Abstract (eng)
The tumour microenvironment (TME) influences the way tumour cells (TCs) migrate, proliferate and how they “immunologically behave”. To elucidate the underlying molecular mechanisms of the cross-talk between TCs and cells of the TME with a particular focus on immune cells, we made use of our novel 3D models consisting of TCs (non-small cell lung cancer (NSCLC)-derived cell lines NCI-H157 and NCI-H1437), stroma cells (fibroblasts) and immune cells (focus on donor-derived primary NK cells; PNKs) embedded in a Matrigel/collagen matrix. Emerging evidence indicates that particularly epigenetic events in TCs influence their immunological behaviour. Understanding the underlying biological mechanisms may lead to identification of novel therapeutic targets inside TCs that increase their immunogenicity. Modulating those regulators would allow restoring the intrinsic immune response towards neoplasms without interfering with or specifically targeting immune cell functions. One of those potential novel epigenetic targets is the bromodomain PHD finger transcription factor (BPTF), a reader of the histone code and an essential subunit of the nucleosomal remodeling factor (NURF). In the course of my Master thesis I studied both, the effect of genetic and pharmacologic inhibition of BPTF on the activation of PNKs. For this, we applied BI90, a potent and specific proprietary inhibitor of the bromodomain (BRD) of BPTF. Treating respective mono-cultures with BI90 did not generate any adverse effect per se. Thus, BI90 could be applied for pharmacological BPTF inhibition in simple spheroid (floater) and complex 3D embedded co-cultures. In all 3D models killing of TCs by activated PNKs was significantly increased upon BI90 treatment accompanied by an increase in GZMB and IFNG levels in the supernatant of the respective 3D cultures. Interestingly, the rate of infiltration and activation of PNKs was lower in 3D embedded models, clearly supporting our previous finding that the cross-talk between TCs and stromal cells (FBs) induces a strong immunosuppressive environment. As pharmacologic inhibition of BPTF most likely modulate only a very focused set of genes in TCs, we next performed genetic perturbation experiments by either knocking down (KD via siRNA) or knocking out (KO via CRISPR/Cas9) BPTF in TCs. Importantly, NCI-H1437 BPTF KO cells (pools) and NCI-H157 BPTF KD cells were again significantly better killed in co-cultures with PNKs compared to the nontarget control group. According to the location of sgRNAs used, domain specific effects in depletion assays were observed. Strongest effects were observed targeting the N-terminal domain, while those recognizing more C-terminal domains (PHD-finger, BRD) showed weaker or no effects. In order to gain a more detailed biological understanding, BPTF KO single cell clone cultures were isolated and gene expression profiling of five different BPTF KO single cell clones was performed. Similarly, a detailed time-resolved transcriptional analysis of the BI90-treated cultures was performed. According to a recent publication by Mayes and colleagues, heparanase (HSPE) might be the reason for the executed immunosuppressive effect by TCs via BPTF. HSPE cleaves heparan sulfates, known co-ligands of PNKs natural cytotoxicity receptor. In contrast, we found that HPSE transcript abundance stayed unaffected both in BI90-treated as well as in BPTF KD or KO cultures. However, we found many genes affected (down-regulated) by BPTF KO which are mainly involved in cell cycle regulation. Interestingly, it is known that inhibitors targeting cell cycle regulators, not only arrest cells but also increase their immunogenicity. Comparison of transcripts being up- or down-regulated in at least three out of five BPTF KO single cell clones with the those of BI90-treated tumour cells revealed only an overlap of six down-regulated (CLSPN, RARRES1, KNTC1, KIF11, KIAA0101 and XRCC2) and two up-regulated genes (IFIH1 and GMFG). In particular, IFIH1, also known as MDA5 (Melanoma Differentiation-Associated protein 5) might substantially contribute to the immunogenic visibility of TCs. IFIH1 is part of the RIG-I-like receptor (RLR) family and functions as a pattern recognition receptor (recognizing dsRNA) that is a sensor for viral infection. It is known that cancer cells can mimic a viral infection to activate RNA sensing pattern recognition receptors and interferon response pathways subsequently stimulating cytotoxic immune cells including NK cells. Consequently, TCs get killed via induced cell lysis and apoptosis. Therefore, impaired NURF function – due to BPTF KO/inhibition - may induce cellular stress and expression of endogenous retroviral dsRNA leading to an increase in immunogenicity and triggering an IFIH1-dependet type 1 interferon response. Clearly, further studies are needed to proof this hypothesis. However, despite many open questions in the context of the underlying molecular mechanisms of BPTF-mediated immune escape, we showed that BPTF can be regarded as a highly promising target in the immunological approach of combating human cancer.