(K-437) Evaluating the Impact of Collagen Stiffness on Lymphatic Endothelial Cell Behavior

Introduction:: Lymphatic vasculature is spread throughout the body, maintaining extracellular matrix fluid homeostasis and inflammatory response. Lymphatic endothelial cells (LEC) make up lymphatic capillaries that are directly attached to their surrounding tissue. LECs undergo morphological changes, including proliferation and migration, under pathological diseases and dynamic processes such as fibrosis that create a stiff tissue environment. LECs sense and respond to external stimuli such as fluid flow and tissue stiffness, but there is a significant gap in understanding how changing tissue stiffness may alter lymphatic growth and function separately. There are also specific questions on how Vascular Endothelial Growth Factor Receptor 2 (VEGFR2) expression may play a role in function more than growth. Existing disease models evaluating the effect of tissue stiffness on cells typically only include a stiff and a soft environment and exclude temporal stiffening that is more reflective of fibrosis progression. To address these gaps, our lab has created a collagen-based in vitro model with methacrylated collagen to generate three stiffnesses (soft, temporal, and stiff) that mimic tissue stiffening during fibrosis and disease progression. This project investigated the impact of stiffness on VEGFR2 expression and cell-cell junction formation as an indicator of lymphatic function. A portion of this project also focused on determining an appropriate method for assessing cell coverage and cell-cell junction formation to evaluate barrier integrity morphologically.

Materials and Methods:: Methacrylated type 1 collagen gels, ColMA (8mg/mL) (Advanced BioMatrix) with ruthenium and sodium persulfate photoinitator were made according to the manufacturer’s protocol, protected from light in a 96-well plate, and kept in incubation with 1X phosphate buffer saline (PBS) solution until next steps. The 3D ColMA gels were then coated with fibronectin to promote cell attachment before human dermal lymphatic endothelial cells (LEC) were seeded on the gel surface at 3,000 cells/cm2. Photo-crosslinking with 405nm blue light exposure for 90 seconds was used to generate pathological stiffness conditions: static crosslinked on Day 0 prior to cell seeding (~4kPa) and temporal crosslinked on Day 6 (after cell seeding on Day 0). An un-crosslinked gel was used as a soft stiffness level to represent normal tissue stiffness (~0.5kPa). Cells were fixed and stained with VE Cadherin (cell-cell junction) or Phalloidin (F-actin) antibody with DAPI nuclear counterstain, immediately imaged with 20X magnification on a Keyence fluorescence microscope, and analyzed using ImageJ and CellProfiler to calculate the LEC coverage ratio and determine a more definitive measure of cell junction formation. Separate samples were made to evaluate changes in VEGFR2 expression using western blotting techniques (GAPDH loading control) with immunofluorescence. Samples were imaged using Azure600 fluorescent labels.

Results, Conclusions, and Discussions::
VE-Cadherin-stained images showed more definitive cell-cell junction formations and cell membrane, while phalloidin-stained samples displayed the actin cytoskeleton clearly (Figure 1). After running cell coverage analysis more than once to increase object size and detect as much of the cell body as possible, VE-Cadherin images showed a 39.4% cell area to total area coverage that was more representative of what we observed visually. Phalloidin-stained samples showed lower coverage at 27.6%. The VE-Cadherin stain emerged as a better tool to assess cell coverage ratio than phalloidin due to less defined cell boundaries in phalloidin stained samples. Figure 2 shows DAPI staining of LEC distribution. The soft ColMA gel appeared to support LEC growth with more even distribution, while LEC distribution on stiffer ColMA was not uniform. We tried to visualize VE-Cadherin staining to evaluate cell coverage on 3D ColMA, but there were difficulties in detecting positive staining during imaging. In Figure 3, we observed prominent bands at 36kDa for GAPDH loading control and faint bands around 230kDa (soft and temporal samples) and 115kDa (stiff samples), suggesting low-level expression of VEGFR2. There may be variable expressions with stiffness, but dark spots on the blot made resolving bands difficult. The 230 kDa band is expected for mature VEGFR2, but the lower molecular weight band may be a glycosylated form of VEGFR2 or non-specific binding.

Overall, results show that LECs are responsive to changes in ColMA stiffness. Although we successfully identified a cell coverage approach using VE-Cadherin on LECs seeded on fibronectin-coated wells, we were unable to clearly visualize VE-Cadherin in LECs on 3D ColMA to compare cell-cell junctions between ColMA groups. Western blot results are promising, but protein concentration may have been too low, and we need a positive VEGFR2 control. Additional future steps are to troubleshoot staining and imaging for both VE-Cadherin and western blots and quantify total and phosphorylated VEGFR2 expression.

We can still use these findings to build on lymphatic and disease models. Having a clearer indication of how LEC functions may be affected in fibrotic disease models can help improve treatments that may include VEGFR2 as a new target.



Acknowledgements (Optional): :
The authors would like to thank Dr.Jeannine Coburn, Athenia Jones, and Stephen Larson for their generous insight and support in various aspects of this study. We would like to acknowledge the National Science Foundation Engineering Research Initiation award (#2138841) and National Science Foundation REU (EEC 21 50076) for funding. Images were created using BioRender.com.



References (Optional): :