Tissue Engineering of Heart Valves: Decellularized Porcine and Human Valves Differ Importantly in Residual Potential to Attract Monocytic Cells

Erwin Rieder, Gernot Seebacher, Marie-Theres Kasimir, Eva Eichmair, Birgitta Winter, Barbara Dekan, Ernst Wolner, Paul Simon and Guenter Weigel

All Figures and Captions from original paper

Introduction

Porcine (pig) and human heart valves have already been implanted into humans, but have experienced early failure do to calcification and severe immune response. Decellularized porcine and human valves have also been adapted to avoid a negative host response and increase the longevity of the implant. In general, human valves have proved to have a higher biological performance compared to porcine valves. In an attempt to characterize the crucial differences between decellularized and native valves as well as the differences between porcine and human valves, this study investigated the migratory response of human monocytes for each of these conditions. Monocytes are attracted to the endothelial lining and once attached, leukocytes also begin to migrate toward the tissue causing an immunological response eventually leading to failure of the heart valve scaffold. By understanding the initial inflammatory response, more durable heart valve replacements can be designed in the future.

Methods

First, 12 porcine and 8 human pulmonary heart valves were obtained. 6 porcine valves and 4 human valves were decellularized using a detergent-based method. Two 5mm longitudinal slices of each valve were then excised and the rest of the samples were snap-frozen. The slices were then visualized using hematoxylin-eosin (HE) staining and TOPRO-2 dye to specifically detect cell nuclei. Immunohistostaining using polyclonal anti-porcine collagen type I and III (also reacts with human types) and monoclonal anti-elastin antibodies were used to better visualize the cell structure as a whole.

100mg from the snap-frozen samples of both the leaflet and pulmonary wall from the porcine and human were homogenized using a mortar and pestle and then centrifuged. The supernatant was collected and protein quantified through a Bradford assay. The migration assay was then preformed using a PET membrane with 3mm pores in a 6-well culture plate. The bottom of each well was covered with 1.5mL of protein extract from the samples and 1.5 x 10⁶ U-937 monocytic cells were placed on the upper side of the filter membrane. After a 24-hour incubation at 5% CO₂ atmosphere at 37°C, the cells at the bottom of each well were scraped and subsequently centrifuged. Cells were stained using a crystal violet solution and counted using a hemocytometer. This process was repeated using 40mg instead of 100mg of the original tissue samples.

200mg of each sample was also homogenized for protein electrophoresis. The total protein content was again determined and 75mg of protein for each sample was combined with 75 mL SDS-PAGE buffer. After boiling for 3 minutes, the samples were run on a 10% SDS-PAGE resolving gel with a 4% acrylamide stacking gel. The gel was then stained with Coomassie Brillant Blue, scanned, and visualized using Adobe Photoshop 5.0

Results & Conclusion

No cells were observed for the decellularized heart valves or conduit walls for both the human and porcine samples. However, elastin and collagen were still present for all the decellularized cases indicating a successful decellularization method.

Figure 1- A, B, HE staining of native (left) and decellularized (right) porcine pulmonary heart valves. Shown are conduit wall (A) and leaf- let (B). Original magnificatio X400. C, D, Native (left) and decellularized (right) porcine pulmonary heart valves, staining for collagen I and III (green), elastin (red), and DNA (white). Confocal laser scanning microscopy, original magnification X400. Shown are conduit wall (C) and leaflet (D).

As seen in the figure below, decellularization in general decreased monocyte migration for all 4 cases. The decellularized human leaflets and conduit walls essentially eliminated the attraction of monocyte cells having a result comparable to the negative control case. The decellularized porcine samples had a reduced attraction of monocyte cells compared to the native case but still experienced significant monocyte accumulation. The native human cusp had a similar attraction of monocyte cells compared to the decellularized porcine cusp.

Figure 2- Migration of U-937 cells toward extracts of pulmonary conduit wall (A) and cusps (B). nat porc indicates native porcine; dec porc, decellularized porcine; nat hum, cryopreserved pul- monary homograft; and dec hum, decellularized pulmonary homograft. *P < 0.05, **P < 0.01, Wilcoxon rank-sum test. (Note: Figure 1 not included)

The SDS-PAGE results shown below demonstrate that there are still significant amounts of protein after decellularization for each case. The remaining protein in the porcine samples likely contributes to the migratory response, while the remaining protein in the human sample does not. The residual proteins from the porcine sample should be identified to determine the specific proteins that cause monocyte attraction. Overall, this research revealed that porcine heart valves are not as biocompatible as they were originally considered to be. Additionally, although decellularization reduces monocyte migration, specific residual proteins still contribute to increased monocyte concentration at the heart valve surface, thereby reducing the longevity and performance of the implant.

Figure 3- SDS-PAGE of homogenized pulmonary conduit wall (A) and cusp (B) tissue. I indicates native porcine; II, decellular- ized porcine; III, cryopreserved pulmonary homograft; and IV, decellularized pulmonary homograft.

Commentary

This research provides an initial look into the factors affecting biocompatibility of porcine and human heart valves. Although decellularization was known to decrease the host immune response, this research identified the specific effects decellularization and sample origin have on monocyte migration and thus initial immune response. My main critique with this research is its emphasis on the applicability of the results to the functioning of the heart valve in vivo. Although the monocyte migrations were altered with the decellularization treatment, it is unclear whether this result would apply to a functioning heart valve in the body. There are various proteins in the blood that could adsorb to the decellularized heart valve surface drastically altering its monocyte attraction potential. This same principle applies to the differences noted between porcine and human heart valves. This paper seemed to jump to the conclusion that their result will also be observed in vivo and that monocyte recruitment is the reason for calcification and early failure of an implanted heart valve.

Also, I am skeptical as to whether their decellularization methods were as successful as they reported. Although their presented images did not reveal any cells, the images were said to be “representative “ and could have been subject to user bias. Different decellularization methods involving the use of chelating substances, antibiotics, or enzymes instead of detergents should also be performed and run on an SDS-PAGE gel to compare results.

Finally, I believe the researchers could have conducted preliminary SDS-PAGE analysis involving the identification of protein bands from their results. By knowing some of the residual proteins after decellularization, specific mechanisms concerning monocyte attraction could be inferred. Overall, this paper provides an intriguing glance at the possible reasons behind early heart valve replacement failure; however, much more research is needed to apply this knowledge to the selection and design of suitable heart valve replacements.