Aligned Nanotopography Promotes a Migratory State in Glioblastoma Multiforme Tumor Cells
Glioblastoma multiforme (GBM) is an aggressive, Grade IV astrocytoma with a poor survival rate, primarily due to the GBM tumor cells migrating away from the primary tumor site along the nanotopography of white matter tracts and blood vessels. It is unclear whether this nanotopography influences the biomechanical properties (i.e. cytoskeletal stiffness) of GBM tumor cells. Although GBM tumor cells have an innate propensity to migrate, we believe this capability is enhanced due to the influence of nanotopography on the tumor cells’ biomechanical properties. In this study, we used an aligned nanofiber film that mimics the nanotopography in the tumor microenvironment to investigate the mechanical properties of GBM tumor cells in vitro. The data demonstrate that the cytoskeletal stiffness, cell traction stress, and focal adhesion area were significantly lower in the GBM tumor cells compared to healthy astrocytes. Moreover, the cytoskeletal stiffness was significantly reduced when cultured on aligned nanofiber films compared to smooth and randomly aligned nanofiber films. Gene expression analysis showed that tumor cells cultured on the aligned nanotopography upregulated key migratory genes and downregulated key proliferative genes. Therefore, our data suggest that the migratory potential is elevated when GBM tumor cells are migrating along aligned nanotopographical substrates.As GBM is categorized as a Grade IV astrocytoma, the difference in cytoskeletal stiffness between GBM tumor cells and non-cancerous healthy primary astrocytes was measured using atomic force microscopy (AFM). The cytoskeletal stiffness was tested on two GBM cell lines (U87MG and A172), and primary GBM CICs (BT145). Conventional thought is that primary GBM tumor cells were derived directly from genetically mutated astrocytes or glial precursor cells (i.e. EGFR amplification/mutation, PTEN loss/mutation, etc.)36. Therefore, primary rat post-natal day 2 astrocytes and mouse neural stem cells were used as the non-cancerous, healthy cells. Average stiffness measurements and representative images for each cell type are shown in Fig. 1. Astrocytes were significantly stiffer than each GBM tumor cell type, with an average stiffness of 4184 ± 102.3 Pa (p < 0.0001). There was no statistical difference between the two GBM cell lines, U87MG and A172 tumor cells, which had an average stiffness of 1315 ± 39.98 Pa and 1138 ± 68.58 Pa, respectively. Primary GBM CICs, BT145, were statistically less stiff than the primary astrocytes and the GBM tumor cell lines (p < 0.01), with an average stiffness of 653.3 ± 35.37 Pa (Fig. 1A). Finally, NSCs had a similar stiffness to the CICs when plated on laminin (data not shown). Morphologies of the cells were also noticeably different between the various cell types (Fig. 1D). Astrocyte morphology was more spread on the TCPS compared to the spindled morphology exhibited in the tumor cells.As a healthy cell undergoes oncogenic mutations, many cellular attributes are altered resulting in the abnormal growth and migratory behavior of cells15. Due to the role of the cytoskeleton on cell migration, previous studies identified biophysical attributes, specifically cytoskeletal stiffness, as a biomarker for invasive potential in a variety of cancers. Tumor studies using AFM to measure cytoskeletal stiffness have shown that more invasive breast, ovarian, and prostate cancer cells are less stiff than their less invasive, benign, or healthy cell counterparts18,20,37. Significant reductions in stiffness, 2–5 fold, were observed between immortalized ovarian surface epithelial cells and different ovarian cancer cell lines20. In addition, Andolfi et al. showed that high and low grade glioma stem cells are less stiff than the non-tumorigenic glioma-associated-stem cells23. Similar results were observed in our experiments with the GBM tumor cells, as the highly invasive CICs were more compliant than the U87MG and A172 tumor cell lines, which together, were significantly softer than non-cancerous astrocytes. During invasion, cancer cells need to deform their bodies in order to conform to the surrounding tissue and migrate. The reduced stiffness indicates a greater capability to deform within the aggressive tumor cells, which facilitates the migration and invasion through the surrounding ECM leading to secondary tumor sites15. In addition, NSCs have also been shown to be highly migratory within the brain. Due to their enhanced migratory capacity toward cancerous/diseased tissue, these cells have been researched as a tool to deliver therapies to tumor cells38,39.Human GBM tumor cell lines, U87MG and A172 (ATCC, Manassas, VA), were maintained and cultured in DMEM (Corning, Corning, NY), supplemented with 10% fetal bovine serum (FBS, Gemini Bio-Products, West Sacramento, CA), 1% penicillin-streptomycin, 1% L-glutamine, and 1% non-essential amino acids (Corning). U87MG cells expressing enhanced green fluorescent protein (eGFP) were generously donated by Dr. Ravi Bellamkonda (Georgia Institute of Technology). A172 cells were transfected using GFP-Actin fusion lentiviral particles (GenTarget Inc., San Diego, CA) and positive colonies were selected for expansion. Primary CICs, BT145 were generously gifted by Dr. Rosalind Segal (Dana-Farber Cancer Institute). The GBM CICs were cultured in DMEM/F12 (Corning), supplemented with B27 (Gibco, Life Technologies Grand Island, NY), 15 mM HEPES (Alfa Aesar, Ward Hill, MA), 20 ng/mL EGF (Invitrogen, Carlsbad, CA), and 20 ng/mL FGF (Invitrogen). Differentiated BT145 tumor cells were grown in DMEM supplemented with 10% FBS. Primary rat astrocytes were isolated from Neonatal Sprague-Dawley pups (post-natal day 2) and maintained in Neurobasal Medium (Gibco), supplemented with 10% FBS and 1X GlutaMAX (Gibco). Mouse neural stem cells (Cyagen) were maintained and cultured in NeuroCult™ NSC Basal Medium (STEMCELL Technologies) supplemented with 1% penicillin-streptomycin, B27, 5 μg/mL FGF, 10 μg/mL EGF, and 0.2% v/v heparin.We would like to thank Dr. Rosalind Segal (Dana-Farber Cancer Institute) and Dr. Ravi Bellamkonda (Georgia Institute of Technology) for their generously gifting the BT145 and U87MG-eGFP tumor cells, respectively. We would also like to thank Dr. Sakthikumar Ambady (WPI) for his assistance in generating the A172-GFP Actin cells, Kimberly Ornell (WPI) for maintaining and seeding the CICs, Dalia Shendi (WPI) for maintaining and seeding the neural stem cells, Dr. Robert Gegear (WPI) for his assistance with statistical analysis, and Dr. Roger Ristau (University of Connecticut) and Anbo Wang (WPI) for their assistance using the SEM.Author Contributions A.J. conceived and designed the study. A.B. performed the experiments. G.T. helped with performing the measurements using the AFM and J.G. helped obtain and analyze the measurements using TFM. Q.W. provided guidance for the AFM and TFM experiments. The manuscript was written by A.B. and A.J. which was edited and reviewed by all authors.