Unraveling Tumor Blood Vessel PD-L1 Regulation via Trans-Endothelial Flow in a Novel Microvasculature-on-a-Chip Model

Introduction: Programmed cell death-1 (PD-1)/programmed death-ligand 1 (PD-L1) immune-suppressive signaling is often hijacked by cancer cells to evade immune surveillance in the tumor microenvironment. Endothelial PD-L1 expression is higher in the tumors than in normal tissues and is upregulated in response to biochemical cues such as interferon-gamma (IFNγ) and tumor necrosis factor-alpha (TNFα). However, the role of biophysical factors in regulating endothelial PD-L1 expression is not well understood.
Tumoral vasculature is often leakier than normal vessels, leading to higher levels of trans-endothelial flow (also known as transmural flow) across the endothelial cell monolayer. Yet, the impact of this elevated transmural flow on PD-L1 expression in endothelial cells remains elusive.
Microvasculature-on-a-chip systems have emerged as valuable tools for various investigations, including vasculogenesis, metastasis, and immune cell trafficking. In this study, we developed a new microfluidic device to investigate the relationship between transmural flow and PD-L1 expression in microvascular networks (MVNs).

Materials and

Methods: The key feature of this microfluidic device is the two parallel gel channels running the length of the device that enabled the left gel channel to be filled with a cell-free fibrin gel and the right gel channel to be filled with endothelial cells and stromal cells in a solution of fibrinogen and thrombin which spontaneously self-assembled into MVNs. The MVNs formed openings to the two medium channels on the right side. Addition of medium to the top right medium channel creates a hydrostatic pressure difference driving luminal flow to the bottom right medium channel through the MVN. Openings in the MVN on the left side are blocked at the cell-free fibrin gel. Conversely, adding equal amounts of medium to both right-side media channels eliminate the driving force for luminal flow leaving the MVN under the desired hydrostatic pressure to drive transmural flow.
MVNs were established using immortalized human umbilical vein endothelial cells (HUVECs) paired with normal human lung fibroblasts, or with human dermal microvascular endothelial cells and fibroblasts. Quantification of PD-L1 expression was conducted using immunostaining and reverse transcription-polymerase chain reaction (RT-PCR) techniques.

Results, Conclusions, and Discussions: Using this new system, we found that treating MVNs with physiological levels of transmural flow (with a speed of ~ 3 µm/s) for 24 hours upregulated PD-L1 expression in endothelial cells, but not in stromal cells. The elevated endothelial PD-L1 in the vasculature suppressed CD8 T cell activation, including extravasation and IFNγ secretion. Furthermore, we demonstrated that transmural flow also increased PD-L1 expression even in vessels formed in the tumor microenvironment. Mechanistically, through the use of blocking antibodies and CRISPR knock-out assays, we identified integrin αVβ3 as a potential mediator of transmural flow-driven PD-L1 upregulation in the vasculature. Without functional integrin αVβ3, transmural flow failed to upregulate PD-L1 expression in endothelial cells.

In summary, utilizing this innovative microvasculature-on-a-chip platform, we elucidated that trans-endothelial flow enhances endothelial PD-L1 expression through integrin αVβ3. Our findings provide a new biophysical mechanism contributing to the high levels of PD-L1 expression in tumoral vasculature and suggest that targeting this pathway may be a promising approach for enhancing the immune system's ability to fight cancer.

Acknowledgements (Optional):