Classing it up to get noticed

MHC class 1 antigen display in dendritic cells and neuroblastoma

Spel, Lotte

Promoter:
Prof.dr. A.B.J. (Berent) Prakken
Co-promoter:
Dr. M.L. (Marianne) Boes, Dr. S. (Stefan) Nierkens & Dr. J.J. (Jaap Jan) Boelens
Research group:
Prakken , Boes , Nierkens
Date:
February 8, 2018
Time:
12:45 h

Summary

In this thesis I have explored the process of MHC-1-mediated antigen presentation in two distinctive cell types. In the first chapters I describe mechanisms in dendritic cells that affect antigen (cross)-presentation and its application for vaccination. In the later chapters I focus on neuroblastoma tumor cells and possible ways to increase their antigen presentation levels towards cytotoxic T-cells.

Dendritic cells are pivotal players that bridge innate and adaptive immunity. As sessile sentinels in barrier tissues, they scan their local environment using the receptors on their cell surface. When sensing inflammatory or danger signals, DCs engulf the receptor-bound material and process it for antigen presentation to T-cells. This process of antigen presentation has great therapeutic potential in the cancer field, driving the development of innovative immunotherapies.

DCs are able to engulf tumor-derived material and cross-present tumor-derived antigen fragments to CD8+ T-cells. In Chapter 2, I give an overview of recent literature about cross-presenting tumor cell material by dendritic cells. Especially necrotic or apoptotic tumor cells are recognized by dendritic cells and processed for antigen cross-presentation. Signals on or secreted by the dying cells are important for the intracellular routing of antigens towards the cross-presentation pathway. Within this pathway, a balanced play between antigen degradation and antigen preservation serves to maximize cross-presentation of the phagocytosed cargo and thereby the elicitation of an anti-tumor T-cell response.

Tipping the scale towards antigen preservation causes a complete stop of antigenic peptide production and subsequent loss of antigen cross-presentation, as described in Chapter 3. Here, cowpox virus-derived protein CPXV012 inhibits direct MHC-1 antigen presentation in infected cells. In this chapter CPXV012 is delivered into the endosomal pathway of dendritic cells as a soluble protein. In this endosomal environment, soluble CPXV012 colocalizes with endocytosed antigen. By blocking endosomal acidification, it prevents the degradation of antigen, a process that is required to liberate antigenic peptide to be cross-presented by MHC-1.

Peptides that are produced by endosomal processing of antigen do not only reach MHC-1 for presentation, but also MHC 2. MHC 2 presents antigen to helper CD4+ T-cells that aid in the priming of CD8+ T-cells. The benefit of reaching both CD4+ and CD8+ T-cells when initiating an immune response is investigated in Chapter 4. Antigen-specific CD4+ T-cells stimulate DC-mediated priming of antigen-specific CD8+ T-cells recognizing an epitope of the same antigen. This is accomplished by the production of IFNγ upon binding peptide/MHC 2-complexes on the DC cell surface. These findings are relevant, as after stem cell transplantation, when a newly infused immune system is in development, patients may suffer from viral reactivations such as CMV. In these patients, the presence of CMV-specific CD4+ T-cells precedes the induction of CMV-specific CD8+ T-cells. After viral clearance, only the CMV-specific CD8+ T-cells are still detected.

Given the pivotal role of dendritic cells in eliciting cellular immunity, they may be useful for therapeutic purposes. In Chapter 5, I explore this option using a viral vector that targets dendritic cells in vitro and in vivo. This viral vector is a derivative from the bunyavirus Rift Valley Fever virus and genetically engineered to express either pp65 (human) or OVA (mouse) antigen. Using the RVFV vector we detected antigen-specific CD8+ T-cell activation, both using in vitro cultures and using in vivo mouse models. Furthermore, in prophylactic and therapeutic settings of vaccination this viral vector was able to confer protection against a lymphoma tumor challenge.

Neuroblastoma is the most deadly pediatric solid tumor and currently lacks a cellular immunotherapeutic treatment strategy. Tumor-specific CD8+ T-cells may be effective to destroy neuroblastoma tumor cells. However, neuroblastoma tumors were shown not to be immunogenic due to low MHC-1 expression levels and lack of broadly- expressed antigens. How to increase neuroblastoma immunogenicity is discussed in Chapter 6. Cells with no/low MHC-1 levels can become targets for natural killer (NK) cells. Indeed, part of the neuroblastoma cells could be killed by NK cells in vitro. Those cells surviving the NK cell attack showed upregulation of MHC-1 and could therefore become targets for CD8+ T-cells. This hypothesis required the identification of a neuroblastoma-expressed antigen. I show that the preferred antigen in melanoma (PRAME) is expressed in >90% of high-risk neuroblastoma tumors and PRAME-specific T-cells are able to recognize neuroblastoma cells, but only when MHC-1 levels are upregulated e.g. by previous NK cell attack.

The regulation of MHC-1 expression in neuroblastoma is further investigated in Chapter 7. MHC-1 expression, and thus T-cell recognition of neuroblastoma, appears to depend on the activation of the transcription factor NFκB. Through systematic gene deletions, I have identified factors that suppress NFκB activity in neuroblastoma cells. TNXP1* limits NFκB activity in neuroblastoma by destabilizing the canonical IKK-complex. NYBP1* on the other hand shows an inhibitory effect on NFκB that is independent of the canonical IKK-complex. NYBP1 aids OTUD7B to de-ubiquinate TRAF3, thereby stabilizing TRAF3 protein and ensuring blockage of both canonical and non-canonical NFκB pathways. Depletion of NYBP1 results in degradation of TRAF3 and subsequent accumulation of NIK, which initiates activation of non-canonical NFκB. Both TNXP1* and NYBP1* tumor expression levels are correlated with worse survival of neuroblastoma patients, suggesting they could be relevant targets for the development of future (immuno)therapies against neuroblastoma.

Finally, the role of NFκB suppression in neuroblastoma development, immunogenicity and immunotherapy is reviewed and discussed in Chapter 8.

*Gene names are changed.

Full text