Page 357 - IJB-10-3

International Journal of Bioprinting hNVU chip for brain modeling and drug screening

3.2.1. Bioink biometric ECM of NVU cells in brain by (NG2), glioma cells (GFAP), microglias (IBA1), and
modulating 3D bioprinting parameters. neurons (β3-tubulin) depicted cell spreading and even
The bioprinting process faces a challenge arising from distribution in the brain region (Figure 4D). These results
the contradiction between the high modulus required for indicated that the modulus of the hydrogel was appropriate
3D printing of hydrogels and the low modulus of brain for the five-cell co-culture and iPSC-NPCs differentiation.
tissue, which typically ranges from 100 to 3000 Pa. 33,34 Meanwhile, the upregulation of the gene expression of
Collagen-hydrolyzed gelatin and its derivative, GelMA, the neuron marker β3-tubulin and nestin indicated that
provide cell adhesion sites and have long been used for iPSC-NPCs not only differentiated into neurons (Figure
constructing tumor angiogenesis models. In this study, S9 in Supplementary File) but also potentially proliferated
30
we utilized temperature-responsive gelatin and low- (Figure 4E). These characteristics mimic the pathological
concentration GelMA that can be dual-controlled by light features observed in the brain, which involve the presence
and temperature to create a 3D printing-friendly and low- of neural stem cells.
modulus ECM mechanical environment. Precise control
of parameters such as light exposure and temperature 3.2.3. The cells in the brain regions of the hNVU chip
during printing allowed us to achieve optimal results. remodel the ECM
Rheological analysis demonstrated that a mixture of 5% The assembly of microstructures within the structure
GelMA and 5% gelatin achieved a modulus ranging from is achieved through the remodeling of the ECM and the
37-39
100 to 1000 Pa within the printing window of 20°C–29°C cell–ECM interactions. Immunofluorescence staining
(Figure 4A). To recapitulate the pathological ECM showed high protein expression of COL4A1 and laminin,
microenvironment, we incorporated fibrinogen into the which are the main constituent ECM proteins in the brain
bioink cocktail, which is known to be increased in the (Figure 4F). Furthermore, we conducted an investigation
glioma area and associated with perivascular growth and into the gene expression associated with ECM and cell–
glioma infiltration. To provide adhesion sites for NPCs, ECM interactions to gain a deeper understanding of brain
36
35
we also added laminin to the ink. Finally, GelMA (5% region models (Figure 4G). Analysis of gene expression
+
GelMA + 2.5 mg/mL fibrinogen + 5% gelatin + 20 μg/ patterns revealed an upregulation of key ECM protein
mL laminin + 3 U/mL transglutaminase) was used as the components, namely fibronectin (FN1), collagen IV, and
primary hydrogel material. After extrusion-based printing, laminin, indicating cellular remodeling and a tendency
photocrosslinking was employed to confer long-term toward biomimicry in the models. Additionally, the genes
stability to the printed structures. ND1 (mitochondrially encoded NADH dehydrogenase
1) and SPARC (secreted protein acidic and cysteine rich),
3.2.2. Brain regions with the biomimetic cellular known for their involvement in cell–ECM interactions,
composition of neurovascular unit cells in brain exhibited increased expression levels. ND1, widely
To better replicate microstructural assembly and the distributed in basement membranes, tightly interacts with
high cell density in the brain region, our study focused laminin, collagen IV, and perlecan. 40,41
on investigating the different ratios and cell densities of Taken together, these findings indicated that the brain
five distinct cell types, namely ECs, pericytes, astrocytes, region structure can replicate the cellular composition and
microglias, and iPSC-NPCs. The findings revealed that ECM of brain tumors.
a cell ratio of 12:4:4:4:1 and a total cell density of 2.5 ×
10 /mL demonstrated optimal spreading of endothelial 3.3. The hNVU chip exhibits pathological features of
7
cells (Figure S7 in Supplementary File) as well as the NVU cells in brain
highest cell survival rate (Figure 4B and C). A cell ratio To assess the pathological characteristics of glioma in the
of 12:4:4:4:1 and a total cell density of 2.5 × 10 /mL were hNVU model, we evaluated the dynamic changes in the
7
used in subsequent experiments. A schematic diagram gene expression levels of typical markers and revealed
of the constructed structure with a hierarchical cell the biological processes occurring within the cells in the
distribution is shown in Figure S8A (Supplementary File), brain region, which are related to angiogenesis and glioma
and the experimental picture is shown in Figure 4B (red: mass transport.
mixture of pericytes and astrocytes, layer 2 and layer 4 in
Figure S8A in Supplementary File; green with tube: “brain The gene expression levels of the endothelial cell
region”, layer 3 in Figure S8A in Supplementary File; green marker CD31, tight junction protein ZO-1, and pericyte
(up layer and down layer): endothelial cells, layer 1 and markers SM22A and aSMA were upregulated (Figure 5A),
layer 5 in Figure S8A in Supplementary File). indicating the trend of angiogenesis within the model.
The genes expression patterns of the glial marker GFAP
After 7 days of culture, immunofluorescence (IF) and S100B suggested a strong association between
staining of identity markers for ECs (CD31), pericytes angiogenesis and glioma cells. Furthermore, the

Volume 10 Issue 3 (2024) 349 doi: 10.36922/ijb.1684