Biomimetic on-chip assay reveals the anti-metastatic potential of a novel thienopyrimidine compound in triple-negative breast cancer cell lines

Abstract

Introduction: This study presents a microfluidic tumor microenvironment (TME) model for evaluating the anti-metastatic efficacy of a novel thienopyrimidines analog with anti-cancer properties utilizing an existing commercial platform. The microfluidic device consists of a tissue compartment flanked by vascular channels, allowing for the co-culture of multiple cell types and providing a wide range of culturing conditions in one device. Methods: Human metastatic, drug-resistant triple-negative breast cancer (TNBC) cells (SUM159PTX) and primary human umbilical vein endothelial cells (HUVEC) were used to model the TME. A dynamic perfusion scheme was employed to facilitate EC physiological function and lumen formation. Results: The measured permeability of the EC barrier was comparable to observed microvessels permeability in vivo. The TNBC cells formed a 3D tumor, and co-culture with HUVEC negatively impacted EC barrier integrity. The microfluidic TME was then used to model the intravenous route of drug delivery. Paclitaxel (PTX) and a novel non-apoptotic agent TPH104c were introduced via the vascular channels and successfully reached the TNBC tumor, resulting in both time and concentration-dependent tumor growth inhibition. PTX treatment significantly reduced EC barrier integrity, highlighting the adverse effects of PTX on vascular ECs. TPH104c preserved EC barrier integrity and prevented TNBC intravasation. Discussion: In conclusion, this study demonstrates the potential of microfluidics for studying complex biological processes in a controlled environment and evaluating the efficacy and toxicity of chemotherapeutic agents in more physiologically relevant conditions. This model can be a valuable tool for screening potential anticancer drugs and developing personalized cancer treatment strategies.

Keywords: endothelial cell; intravasation; microfluidics; permeability; tumor microenvironment.

FIGURE 1

The (A) microfluidic device is fabricated in transparent PDMS and bonded to a glass slide. The device comprises (B) two semicircle vascular channels (basolateral side) surrounding a tissue compartment (apical side), all of which are individually perfused via inlet/outlet ports connected with Tygon tubing. The vascular channels and the tissue compartment are separated by a (C) porous interface comprising of microfabricated porous structure, which facilitates tumor-endothelial co-culture and information/mass exchange.

FIGURE 2

Flowchart shows the overall experimental setup for creating the HUVEC-TNBC co-culture model under flow perfusion and the timeline for using the biomimetic 3D tumor model to test the anti-TNBC efficacy of chemotherapeutics, TPH104c, and PTX using various biological assays.

FIGURE 3

Brightfield images (BF) showing the progression for creating co-culture in the microfluidic device. (A) HUVECs, immediately after seeding, (B) HUVECs firmly attached to the fibronectin-coated vascular channel surface 4 h after initial seeding, no flow during this stage. (C) HUVECs aligned in the direction of flow in the vascular channels after 24 h of perfusion (0–1.75 dyne/cm2). (D) SUM159PTX cells, resuspended in 5% Matrigel, were seeded into the tissue compartment 24 h after HUVEC seeding. (E) SUM159PTX cells were allowed to attach and expand; 24 h after seeding. (F) The HUVEC-TNBC co-culture was fully established after 48 h after SUM159PTX seeding. Continuous media perfusion was maintained in the vascular channels, whereas bolus media injection was given every 8 h to the tissue compartment. Scale bar = 400 µm.

FIGURE 4

Confocal Images of the established tumor microenvironment on-chip. (A) Top view, (B) Side view (90°), (C,D) Angled view (40° and 240°). HUVECs formed a confluent monolayer in the vascular channels, whereas TNBC formed in the 3D tissue compartment. Red: TNBC cells (CM-DiI), Blue: Cell nucleus (Hoechst 33342), and Green: adherens junction between HUVECs (VE-Cadherin).

FIGURE 5

After 72 h of drug treatment, TPH104c exhibited superior selectivity to TNBC cells than ECs compared to PTX in the microfluidic TME, as indicated by (A) BF and fluorescence imaging after applying CellTiter-Blue cell viability assay (B,C) on HUVECs and TNBC cells. Data are presented as mean ± SD (n = 3). * p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, by ANOVA. Scale bar = 200 µm.

FIGURE 6

Drug treatments successfully reached the tumor (central tissue compartment) from their initial delivery site (the vascular channels). Images were taken at various time points during the 72-h drug treatment period, and the cell numbers were analyzed using the “Count and Measure” feature in cellSens Dimension software using the CM-DiI fluorescence signal. (A) Representative fluorescence images showing HUVECs (green), TNBC cells (red), and combined images with various concentrations of drug treatments. (B) Quantification of the number of surviving cells suggests concentration as well as time-dependent cancer cell killing for both treatments. (C) No significant efficacy difference was detected between the two treatments. Data are presented as mean ± SD (n = 3). Red: TNBC cells (CM-DiI) and Green: HUVECs (CMFDA). Scale bar = 400 µm.

FIGURE 7

Endothelial barrier integrity was better preserved after the TPH104c treatment than PTX, as indicated by the permeability assay. The permeability of 4 kDa fluorescent dextran from the vascular channels to the tissue compartment was determined by measuring the fluorescent intensity of dextran in the tissue compartment relative to the vascular channel as the dextran diffuses from the vascular channel to the tissue compartment. (A) Representative fluorescence images of 10 $\mu$ M treatment groups compared to empty device, EC only control, and co-culture control at various time points. Increased permeability after (B) TPH104c or (C) PTX treatments suggest concentration-dependent damage to EC barrier integrity. Notably, (D) this increase was much more prominent in PTX treated than TPH104c treated groups. Data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by ANOVA. Scale bar = 400 µm.

FIGURE 8

TPH104c significantly reduced the intravasation of TNBC cells from the primary tumor into the vasculature compared to PTX. Representative images of TNBC intravasation were shown in Figure 6A. (A) Quantification of the number of intravasated TNBC cells reveals that both treatments exhibit a concentration and time-dependent inhibitory effect. (B) By first normalizing the number of intravasated TNBC cells to the total number of TNBC cells in all the compartments, and then compare with control, we derive the “intravasation percentage” as a metric signifying TNBC intravasation. This index distinctly highlights TPH104c’s superior inhibitory effect in contrast to PTX. Interestingly, the intravasation percentage of TNBC cells increased after PTX treatment at 0.1 and 1 µM. Data are presented as mean ± SD (n = 3). *p < 0.05, **p < 0.01, by ANOVA.