A perfused multi-well bioreactor platform to assess tumor organoid response to a chemotherapeutic gradient

Abstract

There is an urgent need to develop new therapies for colorectal cancer that has metastasized to the liver and, more fundamentally, to develop improved preclinical platforms of colorectal cancer liver metastases (CRCLM) to screen therapies for efficacy. To this end, we developed a multi-well perfusable bioreactor capable of monitoring CRCLM patient-derived organoid response to a chemotherapeutic gradient. CRCLM patient-derived organoids were cultured in the multi-well bioreactor for 7 days and the subsequently established gradient in 5-fluorouracil (5-FU) concentration resulted in a lower IC₅₀ in the region near the perfusion channel versus the region far from the channel. We compared behaviour of organoids in this platform to two commonly used PDO culture models: organoids in media and organoids in a static (no perfusion) hydrogel. The bioreactor IC₅₀ values were significantly higher than IC₅₀ values for organoids cultured in media whereas only the IC₅₀ for organoids far from the channel were significantly different than organoids cultured in the static hydrogel condition. Using finite element simulations, we showed that the total dose delivered, calculated using area under the curve (AUC) was similar between platforms, however normalized viability was lower for the organoid in media condition than in the static gel and bioreactor. Our results highlight the utility of our multi-well bioreactor for studying organoid response to chemical gradients and demonstrate that comparing drug response across these different platforms is nontrivial.

Keywords: bioreactor 3D cell culture; colorectal (colon) cancer; drug transport; flow transport; tumor model.

FIGURE 1

Chemotherapeutic efficacy was compared across three different platforms. (A) Organoids in media, (B) organoids in hydrogel, (C) organoids in perfused multi-well bioreactor platform.

FIGURE 2

Perfused multi-well bioreactor platform. (A) Exploded view showing molded plug gasket, maltodextrin sacrificial channels, and custom machined aluminum lid and base to seal the commercially available 12 well plate. (B) A peristaltic pump was used to recirculate organoid media from a reservoir to each well of the bioreactor at a flow rate of 20 μL/min.

FIGURE 3

Transport of 5-Fluorouracil varies between platforms. (A) Transport of 5-Fluorouracil for organoid in gel condition (no flow) results in a homogenous concentration after 24 h. (B) An isometric view cut along length of channel shows that transport of 5-Fluoruracil in the multi-well bioreactor results in a gradient in drug concentration even after 7 days. (C) Top-down view taken at channel center. (D) Region 1 (R1), close to the bioreactor channel, has a higher average concentration over time compared to region 2 (R2) which is far from the channel.

FIGURE 4

Organoid morphology and size does not change with drug concentration. (A) Experimental timeline and process. (B) Organoids at the lowest and highest drug concentrations in all three platforms for day 7. Organoid area (C) and circularity (D) for all platforms.

FIGURE 5

IC₅₀ curves for the (A) bioreactor, (B) static gel, and (C) organoid in media conditions determined using the CellTiter-Glo assay. (D) IC₅₀ values for all culture conditions.

FIGURE 6

Temporal organoid response to 5-FU can be determined in a nondestructive manner using the LactateGlo assay. Lactate secretion from organoids is dose dependent over time in the (A) bioreactor, (B) static gel, and (C) media conditions. (D) EC₅₀ values for all culture conditions.

FIGURE 7

(A) Total 5-FU delivered (AUC) by dose and culture method calculated using average concentration over time from the simulations. (B) IC₅₀ curves for all culture conditions normalized for total 5-FU delivered (AUC).