Freestanding hydrogel lumens for modeling blood vessels and vasodilation

Abstract

Lumen structures exist throughout the human body, and the vessels of the circulatory system are essential for carrying nutrients and oxygen and regulating inflammation. Vasodilation, the widening of the blood vessel lumen, is important to the immune response as it increases blood flow to a site of inflammation, raises local temperature, and enables optimal immune system function. A common method for studying vasodilation uses excised vessels from animals; major drawbacks include heterogeneity in vessel shape and size, time-consuming procedures, sacrificing animals, and differences between animal and human biology. We have developed a simple, user-friendly in vitro method to form freestanding cell-laden hydrogel rings from collagen and quantitatively measure the effects of vasodilators on ring size. The hydrogel rings are composed of collagen I and can be laden with human vascular smooth muscle cells, a major cellular and structural component of blood vessels, or lined with endothelial cells in the lumen. The methods presented include a 3D printed device (which is amenable to future fabrication by injection molding) and commercially available components (e.g., Teflon tubing or a syringe) to form hydrogel rings between 2.6-4.6 mm outer diameter and 0.79-1.0 mm inner diameter. Here we demonstrate a significant difference in ring area in the presence of a known vasodilator, fasudil (p < 0.0001). Our method is easy to implement and provides a foundation for a medium-throughput solution to generating vessel model structures for future investigations of the fundamental mechanisms of vasodilation (e.g., studying uncharacterized endogenous molecules that may have vasoactivity) and testing vasoactive drugs.

Keywords: Biological assays; Hydrogel; Lumen.

Figure 1.

An arrayable method for fabricating cell-embedded free-standing collagen I lumens. (Ai) Photographs of device setup and basic workflow for Method 1. (Aii) A rubber spacer was added to the 3D printed device to hold up the Teflon tubes on each post. F127 hydrogel seals small gaps so the collagen I does not leak through. (Aiii) The rack was submerged in phosphate buffered saline, and the tubes were pushed down to remove the collagen rings and dissolve F127. (Bi) Fluorescence image of primary human umbilical artery smooth muscle cells embedded in a collagen I ring. Cells are stained with calcein AM (live, green) and ethidium-homodimer I (dead, red). (Bii) Image of an array of 3 mm OD and 1 mm ID hydrogel rings in a 96-well plate.

Figure 2.

Reproducibility of collagen I rings designed with an outer diameter of 3.0 mm and inner diameter of 1.0 mm. (A) Collagen rings of equal size in a 24-well plate. (Bi) The average outer diameter from each device was 2.48±0.17 mm, 2.63±0.12 mm, 2.66±0.15 mm, 2.73±0.08 mm and (Bii) their respective inner diameters were 0.79±0.04 mm, 0.81±0.05 mm, 0.76±0.05 mm, and 0.79±0.05 mm. Each symbol pairs the OD measurements to ID measurements across Bi and Bii. Results are plotted from two independent experiments, each with two racks (one rack is an array of 6 lumens). Error bars are mean ± SD. Full set of data available in Table S1.

Figure 3.

Addition of endothelial cells to the lumen of collagen I rings. (A) Workflow for adding the endothelial cell suspension to the gelled collagen rings after adapting Method 1. (B) Image of the lumen was taken after 24 hours of cell seeding to show the viability of the endothelial cells. Cells are stained with calcein AM (live, green) and ethidium- homodimer I (dead, red). Scale bar is 100 μm.

Figure 4.

Vasoactivity data of human umbilical artery smooth muscle cells seeded in hydrogel rings when a vasodilator (fasudil) was added. (A) Method 2 workflow using a commercially available syringe with a 3D printed insertable core that has been adapted to make cell laden collagen rings. (Bi) The hydrogel rings were recorded for 20 minutes after addition of fasudil, and their percent change in area was calculated using ImageJ. (Bii) Percent change in ring area data for hydrogel rings treated with buffer (control) or vasodilator (fasudil). Data points are from 8 rings across 2 independent experiments; error bars are mean ± SD. A two-sample unpaired t-test (two-tailed) was used. ****p<0.0001.