Dynamic, nondestructive imaging of a bioengineered vascular graft endothelium

Abstract

Bioengineering of vascular grafts holds great potential to address the shortcomings associated with autologous and conventional synthetic vascular grafts used for small diameter grafting procedures. Lumen endothelialization of bioengineered vascular grafts is essential to provide an antithrombogenic graft surface to ensure long-term patency after implantation. Conventional methods used to assess endothelialization in vitro typically involve periodic harvesting of the graft for histological sectioning and staining of the lumen. Endpoint testing methods such as these are effective but do not provide real-time information of endothelial cells in their intact microenvironment, rather only a single time point measurement of endothelium development. Therefore, nondestructive methods are needed to provide dynamic information of graft endothelialization and endothelium maturation in vitro. To address this need, we have developed a nondestructive fiber optic based (FOB) imaging method that is capable of dynamic assessment of graft endothelialization without disturbing the graft housed in a bioreactor. In this study we demonstrate the capability of the FOB imaging method to quantify electrospun vascular graft endothelialization, EC detachment, and apoptosis in a nondestructive manner. The electrospun scaffold fiber diameter of the graft lumen was systematically varied and the FOB imaging system was used to noninvasively quantify the affect of topography on graft endothelialization over a 7-day period. Additionally, results demonstrated that the FOB imaging method had a greater imaging penetration depth than that of two-photon microscopy. This imaging method is a powerful tool to optimize vascular grafts and bioreactor conditions in vitro, and can be further adapted to monitor endothelium maturation and response to fluid flow bioreactor preconditioning.

Figure 1. Vascular scaffold fabrication, MIC integration, and FOB imaging method.

A–C) Illustration of the layer-by-layer electrospinning fabrication technique used to embed MICs within the wall of the vascular scaffold to facilitate imaging. D,E) Images of a vascular scaffold with MICs embedded within the wall. F) Image of a vascular scaffold sectioned lengthwise to validate endothelium imaging (half-vessel scaffold). G) Image of the fiber optic micro-mirror inserted within an MIC. Light is directed at a 90° angle relative to the fiber optic to allow excitation of ECs on the lumen of the vascular graft. H) Photograph of the vascular scaffold during an imaging experiment with the fiber-guided excitation light on. I) Schematic of the fiber optic imaging method used to create FOB images through the scaffold wall. Excitation laser light was guided to the fiber optic micro-mirror to excite GFP-ECs on the lumen of the vascular scaffold. The excitation spot created by the laser light was incrementally rastered on the lumen surface and GFP fluorescence was measured through the scaffold wall using an EMCCD detector. Linking the excitation spot to the GFP fluorescence values enabled fluorescence mapping to create a fiber optic-based (FOB) image. A corresponding control image was obtained by a direct-line-of-sight control camera to validate the imaging method on half-vessel scaffolds.

Figure 2. Quantification of GFP-EC proliferation and lumen coverage.

A) A vascular scaffold sectioned lengthwise was seeded with GFP-ECs and 3 separate 1 mm×250 µm areas of the graft were imaged daily until the lumen reached confluency at Day 4 (representative images shown). B) Images were analyzed to quantify the amount of cell coverage on the vascular lumen (n = 3 images/time point). The results show that there is good agreement between amount of cell coverage for the FOB images, obtained through scaffold wall, and direct-line-of-sight control images at each time point. Data are presented as mean ± one standard deviation and values marked with the same letter are not significantly different (p<0.05).

Figure 3. Quantification of GFP-EC detachment and apoptosis.

A) Representative images of a near confluent endothelium that was subjected to a dilute concentration (1 µM) of trypsin EDTA to induce cell detachment from the lumen. An area of 250 µm×250 µm was continuously scanned to image GFP-EC detachment from the lumen over time. B) Extent of GFP-EC detachment from the lumen was quantified by analyzing the area of cell coverage from the fluorescent images (n = 3 images/time point). The FOB and corresponding control images had equivalent amounts of cell coverage at each time point. C) A near confluent endothelium was exposed to varying concentrations of camptothecin to induce GFP-EC apoptosis. A red fluorescent annexin-V conjugate was then added to visualize the extent of EC apoptosis on the vascular scaffold lumen. Three different 500 µm×250 µm areas were imaged to detect both GFP and red fluorescent annexin-V (apoptotic cells) on each scaffold for each condition. Results showed that varying levels of GFP-EC apoptosis could be noninvasively imaged and D) the fraction of apoptotic cells could be quantified through the scaffold wall using the FOB imaging method (n = 3 images/condition). Data are presented as mean ± one standard deviation and values marked with the same letter are not significantly different (p<0.05).

Figure 4. Effect of lumen topography on graft endothelialization.

Full-vessel grafts with varied electrospun fiber diameters were fabricated to study the effect of lumen fiber diameter and varying topography on EC proliferation and endothelialization. A) Mean fiber diameters of 3.32 µm, 1.00 µm and 0.36 µm were electrospun by varying electrospinning parameters as shown in Table 1 (images captured using SEM). After cell seeding, three separate 500 µm×250 µm areas of each vascular graft were imaged daily to assess luminal coverage and EC growth over time. Representative FOB images of GFP-EC coverage are shown for each scaffold fiber diameter at B) Day 0, C) Day 3, and D) Day 7. E) After FOB imaging on Day 7, scaffolds were removed from the bioreactor, sectioned in half, stained for F-actin (red) and cell nuclei (blue), and imaged using a conventional florescence microscope to provide a comparison to the FOB images at Day 7. F) A total of n = 3 scaffolds for each fiber diameter were imaged, and n = 3 areas (500 µm×250 µm) for each scaffold, yielding a total of n = 9 images for each scaffold group every day. The FOB imaging method provided a noninvasive, accurate measurement of vascular graft lumen endothelialization over time. Data are presented as mean ± one standard deviation.

Figure 5. Imaging penetration depth: comparison of FOB imaging method to Two-photon microscopy.

A) EC-seeded half-vessel scaffolds were imaged with a two-photon microscope in direct-line-of-sight to the lumen. B) Representative image of GFP-ECs in direct-line-of-sight to the vessel lumen. C) The two-photon microscope was then used to image GFP-ECs on the lumen through the wall of vascular scaffolds of varying thicknesses (t = 95, 230 and 520 µm). Images of GFP-ECs on the vessel lumen through a D) 95 µm and E) 230 µm thick scaffold. Although cells were visible through the 95 µm thick scaffold, GFP-ECs could not be resolved through the 230 and 520 µm thick scaffolds. The FOB imaging method, however, was capable of resolving individual GFP-ECs through a ∼510 µm thick scaffold wall (G) and the overall cell distribution of the same region of interest matched very closely to the direct-line-of-sight fluorescence image (F).