Developing a Biofabricated Vascular Model to Study Endothelial Cell Roles in Breast Cancer Progression Following Cardiovascular Disease

Introduction: : Cardiovascular disease (CVD) and cancer are the leading causes of death in the U.S. and are linked in many ways [1]. Heart failure was found to increase the risk of breast cancer recurrence and progression by ~60% in a retrospective clinical trial following breast cancer patients in remission [2]. Mouse studies have shown that CVD releases circulating factors into the bloodstream which creates a pro-tumorigenic environment at distal cancer sites [2-4]. However, it is unknown what role tumor microenvironment cells, such as vascular endothelial cells (ECs), play in creating a pro-tumorigenic environment after interacting with CVD circulating factors. ECs line the curved surface of blood vessels and regulate inflammation in response to soluble factors in the bloodstream, including C3a, a CVD circulating factor shown to have a pro-inflammatory effect on ECs. The magnitude of endothelial response to circulating factors is dependent on dimensional and mechanical factors in the environment such as extracellular matrix stiffness and tissue-level curvature. To facilitate studying pro-tumorigenic responses from ECs exposed to CVD circulating factors in a biomimetic environment, our study objective is to develop a 3D microfluidic endothelialized vessel model in an open chip design that allows controlled placement of adjacent breast epithelial spheroids in the surrounding matrix. This model will be utilized to determine potential pro-inflammatory responses of ECs to CVD circulating factors and the resulting modulation of breast cancer progression. This model will allow us to increase our understanding of the role of ECs in spurring breast cancer progression after a cardiovascular event.

Materials and

Methods: : To create the biofabricated vessel, a base PDMS compartment was adhered to a glass slide. A needle was inserted into the compartment with a 10% Gelatin methacrylate (GelMA) hydrogel solution. After crosslinking, the needle was withdrawn to leave a hollow channel. Human umbilical vein endothelial cells (HUVECs) were seeded on the fabricated channel by adding a HUVEC solution and rotating three times every 30 minutes to ensure cell coverage. Cell coverage was confirmed 48 hours post-seeding using GFP-HUVECs and viability was assessed by staining with calcein-AM (live) and ethidium homodimer-1 (dead). HUVEC phenotype was characterized via VE-cadherin, actin (phalloidin), and Hoechst staining.
To precisely place spheroids in the chip, aspiration-assisted bioprinting (Sartorius CellCelector) was used to position 200 µm diameter MCF10A breast epithelial spheroids adjacent to the needle template in a step-wise fashion in the uncrosslinked GelMA prior to channel endothelialization. After crosslinking, spheroid placement was confirmed by staining for Hoechst, imaging, and measuring the distance from the edge of the channel to the spheroid edges. Spheroids were stained with Ki-67 (proliferative marker), E-cadherin (adherens junctions), and Hoechst using a lab-grade microwave to visualize phenotypic markers. All imaging of stained vessels and spheroids was performed using a confocal laser scanning microscope.
As an initial step to validate the pro-inflammatory effect of C3a, HUVECs were treated with 500 nM C3a for 18 hours and an IL-8 ELISA was performed on the cell culture media.

Results, Conclusions, and Discussions:: Using the needle templating technique, a hollow channel was successfully formed in the 10% GelMA hydrogel on the fabricated PDMS chip (Figure 1A-B). The fabricated channels were endothelialized and visualized using GFP-HUVECs, showing complete cell coverage (Figure 1C). The vessels displayed high HUVEC viability 48 hours post-seeding with minimal appearance of dead cells as visualized by ethidium homodimer-1. The open chip design facilitated on-chip immunostaining and high-resolution images, and staining of endothelialized channels showed actin cytoskeleton filaments and expression of endothelial marker VE-cadherin at cell-cell junctions, along with complete lumen formation (Figure 1D). C3a dosing of HUVECs was shown to significantly increase IL-8 secretion from HUVECs, validating a pro-inflammatory response (Figure 1E). Aspiration-assisted bioprinting enabled placement of breast epithelial spheroids adjacent to endothelialized channels with high cell viability (Figure 1F). Placement of spheroids was found to be accurate between distances of 100 – 500 µm from the channel edge (Figure 1G). MCF10A spheroids were successfully immunostained in the GelMA hydrogel showing presence of proliferation marker Ki-67 and cell-cell junction marker E-cadherin (Figure 1H).

A 3D vascular model was successfully manufactured in an open chip design that allowed accurate placement of breast epithelial spheroids adjacent to the vessel. The biofabricated vessel showed high cell viability, complete endothelial lumen formation, and presence of endothelial phenotypic markers. Breast epithelial spheroids were precisely placed adjacent to the vessel using aspiration-assisted bioprinting at distances between 100 – 500 μm. Further, HUVECs treated with CVD circulating factor C3a showed a statistically significant increase in inflammatory IL-8 secretion following C3a treatment. Controlling the distance from spheroid to vessel will allow us to place the spheroids within range of soluble signaling to interrogate the crosstalk between spheroid and vessel within a 3D matrix. Ongoing work includes treating the endothelialized vessel with CVD proteins to determine potential changes in cytokine secretion in the 3D model, as well as the effect on endothelial barrier permeability, and breast cancer cell response to a pro-inflammatory endothelium. This model system will facilitate future understanding of the endothelial contribution to enhanced tumorigenesis after interaction with CVD circulating factors in a physiologically-relevant 3D environment.

Acknowledgements and/or References (Optional):: 1. Xu et al. NCBI Data Brief. 2022;(456).
2. Koelwyn et al. JMCC. 2022;163:1-8.
3. Meijers et al. Circ. 2018;138(7):678-691.
4. Avraham et al. Circ. 2020;142:670–683