of CLL (Wierz et al, 2018). However, direct translation of these type of findings in the clinic is difficult due to the lack of human CLL-based model.  

Aim and Hypothesis

This project will aim at developing a PDX model of CLL in humanized NSG mice to test several immunotherapy combinations aiming at reactivating the exhausted immune system. Artificial lymph nodes and other 3D co-culture systems will be used to screen new therapies in complex models mimicking the tumor and its microenvironment. 

Methods

(i) establish a humanized mouse and set up the PDX of CLL cells, (ii) test immunotherapies in vivo and analyze the immune system (CyTOF, Hyperion), (iii) develop an artificial lymph node / 3D model with CLL and normal bystander cells (air-liquid interface, cell printing), (iv) perform high-throughput screening of drugs and immunotherapies on 3D organoids, and (v) validate hits in vivo in Eµ-TCL1 and CLL-PDX mice.

Project 2.4. - Impact of gut microbiome on cancer immunotherapy 

Introduction

Immune checkpoint blockades' (ICB)-based cancer immunotherapy has led to impressive and durable clinical responses in diverse cancer patients including melanoma. However, the groundbreaking success of ICB has been seriously challenged by clinical observations showing that only small fraction of patients' gains durable and sustained clinical benefit from this therapy. The gut microbiota has been recently implicated in the success of cancer immunotherapy, yet little is known about the mechanisms behind the impact of individual members of the gut microbiota on the efficacy of the treatment.

Aim and Hypothesis

This project aims to understand the mechanisms of how fine changes in the gut microbial composition and metabolism are causally connected to the success of immunotherapy. An important aspect of this project relies on investigating how changes in the microbial metabolism, brought about by the changes in the dietary patterns including fiber consumption, affect the therapy. Such understanding will provide valuable strategies and therapeutic tools to correct potential defects in the microbiome that compromise therapeutic efficacy of ICB. Overall, this project addresses an urgent and a yet unmet clinical need of how to harness the potential of the gut microbiome to enhance the ICB-based cancer immunotherapy.

Methods

This project relies on the use of gnotobiotic mouse models containing characterized synthetic human gut bacterial communities and dietary intervention. We will employ state-of- the arte Cytometry by Time-Of-Flight (CyTOF) technology will be used for comprehensive analysis of the immune landscape.

Project 2.5. - Development of chemokine receptor modulators for improved cancer immunotherapy

Introduction

ICB-based therapies have emerged as promising treatment for aggressive cancers, but their success requires efficient infiltration of cytotoxic immune cells into the tumor. Therefore, a major challenge in the field of immuno-oncology is to develop strategies, to be combined with ICB, promoting tumor infiltration by these cells. IPI and TIME jointly identified one chemokine receptor as a key receptor involved in preventing the recruitment of cytotoxic immune cells to tumors.

Aim and Hypothesis

Targeting this chemokine receptor using neutralizing monoclonal antibodies (mAbs), antibody fragments such as Nanobodies (Nbs) or small molecules represents an innovative strategy to induce infiltration of immune cells into tumors, rendering them eligible for ICB-based therapy. This project aims at (I) developing receptor modulators (mAbs, Nbs or small molecules) and (II) validating their pharmacological/functional properties in cellular assays using the know-how of IPI group and (III) evaluating their benefit for combined ICB therapy in relevant cancer models using the expertise of TIME group.

Methods

mAbs and Nbs against mouse and human receptor will be elicited using the magnetic GPCRliposome-based technology developed by IPI group allowing to purify and display high density of properly folded receptors for mouse or llama immunization. Mabs and Nbs will be selected by hybridoma and phage-display technologies, respectively. The validation of the binders and the characterization of their activity towards the receptor will be performed using tailor made cellular assays based on the NanoBiT/NanoBRET technologies at the DII. These assays will also be adapted for screening of small receptor modulators using the drug-screening platform of Dr. Yong-Jun Kwon. Finally, selected modulators will be evaluated alone or in combination with anti-PD-1 therapy using mouse melanoma and colon cancer models.

