Objectives
The microcirculation system plays a crucial role in cancer progression. Due to its multi-cellular, structural, and dynamic complexity, microfluidic models have been exploited to study microvascular infiltration, intra/extravasation, and cellular interactions. However, most existing models focus primarily on blood vessels and use endothelial cell-coated channels, which do not fully replicate physiological conditions. To address this, we present a novel human microcirculation-on-a-chip model featuring 3D self-organized blood and lymphatic microvasculature alongside tumor spheroids. This platform enables the study of interactions between multi-cellular tumors and both microvascular networks. Using lung cancer as a case study, we explored how tumor-derived mediators and cellular interactions influenced vessel reorganization and invasion capacity, identifying the key molecular factors involved.
Methods
A 5-channel microfluidic chip was fabricated in PDMS via UV-photolithography and soft-lithography. Human cell lines, including pulmonary blood and lymphatic microvascular cells, and non-lymphangiogenic (H2073) and lymphangiogenic (Calu-1) lung cancer cells, were compartmentalized within the channels to mimic native cellular arrangement. Multi-cellular cancer spheroids co-cultured with MRC5 lung fibroblasts were embedded into Coll I hydrogel and injected into the central channel, while microvascular cells in ECM hydrogel and culture media were added to adjacent channels to recreate 3D microvascular networks and physiological flows. Immunostaining (Podoplanin, CD31, VE-Cad, F-actin, DAPI) visualized the expression of key biomarkers, while Luminex quantified secreted pro-angiogenic, lymphangiogenic, and inflammatory factors. Biophysical parameters (vessel density, sprouting, segment length, diameter) were analyzed to assess microvascular integrity.
Results
Our microcirculation-on-a-chip model successfully recapitulated key features of native lung tissue, rapidly forming 3D networks and biomarker expression of pulmonary (PMV) and lymphatic (LMV) microvessels. Following tumor injection, PMV and LMV exhibited distinct structural alterations depending on the lymphangiogenic potential of the tumor. Specifically, lymphangiogenic Calu-1 spheroids promoted greater lymphatic infiltration, tumor interaction, and microvascular growth than non-lymphangiogenic H2073 cells. Biophysical analyses showed that LMV exhibited larger diameters, lengths, densities, and sprouting than PMV, and that the magnitude of these parameters correlated with tumor lymphangiogenic potential. These structural differences suggested that tumor-specific secreted factors drive microvascular
remodeling. Cytokine analysis supported this, showing upregulation of G-CSF, IL-8, IL-6, and HGF, specifically in lymphangiogenic Calu-1 tumors. These findings indicate distinct molecular signatures associated with tumor invasiveness and lymphangiogenesis, with inflammatory mechanisms facilitating vascular remodeling in more aggressive lung cancers. A cytokine heatmap further linked specific cytokines to tumor invasiveness and vascular remodeling, pointing to potential therapeutic targets.
Conclusions
Our microcirculation-on-a-chip provides a robust tool for investigating and modeling critical events of lung cancer neo-vascularization, for deciphering unknown fundamental mechanisms of cancer cell invasion into the microvasculature, and for future drug screening applications.
| 
