Tuning three-dimensional collagen matrix stiffness independently of collagen concentration modulates endothelial cell behavior

Abstract

Numerous studies have described the effects of matrix stiffening on cell behavior using two-dimensional synthetic surfaces; however, less is known about the effects of matrix stiffening on cells embedded in three-dimensional in vivo-like matrices. A primary limitation in investigating the effects of matrix stiffness in three dimensions is the lack of materials that can be tuned to control stiffness independently of matrix density. Here, we use collagen-based scaffolds where the mechanical properties are tuned using non-enzymatic glycation of the collagen in solution, prior to polymerization. Collagen solutions glycated prior to polymerization result in collagen gels with a threefold increase in compressive modulus without significant changes to the collagen architecture. Using these scaffolds, we show that endothelial cell spreading increases with matrix stiffness, as does the number and length of angiogenic sprouts and the overall spheroid outgrowth. Differences in sprout length are maintained even when the receptor for advanced glycation end products is inhibited. Our results demonstrate the ability to de-couple matrix stiffness from matrix density and structure in collagen gels, and that increased matrix stiffness results in increased sprouting and outgrowth.

Figure 1

Collagen gel mechanical properties. The equilibrium compressive moduli of 1.5 mg/mL collagen gels were measured by confined compression testing. Data presented as mean ± SEM

Figure 2

The effects of non-enzymatic glycation on collagen fibril arrangement and organization. (a) Confocal reflectance microscopy was used to image the collagen gels formed from solutions that had been incubated with 0, 50, 100, 150, 200, or 250 mM ribose. (b) Collagen fiber organizations were compared using the image autocorrelation mean radius. Scale is 20 μm

Figure 3

The effects of non-enzymatic glycation on fluorescently-labeled collagen fibril arrangement. TRITC-labeled collagen was incubated with 0, 50, 100, 150, 200, or 250 mM ribose and imaged with fluorescence microscopy. Scale is 20 μm

Figure 4

The effects of non-enzymatic glycation on the polymerization dynamics of collagen. (a) Collagen polymerization dynamics were measured as a function of ribose concentration based on absorbance readings at 500 nm during polymerization. (b) The fibril formation rates of glycated collagen polymerization were found by fitting a sigmoidal curve to the polymerization data in (a) and reporting the slope of the linear section of the curve. Data presented as mean ± SEM

Figure 5

Endothelial cell proliferation within glycated collagen gels. ECs were embedded within collagen gels that had been glycated with 0, 50, 100, 150, 200, or 250 mM ribose. Cellular viability and proliferation was assessed by measuring the DNA content of gels at 0, 7, 14, or 21 days.

Figure 6

Single cell response to matrix stiffness. Isolated ECs were embedded within collagen gels formed from solutions that had been incubated with 0, 50, or 100 mM ribose. Cells were allowed to spread for 24 hours and then were fixed and stained for actin. (a) ECs were imaged using confocal microscopy and the projected cell (a) area and (b) perimeter were determined. Data presented as mean + SEM, * indicates p<0.05, Scale is 20 μm

Figure 7

Angiogenic outgrowth response to matrix stiffness. Multi-cellular spheroids were embedded within collagen polymerized from solutions treated with 0, 50, or 100 mM ribose. (a) Spheroids were imaged using brightfield microscopy at 0, 1, and 2 days after embedding. (b) The total length of extensions and (c) the average number of extensions per spheroid were measured at day 1. (d) The projected spheroid area, including the extensions, was measured over the course of 5 days. Data presented as mean + SEM, * indicates p<0.05, Scale is 200 μm

Figure 8

Long-term angiogenic outgrowth response to matrix stiffness. Multi-cellular spheroids were grown in collagen gels polymerized form solutions treated with 0, 50, or 100 mM ribose for 8 days. Spheroids were fixed and stained for actin and DAPI and imaged using confocal microscopy. Scale is 200 μm

Figure 9

The effects of RAGE inhibition on spheroid outgrowth. Multi-cellular spheroids were embedded within collagen polymerized from solutions treated with 0 or 100 mM ribose and fed at 1 hour with either complete media with or without 10 μg/ml anti-RAGE blocking antibody. (a) The projected spheroid areas were measured at days 0, 1, and 2 after embedding. Data presented as mean + SEM, * indicates p<0.05