Project 2.6. - Pre-clinical evaluation of immunoconjugates activating NK cells in different cancer models

Introduction

As a novel cancer immunotherapeutic approach, we generated two types of immunoconjugates, called CoMiX, that enhanced complement directed activity and cytotoxicity at the surface of HER2-tumour cells or CD20-overexpressing lymphomas (Seguin-Devaux et al., 2019). Based on the CoMix technology (patent PCT/EP2017/062283), we have generated recombinant molecules targeting and activating NK cells against different cell types infected with various pathogens. These molecules increase also the cytotoxic activity of NK cells against K562 and Raji cancer cells through IFN-y release and cytotoxic protein degranulation. 

Pancreatic cancer is a cancer with high unmet medical needs. Pancreatic cancer has a catastrophic prognosis (recent five year survival in the USA: 9.3%) and moreover is expected to become the second most frequent cancer in 2030. Treatment is difficult, as in most cases diagnosis is made at late stages when tumor cell dissemination has taken place. Chemo- and targeted therapies provide only a limited increase of overall survival for these patients due to therapy resistance and a limited response to checkpoint inhibitors. The CEA expression is generally very high in pancreatic adenocarcinoma, Many pancreatic cancers are resistant to immunotherapeutic approaches due to their immune-hostile microenvironment 

Aim and Hypothesis

We propose here to evaluate novel immunoconjugates activating NK cell function in different models of cancer to evaluate their potency as compared to current therapies and in combination with checkpoint inhibitors. Based on our preliminary results, we will evaluate our therapeutic molecules against a hematopoietic cancer (B cell lymphoma), and pancreatic tumors. These immunoconjugates will allow investigating the therapeutic efficacy of innate immunity and whether our recombinant molecules can overcome the current resistance to the chemo- and targeted therapies in a model of pancreatic cancer. This PhD project may provide additional applications for CoMIX in cancer therapies and consolidate the potential of CoMIX against other solid tumors.

Methods

The PhD student will develop different in vitro (2D and 3D cellular models) and in vivo (mouse models of lymphoma and pancreatic cancer) cancer models, evaluate several molecules activating NK cells against tumor cells, and investigate their mode of action using different immunoassays.

Project 2.7. - Developing novel personalized therapies for colorectal cancer patients

Introduction

Targeted therapy can enhance the antitumor immune responses by releasing new antigens; this provides a theoretical basis for immunotherapy combined with targeted therapy as recently observed in lung cancer (Akbay et al., 2013; Liang et al., 2018). Along this line, the small REGONIVO phase Ib clinical trial (n=48 patients) very recently showed promising results on the combination of the tyrosine inhibitor Regorafenib and the anti-PD1 therapy Nivolumab in gastric and CRC patients (ASCO 2019). These initial results are promising and encourage to test more combinations of different tyrosine kinase inhibitors and immune-based therapies. 

Aim and Hypothesis

We hypothesize that the screening of single drugs and the combination thereof directly on patient material will lead to the development of new personalized therapy options. In addition, patient-derived models composed of tumor and stromal cells (note that stromal cells, among which fibroblasts, are known to be responsible for immune-suppression), as established over the past years within the MDM group, will help to understand the underlying mechanisms of innate/acquired resistance and thereby prevent therapy resistance in long-term treatments. 

Methods

Fresh biopsies from stage III and stage IV CRC patients will be collected at the different involved hospitals and directly transferred to LIH and the MDM group at UL. Screening of therapy options (including different combinations of targeted and immune-based therapies) directly on patient material will be done by applying the high throughput drug-screening library from LIH. Using in silico modelling, as previously described by our group in collaboration with the Sauter group at UL (Pacheco et al., 2019), we will further refine the list of tested drugs/drug combinations. Additionally, patient- derived models from the same patients (composed of tumor and stromal cells as well as cells from the normal counterpart tissue) will be used to (i) understand the mechanisms of innate and acquired resistance, (ii) to address the efficiency of the identified therapy options (iii) as well as study their potential to induce resistance in long-term treatments, by the means of molecular biology (i.e. bulk and single-cell transcriptomics combined with signaling network analysis) and functional assays (Cell Titer Glo viability assays). Finally, most promising pre-clinical candidates will be validated in our in house orthotopic mouse CRC models. 

