Refine
Has Fulltext
- yes (2) (remove)
Is part of the Bibliography
- yes (2)
Year of publication
- 2015 (2) (remove)
Document Type
- Doctoral Thesis (2)
Language
- English (2)
Keywords
- Krebs <Medizin> (2) (remove)
Institute
- Graduate School of Life Sciences (2) (remove)
The recently discovered human DREAM complex (for DP, RB-like, E2F and MuvB complex) is a chromatin-associated pocket protein complex involved in cell cycle- dependent gene expression. DREAM consists of five core subunits and forms a complex either with the pocket protein p130 and the transcription factor E2F4 to repress gene expression or with the transcription factors B-MYB and FOXM1 to promote gene expression.
Gas2l3 was recently identified by our group as a novel DREAM target gene. Subsequent characterization in human cell lines revealed that GAS2L3 is a microtubule and F-actin cross-linking protein, expressed in G2/M, plays a role in cytokinesis, and is important for chromosomal stability.
The aim of the first part of the study was to analyze how expression of GAS2L3 is regulated by DREAM and to provide a better understanding of the function of GAS2L3 in mitosis and cytokinesis.
ChIP assays revealed that the repressive and the activating form of DREAM bind to the GAS2L3 promoter. RNA interference (RNAi) mediated GAS2L3 depletion demonstrated the requirement of GAS2L3 for proper cleavage furrow ingression in cytokinesis. Immunofluorescence-based localization studies showed a localization of GAS2L3 at the mitotic spindle in mitosis and at the midbody in cytokinesis. Additional experiments demonstrated that the GAS2L3 GAR domain, a putative microtubule- binding domain, is responsible for GAS2L3 localization to the constriction zones in cytokinesis suggesting a function for GAS2L3 in the abscission process.
DREAM is known to promote G2/M gene expression. DREAM target genes include several mitotic kinesins and mitotic microtubule-associated proteins (mitotic MAPs). However, it is not clear to what extent DREAM regulates mitotic kinesins and MAPs, so far. Furthermore, a comprehensive study of mitotic kinesin expression in cancer cell lines is still missing.
Therefore, the second major aim of the thesis was to characterize the regulation of mitotic kinesins and MAPs by DREAM, to investigate the expression of mitotic kinesins in cancer cell line panels and to evaluate them as possible anti-cancer targets.
ChIP assays together with RNAi mediated DREAM subunit depletion experiments demonstrated that DREAM is a master regulator of mitotic kinesins. Furthermore, expression analyses in a panel of breast and lung cancer cell lines revealed that mitotic kinesins are up-regulated in the majority of cancer cell lines in contrast to non-transformed controls. Finally, an inducible lentiviral-based shRNA system was developed to effectively deplete mitotic kinesins. Depletion of selected mitotic kinesins resulted in cytokinesis failures and strong anti-proliferative effects in several human cancer cell lines.
Thus, this system will provide a robust tool for future investigation of mitotic kinesin function in cancer cells.
Background
GDF-15 is a divergent member of the TGF-superfamily, which was first described as macrophage inhibitory cytokine-1 (MIC-1), revealing an immune modulatory function. GDF-15 is a soluble protein which is, under physiological conditions, highly expressed in the placenta and found in elevated levels in blood sera of pregnant women. Apart from the placenta, GDF-15 is expressed in healthy tissue, albeit to a lower extent and overexpressed in many solid tumors. A variety of different functions are attributed to GDF-15 in healthy as well as diseased humans. On the one hand, GDF-15 is required for successful pregnancy and low GDF-15 serum levels during pregnancy correlate with fetal abortion. On the other hand, overexpression of GDF-15, which can be observed in several malignancies is correlated with a poor prognosis. Furthermore, tumor derived GDF-15 leads to cancer associated anorexia-cachexia syndrome in mice. The aim of my PhD thesis was to further investigate the role of GDF-15 as an immune modulatory factor in cancer, in particular, by inhibiting the target molecule in vitro and in vivo. Therefore, the main focus was placed on the generation and characterization of monoclonal GDF-15 specific blocking antibodies, which were tested in vitro and in vivo, which represents a substantial part of my work.
Results
Here, GDF-15 was shown to be highly expressed in human gynecological cancer and brain tumors. We could then demonstrate that GDF-15 modulates effector immune cells in vitro. GDF-15 mediated a slight downregulation of the activating NKG2D receptor on NK and CD8+ T cells, which is crucial for proper anti-tumoral immune responses. Furthermore, we could demonstrate that GDF-15 reduces the adhesion of CD4+ and CD8+ T cells on endothelial cells in vitro. A negatively affected trans-endothelial migration of leukocytes into inflamed tissue could explain the low T cell infiltration in GDF-15 expressing tumors, which were observed in vivo, where mice bearing (shRNA mediated) GDF-15 deficient glioma cells revealed enhanced immune cell infiltrates in the tumor microenvironment, compared with the GDF-15 expressing control group. Those animals further exhibited a decreased tumor growth and prolonged survival. GDF-15 is a soluble protein, secreted by more than 50 % of solid tumors and associated with grade of malignancy. Therefore a neutralizing monoclonal antibody to GDF-15 was assumed to be an auspicious therapeutically anti-cancer tool. Such an antibody was thus generated in GDF-15 knock out mice against human GFD-15. Amongst many clones, the GDF-15 antibody clone B1-23 was found to be applicable in Western Blot as well as in ELISA techniques, detecting a three-dimensional epitope of the mature GDF-15 dimer with high affinity and specificity. To enable the humanization for a later administration in humans, the variable regions of antibody B1-23 were identified by a special PCR method using degenerate primers and cloned into a sequencing vector. The sequence obtained thereby enabled the generation of chimeric and humanized B1-23 variants. After further comprehensive characterization, the original mouse antibody B1-23 as well as the chimeric antibody (ChimB1-23) and the humanized B1-23 antibody (H1L5) were applied in a melanoma xenograft study in vivo. None of the antibodies could significantly inhibit tumor growth. .However of utmost importance, body weight loss mediated by tumor derived GDF-15 could be significantly prevented upon administration of all three GDF-15 specific antibodies, which confirmed the antagonizing functionality of the immunoglobulin.
Conclusion
GDF-15 is a promising cancer target, involved in tumor progression and cancer related cachexia. A monoclonal GDF-15 antibody was generated, which served on one hand as a tool for molecular biological applications (Western Blot, ELISA, etc.) and on the other hand was applied as an antagonizing antibody in vitro and in vivo. Even though tumor growth inhibition by GDF-15 depletion in T cell deficient athymic mice failed using B1-23, the same antibody and derivates thereof (chimeric and humanized) impressively prevented tumor associated cachexia in UACC-257 melanoma bearing nude mice. The missing anti-tumor effect in our own melanoma model in nude mice can only partially be explained by the missing secondary immunity, in particular cytotoxic T cells, in the athymic animals, since in a similar melanoma model, performed by an external company, a tumor reduction in immunocompromised animals was observed, when B1-23 was administered. These findings support the idea that T cells are substantial for an effective tumor immunity and are in line with the results of the syngeneic, T cell comprising, mouse glioma model, where silencing of tumor expressed GDF-15 led to an enhanced intratumoral T cell infiltration and a prolonged survival.
Taken together our data allow for the conclusion that tumor associated cachexia can be combatted with the GDF-15 antibody B1-23. Further, B1-23 might elicit direct anti-tumor effects in immune competent models, which contain T cells, rather than in an athymic, T cell deficient nude mouse model.