(476d) Interaction of Human Umbilical Vein Endothelial Cells (HUVECs) and LN229 Glioblastoma Cells in a 3-Dimensional Hydrogel Perivascular Niche Model

Glioblastoma (GBM) is the deadliest form of brain cancer, with a median survival time of 15 months ¹. Surgical resection, chemotherapy, and radiotherapy have not been able to extend this survival time ¹. One of the problems in resecting GBM is the diffuse invasion of cancer cells through the brain parenchyma. It has been suggested that one of the mechanisms for diffusion of these cells is by migration along blood vessels ¹. Vasculature plays an important role in cell viability in all tissues, as well as in cancer. Cell growth is dependent in part on nutrient supply and diffusion of oxygen, both of which can be controlled by access to the vasculature. The area near a blood vessel has been described as the perivascular niche (PVN) ¹. Understanding how GBM interacts with endothelial cells in the PVN may provide a deeper understanding of this cancer.

Methods

Our goal was to create a 3-dimensional (3D) hydrogel PVN to mimic the natural brain extracellular matrix (ECM) in close proximity to vasculature. To this end we have designed both a collagen I, as well as utilized a commercially available hyaluronic acid-gelatin, Hystem-C® (Sigma) hydrogels. Hyaluronic acid is the main component of the brain ECM ¹, making the Hystem-C® system a biologically relevant scaffold. Collagen I is not commonly found in the normal brain ECM ¹. However, it is found in GBM tumors and may recapitulate the PVN. Therefore, these scaffolds have the potential to recapitulate the mechanical properties and protein composition of normal and cancerous environments.

A human GBM cell line LN229 (ATCC) was utilized since these cells display a mesenchymal subtype, are moderately invasive, and have been shown to display similar mutations to tumors in vivo ². HUVEC networks were used as a vasculature mimic. While the vasculature is comprised of different cell types, endothelial cells are a principal component and have been commonly used as an in vitro model for vascular networks ¹.

3D hydrogel systems were assembled in stratified layers in glass-bottomed well plates. First, a thin gel of collagen (either 1.1 mg/mL or 2.2 mg/mL) was deposited to prevent cells from adhering to the glass surface beneath. This gel contained 1.5% (v/v) blue fluorescent beads (0.1 µm diameter) to enable visualization of the gel and the bottom of the well. Thereafter, either a collagen I or a Hystem-C® solution was added and allowed to gel. The concentrations of collagen were based upon an earlier study that demonstrated the elastic moduli to lie in the 0.2 – 2.2 kPa range for these input concentrations ³. Cells were seeded on fibronectin-coated glass coverslips and allowed to adhere overnight. The coverslips were then inverted and placed above the stacked gels.

Results

The cells were visualized via confocal microscopy every 2 days. LN229 cells were labeled with a non-toxic fluorescent cytoplasmic dye. By day 6 the cells had migrated through the gel to the bottom bead layer. Fluorescent z-stack images were taken from 3 locations per sample. Glial fibrillary acidic protein (GFAP) immunofluorescence staining was performed on 3D stratified hydrogels as well as 2D monolayer cultures. Overall, higher fluorescence was observed in LN229 cells cultured in collagen gels in comparison to Hystem-C® scaffolds. Hystem-C® cultures resulted in cells that exhibited a round morphology and connected via string-like extensions, whereas collagen cultures displayed elongated cells whose morphology was similar to those observed in 2D monolayer cultures.

HUVECs were cultured in the stratified 3D hydrogels up to 14 days. HUVECs began to connect and form network-like structures around day 7. Between days 10-14, 3D connectivity was observed between HUVECs. In some cultures, vascular endothelial growth factor (VEGF), was covalently linked to the bottom gels. While VEGF resulted in enhanced HUVEC proliferation it did not appear to increase network formation.

Future and Ongoing Work

Our ongoing work is focused on investigating the interactions between LN229 and HUVECs in an in vitro perivascular niche model. In the future, we aim to include stromal cells such as macrophages and astrocytes to obtain a comprehensive understanding of the PVN.

References

1 Davide Schiffer, L. A., Christina Casalone, Critstiano Corona, Marta Mellai. Glioblastoma: Microenvironment and Niche Concept. Cancers 11, doi:10.3390/cancers11010005 (2018).

2 Nobuaki Ishii, D. M., Adrian Merlo, Mitsuhiro Tada, Yutaka Sawamura, Annie-Claire Diserens, Erwin G. Van Meir. Frequent Co-Alterations of TP53, p16/CDKN2A, p14^ARF, PTEN Tumor Supressor Genes in Human Glioma Cell Lines. Brain Pathology 9, 469-479, doi:10.1111/j.1750-3639.1999.tb00536.x (1999).

3 Andrew J. Ford, S. M. O., Padmavathy Rajagopalan. Fibroblasts stimulate macrophage migration in interconnected extracellular matrices through tunnel formation and fiber alignment. Biomaterials 209, 88-102, doi:10.1016/j.biomaterials.2019.03.044 (2019).