Page 128 - IJB-9-4

International Journal of Bioprinting 3D bioprinted models in pediatric tumors

produced viable tumors. The printed bioink without cancer 3.3. Bioprinted tumors are resistant to hypoxia
cells was plated in media withCalcein AM,and images Solid tumors, such as neuroblastoma, grow under hypoxic
were taken without(Supplementary File, Figure S2A) and conditions in vivo [17,18] . We aimed to compare the ability
withfluorescence (Supplementary File, Figure S2B) using of 3D-bioprinted tumors to 2D-cultured cells to survive
a FITC laser. No fluorescence was detected under either hypoxic conditions. SK-N-AS cells were printed as layered
condition, indicating that the fluorescence detected in the bioprinted tumors (Figure 1A, left panel) and compared
3D-bioprinted tumors represented live tumor cells, not to 2D-cultured SK-N-AS cells. The cells and bioprinted
autofluorescence of the bioink. models were exposed to 1% O for 5 days. In order to count
2
the number of alive versus dead cells due to hypoxia,
In order to determine if the bioprinted models were fluorescence staining was used to evaluate viability of the
representative of the tumors from the patients or those bioprints, while trypan blue stain was used to assess the
propagated in animals, IHC studies were performed. viability of 2D-cultured cells. The percentage of viable
H&E staining of the bioprints demonstrated a similar cells in the bioprinted tumors grown in hypoxia was
morphology with that observed in the original tumor significantly higher than that observed in cells grown in
and PDX from mouse (Figure 1C and E) for both hypoxic 2D culture. The bioprinted tumors had an average
COA6 and COA109. Specifically, the COA6 bioprint of 69 ± 9% viability compared to cells in the 2D culture,
tumor retained features of small round blue cells, while with 33 ± 7% viability (Figure 3A). A representative image
the COA109 bioprint demonstrated “salt and pepper” of a bioprinted tumor shows areas of tumor cell death
appearance (Figure 1E, right panel), which are typical (Figure 3B, red fluorescence) surrounded by alive tumor
of neuroendocrine tumors . Next, we performed cells (Figure 3B, green fluorescence).
[15]
immunostaining for commonly utilized tumor markers,
including NSE for neuroblastoma and chromogranin A 3.4. Mixed bioprinted models for ex vivo studies
for neuroendocrine tumors. The human neuroblastoma PDXs can better recapitulate human conditions than
COA6 bioprint stained positive for NSE (Figure 1D, established, long-term passaged cell lines; however,
right panel), while the human neuroendocrine COA109 it remains a challenge to perform in vivo studies with
bioprint stained positive for chromogranin A (Figure 1F, PDXs due to their slow and inconsistent growth rates .
[19]
right panel), indicating that the bioprints retained NET We hypothesized that bioprinted PDX models could
and neuroblastoma protein markers. be employed to test potential therapeutics in an ex vivo
fashion. Both COA6 and COA109 bioprinted tumors
3.2. In vivo growth of bioprinted models were produced using the mixed bioprinting method
In order to evaluate the tumorigenicity of the bioprints, (Figure 1A, left panel) and allowed to grow for 5 days.
we used the mixed bioprint method and a murine flank Chemotherapeutic agents including cisplatin (10 μM),
model. We first printed SK-N-AS cells with the mixed which is commonly employed for neuroblastoma, and
model (Figure 2A). After growing in culture for 5 days, trametinib (100 nM), a MEK1/2 inhibitor used to treat
the SK-N-AS bioprinted microtumor was minced and NETs, were added to the media of COA6 or COA109
implanted into the flank of a nude mouse (Figure 2A). The bioprints, respectively. After 10 days, Calcein AM was
growth was consistent with other SK-N-AS orthotopic used to detect viable cells. Viability was quantitated by
tumors grown in mice . Once the implanted microtumor calculating the mean integrated density (MID) of the
[16]
reached 2,000 mm , it was harvested for IHC studies. H&E fluorescence of each bioprint, as these mixed bioprinted
3
staining showed that the tumor growing in the mouse, tumors were larger than the layered bioprinted tumors and
from the bioprinted microtumor, retained SK-N-AS required the inclusion of intensity in determining viability
morphology (Figure 2B, middle panel) as compared to (Figure 4A). Compared to those treated with vehicle,
SK-N-AS flank tumors(Figure 2B, left panel) and stained COA6 bioprinted tumors treated with cisplatin showed
positive for NSE (Figure 2B, right panel), indicating that a significant decrease in MID (1713 ± 268 pixels/mm
2
the bioprint tumor was a neuroblastoma. We repeated a versus 4878 ± 306 pixels/mm , cisplatin versus vehicle,
2
similar experiment using the COA109 PDX (Figure 2C). P ≤ 0.0001, Figure 4A and B). Compared to bioprinted
There were no notable differences on H&E between the tumors treated with vehicle, COA109 bioprinted tumors
COA109 propagated through a mouse (Figure 1E, middle treated with trametinib had significantly less MID,
panel) or the COA109 bioprinted microtumor grown in indicating fewer viable cells (1442 ± 172 pixels/mm 2
the animal (Figure 2D, left panel). The neuroendocrine versus 2336 ± 120 pixels/mm , trametinib versus vehicle,
2
features were retained in the printed microtumor P ≤ 0.001, Figure 4D and E). These results demonstrate
as demonstrated by positive chromogranin staining that 3D-bioprinted models may be used to test PDX cells
(Figure 2D, right panel). ex vivo.

Volume 9 Issue 4 (2023) 120 https://doi.org/10.18063/ijb.723