Fig 1.
Decellularization of the rat liver is achieved following one cycle of DET and EDTA-DET treatments.
(A) Timeframe and infusion of solutions for the DET and EDTA-DET protocols. (B) Macroscopic appearance of the liver scaffolds showed no difference between the DET and EDTA-DET scaffolds. Following dH2O addition, the livers became blanched, with SDC and DNase addition resulting in the livers becoming transparent. (C) H&E staining demonstrated absence of cells in sections from DET and EDTA-DET scaffolds. (D) DNA quantification reduced DNA in DET and EDTA-DET scaffolds compared with fresh tissue (p<0.001). (E-F) Histograms showing the signal intensities, associated to relevant nuclear proteins (E) and cytoplasmic proteins (F); ‡: p<0.001, compared to fresh tissue, #: p<0.001, compared to DET scaffold, scale bar in macroscopic images: 2cm, scale bar on histology: 100μm.
Fig 2.
Decellularization preserves ECM components, with a higher amount of collagen and elastin present in the EDTA-DET scaffold.
(A) MT staining demonstrated acellularity in both scaffolds, showing composition only by connective tissue. (B) PR staining demonstrated composition mostly by collagen fibers. (C) EVG staining revealed maintenance of elastin fibers in the inner surface of blood vessels. (D) AB staining confirmed the presence of GAG in both scaffolds. (E) Collagen was significantly increased in EDTA-DET scaffolds (p<0.01) compared with fresh tissue. (F) Elastin significantly decreased in DET (p<0.001) and EDTA-DET (p<0.01) scaffolds compared with fresh tissue. Elastin was higher in EDTA-DET scaffolds compared with DET scaffolds (p<0.05). (G) GAG quantification showed that while both DET and EDTA-DET scaffolds had significantly reduced GAG amount, EDTA-DET scaffolds had significantly less when compared to DET scaffolds (p<0.001). (H) Venn Diagram showing the number and the distribution over sample types of proteins identified in Fresh, DET and EDTA-DET treated liver tissues; *: p<0.05, †: p<0.01, ‡: p<0.001, compared to fresh tissue, §: p<0.05, ||: p<0.01, compared to DET scaffold, scale bar: 100μm.
Fig 3.
Immunostaining demonstrates the preservation of ECM proteins in the decellularized scaffolds.
(A) Collagen I staining in fresh tissue was positive as fine strands in the parenchymal space as well as around the blood vessels (asterisk). This was preserved following decellularization with a strong signal from vascular structures both in DET and EDTA-DET scaffolds. (B, C) Collagen III and collagen IV staining demonstrated a similar distribution and preservation in fresh tissue and decellularized scaffolds. Weak parenchymal staining was observed, and a strong signal surrounding vascular and biliary structures. EDTA-DET scaffolds demonstrated slightly increased staining in the parenchymal space when compared to DET scaffolds. (D) Fibronectin showed strong staining around the main blood vessels in fresh tissue with a more distributed signal pattern in decellularized scaffolds. (E) Laminin showed strong staining around the main blood vessels in fresh tissue with a more distributed signal pattern in decellularized scaffolds; scale bar: 100 μm.
Fig 4.
Proteomics demonstrate the preservation of ECM components in the decellularized scaffolds.
Histograms showing the signal intensities, based on the summed extracted ions counts, measured for each peptide identified and associated to proteins belonging, accordingly to the GOCC classification, to the class of Collagens (A), Glycoproteins (B), Laminins (C) and Extracellular Matrix Proteins (D).
Fig 5.
Addition of EDTA to the protocol makes the matrix more compact in the parenchyma and vasculature.
(A-C) SEM confirmed acellularity in the scaffolds, demonstrating composition by a three-dimensional network of connective tissue fibers arranged in a honeycomb-like manner. The structure of the portal triads was identified clearly at low magnification (asterisk). (C) Scaffolds prepared with EDTA-DET were more tightly packed compared to the DET scaffolds, as the porous structure appeared to be compressed. (D) Quantification of the hepatocyte pockets in the EDTA-DET scaffold showed a reduction to approximately 50% of the size in the DET scaffolds. (E, F) Synchrotron-based x-ray phase contrast imaging corroborated these results, demonstrating a denser scaffold in scaffolds pre-treated with EDTA. (G) Infusion of trypan blue dye via the portal vein showed maintenance of the vascular network architecture. There was no dye leakage through the walls to the surrounding tissue; the dye followed the regular pathway of flow to the IVC. Macroscopically, the DET scaffolds were seen to possess a denser vascular network with the dye diffusing through the vessels more readily. (H) Dye intensity within the DET scaffolds followed an S-shaped curve, reaching a maximum point at approximately 35 seconds. Dye intensity within the EDTA-DET scaffold increased in a less steep S-shaped manner with the exponential part lasting 45 seconds instead of 15. Maximum intensity was reached at 75 seconds. (I) These results were paralleled by the quantification of surface area of dye distribution with the point of maximal surface area coverage having a difference of 40 seconds between the two scaffolds; #: p<0.001, compared to DET scaffold, scale bar: 1cm.
Fig 6.
Seeding of the DET scaffolds by microinjection demonstrates complete repopulation at 14 days.
(A) DET scaffolds were microinjected with HepG2 cells in a multifocal manner and the seeded constructs harvested at day 1, 4 and 14. (B) H&E staining demonstrated a wide distribution across the scaffold at day 1 and a complete repopulation at 14 days (B). At higher magnification, cells were seen repopulating the hepatic spaces, forming colonies and secreting their own ECM at 14 days (C, asterisk). Ki67 staining demonstrated proliferation at days 1 and 4, with only a few ki67-positive cells at 14 days (D). Quantification of ki67-positive cells as a percentage to DAPI-positive cells (F). Staining for CC-3 showed a few positive cells across the three time-points (E), with quantification demonstrating a non-significant increase (G); scale bar: 100μm.