Establishment and comparison of two-dimensional and three-dimensional models of gliomas

Alexandra Gee, Dr Ashraf Zaman, Professor Joseph Powell, Madhavi Pandya, Dr Walter Muskovic, Emily Holton

Introduction:

Diffuse gliomas are incurable primary neoplasms that originate in supportive glial cells of the central nervous system and are the most common primary brain tumours in adults. Gliomas have complex inter- and intra-tumoural heterogeneity, rendering them highly recurrent and difficult to treat. Consequently, traditional monolayer cultures used for drug screening do not accurately capture these complexities, leading to limited successful drug trials. Out of the search for models beyond two-dimensional cultures have emerged three-dimensional cultures of gliomas.

Aim/Purpose:

The study refines the protocol to create three-dimensional domes, tumour spheroids and explants from patient-derived cryopreserved tissue and compares the growth and architecture of these to monolayer cultures. The study uses single-cell RNA sequencing to compare the transcriptome of 2D cultures to their original fresh tissue specimens.

Method/Description:

A total of seventeen surgically resected and cryopreserved glioblastomas (GBMs), oligodendrogliomas and astrocytomas were obtained from the Charlie Teo Foundation Brain Tumour Bank. The protocol was refined to create 3D domes, tumour spheroids and explants from cryopreserved tissue. For all cultures serum-free media was used to maximise retention of the transcriptional profiles and tumorigenicity of the tissue. Single cell RNA sequencing was performed to compare the fresh tissue with their respective cultures. Cellular states were determined by gene sets defined by Neftel et al.

Results:

 ScRNA-seq revealed the presence of both malignant and stromal cells in fresh tissue specimens. The two sequenced GBM tumours had higher proportions of myeloid cells than the astrocytoma. The GBM tumours also had a higher proportion of malignant cells in the mesenchymal (MES) state. Comparing fresh tissue and 2D cultured tissue, there was a change in the composition of malignant cell states. Samples showed an increased proportion of hypoxia -independent (MES1) and -dependent (MES2) signatures with virtually a complete loss of neural progenitor (NPC) and oligodendrocyte precursor (OPC) states. 3D cultures, namely domes, explants and tumour spheroids were established from cryopreserved patient-derived tissues using media free of serum and growth-factors. There was a gross change in the morphology and size of cultures with time.

Discussion: 

Authentic pre-clinical models of gliomas are needed for the development of new treatment options and the opportunity for personalised medicine. In this study, single-cell RNA sequencing revealed a loss of non-malignant cell populations and a change in the composition of malignant cell states after two-dimensional culture. Non-malignant cells make up the tumour microenvironment which plays an important role in tumour progression and therapeutic resistance, and may influence the composition of cellular states.  Within the malignant cell population, increased proportions of MES cell states in 2D cultures may indicate a transcriptomic shift. Cells in these cultures displayed high gene expression levels for invasiveness. This aligns with the MES subtype correlating to poor survival outcomes.

While studies have suggested 3D cultures may better retain the original tumour heterogeneity, a benefit of 2D cultures is that they are better supporters of tumour cell growth from cryopreserved glioma tissues. This study found challenges in long-term culture maintenance for 3D cultures, including a loss of peripheral cells, likely due to mechanical forces induced by media changes and the movement of plates, as well as the potential development of necrotic cores.

Conclusions:

The study reinforces the notion that 2D cultured tumours, even at a low passage, do not retain the heterogeneity of their primary tumours at a single-cell level. The findings reiterate the importance of optimisation of culture conditions to mimic in vivo conditions in order to preserve the heterogeneity of the original tumour.

References

1. Neftel, C., et al. An Integrative Model of Cellular States, Plasticity, and Genetics for Glioblastoma. Cell 178, 835-849.e821 (2019).
2. Wu, H., et al. Single-cell RNA sequencing reveals tumor heterogeneity, microenvironment, and drug-resistance mechanisms of recurrent glioblastoma. Cancer Sci. 114, 2609-2621 (2023).
3. Ciechomska, I.A., et al. Exploring Novel Therapeutic Opportunities for Glioblastoma Using Patient-Derived Cell Cultures. Cancers (Basel) 15, 1562 (2023).
4. Patel, A.P., et al. Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma. Science 344, 1396-1401 (2014).
5. Wang, L.-B., et al. Proteogenomic and metabolomic characterization of human glioblastoma. Cancer Cell 39, 509-528.e520 (2021).
6. Gisina, A., et al. Glioma Stem Cells: Novel Data Obtained by Single-Cell Sequencing. in International Journal of Molecular Sciences, Vol. 23 (2022).
7. Jacob, F., Ming, G.-l. & Song, H. Generation and biobanking of patient-derived glioblastoma organoids and their application in CAR T cell testing. Nature Protocols 15, 4000-4033 (2020).
8. Singh, S.K., et al. Critical role of three-dimensional tumorsphere size on experimental outcome. BioTechniques 69, 333-338 (2020).