Developing a microfluidic platform for studying the interaction between perivascular niche and glioblastoma

Introduction: Glioblastoma (GBM) is the most common primary malignant brain tumor, accounting for 15% of brain tumors in adults. Malignant GBM tumors display high proliferation and invasion capacity. Studies have shown that the perivascular niche contributes greatly to drug resistance and tumor recurrence, therefore it is important to develop a feasible model to study the interaction between GBM and perivascular cells. Our lab has previously described a model using methacrylamide-functionalized gelatin (GelMA) and demonstrated that a tri-culture system with human umbilical vein endothelial cells (HUVECs), normal human lung fibroblasts (NHLFs), and GBM cells recapitulated the interaction between GBM and perivascular niche. However, NHLFs and HUVECs are not normally present in the human brain. To better recapitulate the microenvironment of human brain, we introduced Brain microvascular endothelial cells (BMVECs), normal human astrocytes (NHA) and Human brain vascular pericytes (HBVP). In addition to the perivascular environment, we also realized that dispersed GBM cells lack the heterogeneity and complex architecture of the mature tumor. In attempt to improve the GBM model, we generated cancer spheroid using the spinner method and combined it with the appropriate perivascular niche. To better profile GBM invasion, perivascular cells and cancer spheroids were introduced into a commercial microfluidic device (AIM Biotech, IdenTx3 chip). While UV exposure has the advantage of fast dynamic crosslinking of GelMA hydrogels, it introduces potential mutagens that could alter the genetic profiles of cells and is difficult to synchronize the amount of exposure across hydrogels with different sizes. Here, we explore a different hydrogel system that relies on enzymatic crosslinking. Thiolated gelatin (Gel-SH) mediated by horseradish peroxidase (HRP) forms hydrogel in the absence of UV light, reducing the variability due to mutation and hydrogel inconsistency.





Materials and Methods: Cell culture: Human umbilical vein endothelial cells (HUVECs) and normal human lung fibroblast (NHLFs) were purchased from Lonza (Walkersville, MD) and used before passage 6. Primary human brain microvascular endothelial cells (BMVECs) was purchased from Cell system and used before passage 4. Human brain vascular pericytes (HBVPs) and normal human astrocytes (NHAs) were purchased from Sciencell and used before passaged 3. HUVECs and BMVECs were culture using Endothelial growth medium 2 and NHLF were cultured using fibroblast growth medium 2 (Lonza, Walkersville, MD), BMVP and NHA were cultured using pericyte growth medium and astrocyte growth medium (Sciencell, Carlsbad, CA)

Hydrogel formation: A mixture of HUVECs and NHLFs (2:1 NHLF: HUVEC, 2 x 10⁶ cells/ml) or a mixture of BMVECs, HBVP, and NHA (3:1:1 BMVECs: HBVP:NHA, 1 x 10⁶ cells/ml) were resuspended in the GelMA solution, then loaded on to the teflon mold or pipetted into the central channel of IdenTx3 chip (AIM Biotech). GelMA was polymerized by the addition of L,AP and exposure to UV light for 45s using a UV lamp (λ = 365 nm, 7.07 mW/cm²). ; Gel-SH was polymerized by the addition of diluted hydrogen peroxide and HRP at room temperature for 60 minutes.

Spheroid formation: Human U-87 MG, 42MGBA GBM cells formed spheroids using the spinner method, briefly, GBM cells were seeded in the Corning® spheroid microplate, incubated at 37°C on a rotator at 60 rpm for 48 hours. Spheroids were initiated with 100, 200, 500, 5000, and 10000 cells and cultured for 48 hours before further experimentation.

Invasion assay: In the invasion assay, a mixture of BMVECs, HBVP, and NHA were re-suspended in the hydrogel, loaded into the central channel, and then cultured for 48 hours before introduction of a GBM spheroid. Spheroids of different sizes were then introduced into the media channel and cultured for 7 days using EGM-2 media, replaced every other day.

Results, Conclusions, and Discussions:: We first investigated whether it is possible to form vasculature networks inside the microfluidic device. HUVECs and NHLFs encapsulated in GelMA and Gel-SH hydrogels formed mature vasculature 48 hours after seeding into the microfluidic device. Culture was maintained for 14 days. The same experiment was repeated using BMVECs, HBVPs, and NHAs. We then generated GBM spheroids of different cell initiating number and identified a logarithmic relationship between spheroid sizes and initiating number. We observed the formation of necrotic cores in spheroids larger than 200 µm, and the size of the necrotic core was dependent on spheroid size. We then monitored spheroid response in co-culture with perivascular cells and found an increase in invasion as measured by number of invading cells and average invading distance. This result coincided with our previous finding that perivascular network increases GBM proliferation and invasion in single GBM cells (Ngo et al, 2020). In addition, we observed a different degree of invasion across GBM spheroids of different size. It has been reported that spheroids with different sizes and shapes displayed enhanced migration and drug resistance; however, the underlying mechanism remain unknown. Our future work aims to investigate the effects of anticancer drug on drug resistance- cancer line, and how it affect the interaction with perivascular cells. We also seek differences in live-death ratio between spheroids of different sizes and investigate the potential changes in the genetic profile.


Acknowledgements (Optional): :

References (Optional): :