Project 2.8. - Precision medicine for Glioblastoma patients: a patient-based pre-clinical platform for immuno-oncology

Introduction

Precision medicine provides a promising avenue for cancer patients, however robust preclinical models to predict sensitivity of novel treatments are key for advancing clinical oncology (Byrne et al., 2017; Gao et al., 2015). Based on a large platform of Glioblastoma patient-derived orthotopic xenografts (PDOXs), which retain histopathological, genetic and epigenetic features of patient tumors superior to in vitro cultures, we have successfully applied biomarker-driven drug treatment in PDOX in vivo and in PDOX-derived 3D tumor organoids ex vivo (Bougnaud et al., 2016; Abdul Rahim et al., 2017; Golebiewska et al., BioRvix, 2020). Both approaches, however, suffer from the reduction and/or absence of immune system components, limiting the opportunity to test novel immune modulating therapies.

Aim and Hypothesis

In this project we plan to expand our GBM PDOX platform to humanized mouse models. The immune component, i.e. peripheral blood mononuclear cells and/or patient-derived microglia/macrophages will also be combined with ex vivo 3D primary tumor organoids.

Methods

Relevant tumor-stroma interactions will be investigated at the single cell level (single cell RNA-seq, Flow and mass cytometry) to monitor tumor progression and treatment response. Standardized short-term tumor-immune co-cultures will be applied for high-throughput drug screening and successful compounds will be validated in humanized PDOX models in vivo.

Project 2.9. - Robotic / stereotactic Intracerebral Drug Infusions (Convection Enhanced Delivery, CED) for neurooncological diseases (Brain Tumors)

Introduction

The treatment of brain tumors remains challenging. Whereas the large majority of benign tumors can be treated surgically or by radiosurgery, a large majority of malignant brain tumors is still not curable. Whereas neuronavigation technologies have been coming up in Neurosurgery as a standard of care within the last two decades, innovative technologies, such as spectroscopic imaging (RS and MRS), as well as precision MRI, offer new possibilities for targeted treatment approaches. However, despite the development of many new and promising drugs, delivery through the Blood Brain Barrier (BBB) remains challenging.

Aim and Hypothesis

The aim of the project is it, to develop heat maps of spatial distribution of biochemical tumor properties by a combination of navigated RS, MRS and diffusional MRI and to combine those with models of drug distribution for CED by computer modelling of diffusional MRI images.

Methods

We will analyze RS with navigated RS fiber probes and integrate those into data sets of multiple 3D-MRI datasets. Therefore, we will develop heat maps to depict the biochemical tumor property distribution. By integration of diffusability (diffusion weighted MRI), we will develop models for fluid (drug) distribution within the individual patients brains. Those models will be tested by labelled Gadolinium particles.

Project 2.10. - Tumor mutational burden for personalized therapeutic decision making in cancer patients

  • Supervisor: Dr Barbara Klink

Introduction

Precision oncology, guided by next-generation sequencing of tumor DNA and RNA, implies the ability to predict which patients will likely respond to specific cancer therapies (Horak et al., 2017). Recent studies demonstrated the utility of tumor mutational burden (TMB) measured by wholeexome sequencing (WES) as a promising biomarker to identify patients most likely to benefit from immunotherapy across a wide range of cancer types (Chan et al., 2019). In the clinical setting, where WES is not routinely available, panel sequencing is emerging as the technology for TMB assessment. However, lack of harmonization between plat-forms and of robust predictive cut-offs are major limitations of panel-based TMB quantification (Fancello et al., 2019). Furthermore, there is an unmet clinical need for the assessment of TMB in liquid biopsies, since measuring TMB in circulating tumor DNA (ctDNA) presents a major detection challenge (Gandara et al., 2018).

Aim and Hypothesis

This project aims to evaluate and validate different strategies for reliable detection of TMB in solid tumor tissues and blood-derived ctDNA to identify patients eligible for immunotherapy. Implementation of TMB-assessment in translational oncology programs will improve personalized therapeutic decision-making and patient management in the future.

Methods

We will perform WES and panel sequencing on tumor tissues and corresponding blood samples as well as ctDNA to identify actionable somatic and hereditary alterations, including TBM. Bioinformatics pipelines to determine TBM from different data sets will be developed utilizing own and public available data sets. Results will be correlated with clinical data. We will integrate our approaches in translational research programs such as PFP and in multidisciplinary tumor boards to enable fast translation into the clinics.