Optimization and validation of a species-specific in vivo approach for studies on growth and progression of neural crest derived tumours

Abstract: Background: Despite great advantages in cancer therapeutics during the last decades, therapy-resistance remains a problem in the majority of malignancies. Hence, development of more effective therapeutics is critical to enhance survival of patients with cancer. A big challenge for therapy development is low efficacy upon clinical translation; therapies proven highly effective in animal trials often exhibit a poor effect or even fail in the clinic. This indicates the need for more relevant preclinical in vivo model systems for cancer. Malignant melanoma and pediatric neuroblastoma, occurring in derivatives of the neural crest, are both exhibiting a high degree of therapy-resistance, with a high number of deaths as a consequence. These malignancies are characterized by a significant plastic developmental capacity with a tumour phenotype extensively depending on the surrounding microenvironment. Thus, they provide excellent opportunity for studies on interactions between the tumour and adjacent supportive tissue. Objective: The overall aim of this thesis is to optimize and validate a human in vivo model system, based on human embryonic stem cell derived teratomas (hEST-model), for growth and progression of neural crest derived malignancies. Further on, I aim to elucidate whether human tumours grafted in such system are having a higher resemblance to clinical tumours, as compared to conventionally used xenografts. Results: Based on a thorough kinetic study of development in hESC derived teratoma from the cell line HS181, we chose day 45 as the time point for tumour inoculation. At this time, mature embryonic tissues are formed and human vascularisation is initiated. Also, this allows further grafting of the tumour cells without affecting the animals well-being and occurrence of necrotic areas. The capacity for teratoma formation was also explored using a subline of HS181 with altered karyotype, i.e. trisomic for chromosome 12. Results revealed that these cells are capable of forming teratomas, however with a skewed tissue contribution with higher frequency of renal formation. Cell lines from malignant melanoma and neuroblastomas were grafted in the hEST-model and compared to the corresponding xenograft. For the melanomas, striking differences in expression of markers related to melanocytic differentiation was seen, indicative of induced differentiation from the xenograft-environment. Also, a dedifferentiated and invasive subpopulation of melanoma cells was detected, but not present in the xenograft, indicating species-specific interactions upon migration and invasion into surrounding stroma. Grafting of three different neuroblastoma cell lines in the hEST-model resulted in a general finding of higher histological heterogeneity and more resemblance to clinical neuroblastomas, as compared to the xenografts. Grafting of both tumour types in the hEST-model induced an extensive neo-vascularisation of human origin in the surrounding mesenchyme indicating species-specific effects. Conclusions: The results presented in this thesis indicate that a human in vivo model system for cancer based on hESC-derived teratomas add significant importance for preclinical cancer studies. The embryonic environment of the teratoma is probably most relevant for grafting of embryonic tumours, indicated also by our results using pediatric neuroblastoma. Altogether, the hEST-model provides unique possibilities to study human tumours in a species-specific environment, and is therefore suggested a well-needed complement to current preclinical in vivo models.