The presented techniques for liver harvesting, cannulation and perfusion using our proprietary device enable sophisticated perfusion set-ups to improve decellularization and recellularization experiments in rat livers.
Decellularization and recellularization of parenchymal organs may enable the generation of functional organs in vitro, and several protocols for rodent liver decellularization have already been published. We aimed to improve the decellularization process by construction of a proprietary perfusion device enabling selective perfusion via the portal vein and/or the hepatic artery. Furthermore, we sought to perform perfusion under oscillating surrounding pressure conditions to improve the homogeneity of decellularization. The homogeneity of perfusion decellularization has been an underestimated factor to date. During decellularization, areas within the organ that are poorly perfused may still contain cells, whereas the extracellular matrix (ECM) in well-perfused areas may already be affected by alkaline detergents. Oscillating pressure changes can mimic the intraabdominal pressure changes that occur during respiration to optimize microperfusion inside the liver. In the study presented here, decellularized rat liver matrices were analyzed by histological staining, DNA content analysis and corrosion casting. Perfusion via the hepatic artery showed more homogenous results than portal venous perfusion did. The application of oscillating pressure conditions improved the effectiveness of perfusion decellularization. Livers perfused via the hepatic artery and under oscillating pressure conditions showed the best results. The presented techniques for liver harvesting, cannulation and perfusion using our proprietary device enable sophisticated perfusion set-ups to improve decellularization and recellularization experiments in rat livers.
Decellularization and recellularization may enable the generation of functional, transplantable organs in vitro 1. By removing cells and antigenic material (e.g., DNA, alpha-Gal epitopes) from an organ, the non- or less-immunogenic extracellular matrix (ECM) can be obtained. This matrix conserves the three-dimensional microanatomy of an organ and can serve as the ideal biomatrix for repopulation with cells of a different, possibly xenogeneic origin 2. Thus, a decellularized rat liver matrix could be repopulated with human liver cells. This humanized micro-liver could serve as an ex vivo model for research on diseases (e.g., inborn metabolic diseases, viral diseases or malignancies) or for preclinical pharmaceutical testing 3.
Several different protocols for rat liver perfusion decellularization have already been published 4-13. In all protocols, decellularization was achieved by perfusion of alkaline ionic or non-ionic detergents via the cannulated portal vein. To the best of our knowledge, we were the first group to report rat liver decellularization by selective perfusion via the portal vein and/or the rat hepatic artery 14. Enabling the selective perfusion of the different vascular systems in the liver may enable better decellularization results and, furthermore, may play an important role in cellular repopulation.
In the study detailed here, livers were perfused in a custom-made proprietary perfusion device, enabling perfusion under oscillating pressure conditions. These pressure conditions mimic the physiologic respiratory-dependent perfusion of the liver: in situ, the liver hangs under the copula of the diaphragm, whose movement during respiration has a direct impact on liver perfusion. Inspiration specifically leads to lowering of the diaphragm and squeezing of the liver, optimizing hepato-venous outflow, whereas expiration leads to elevation of the liver and lowering of the intraabdominal pressure to optimize portal-venous inflow 15.
Our aim was to evaluate whether oscillating pressure conditions have an impact on the homogeneity of rat liver perfusion decellularization by mimicking intraabdominal conditions ex vivo. The homogeneity of the decellularization process may be an underestimated factor in perfusion decellularization. All known agents used for liver decellularization cause alterations to the ECM. Cells in poorly perfused areas remain within the ECM, whereas other areas are already completely decellularized. To dissolve the remaining cells, the perfusion duration or pressure must be elevated, causing more alterations to the well-perfused areas. Thus, detergents for decellularization should be distributed homogenously within the organ.
Animals were kept at the Facility for Experimental Medicine (FEM, Charité, Berlin, Germany), and all experimental protocols were reviewed and approved by the State Office of Health and Local Affairs (LAGeSo, Berlin, Germany; Reg. No. O 0365/11).
1. Liver Harvesting
2. Detergent Solutions
3. Decellularization
Figure 1: Scheme of the Perfusion Device for Selective Arterial and Portal Venous Perfusion of Four Rat Livers under Oscillating Pressure Conditions. (modified from Struecker et al. 14)
Figure 2: Scheme of the perfusion set-up. Link the perfusion chamber (1,400 cm3)(1) to the distal end (4) of the bubble trap (5) via the portal venous access (2) or via the arterial access (3). The bubble trap (5) should be equipped with an obturation (4) at the distal end and a vent (6). Connect the bubble trap (5) to a Heidelberger extension (7). Link the Heidelberger extension to the pump segment (8). Furthermore, connect the pump segment (8) to another Heidelberger extension (9). Insert the last Heidelberger extension (9) into the bottle of detergent solution (10). Connect the respirator (11) to the perfusion chamber (or to the pressure distributor in the case of multiple perfusions). To drain the reactor fluids, connect the outflow to the waste bottle (12).
Figure 3: Perfusion Protocol for All Experimental Groups.
The homogeneity and thus the effectiveness of different decellularization protocols were evaluated by macroscopic observation, histological analysis, and analysis of the remaining DNA content within decellularized liver matrices. Furthermore, corrosion casting was performed to visualize the intact microanatomy of livers after decellularization.
Macroscopy
During decellularization, livers become lucent, indicating the removal of cellular content. Livers perfused via the portal vein showed inhomogeneous elucidation (and thus decellularization), with macroscopically visible remaining cells during and after decellularization (white arrows). Livers perfused via the hepatic artery showed a more homogenous decellularization course and no visible remaining cells. If livers were perfused with the application of oscillating pressure conditions, no remaining cells were visible, irrespective of the perfusion route.
Figure 4: Macroscopic Results of Rat Liver Decellularization without Oscillating Pressure Conditions. Left: Rat liver decellularized via the hepatic artery. The liver appears lucent, without macroscopically visible remaining cells. Right: Liver perfused via the portal vein. The liver appears inhomogeneously decellularized, with macroscopically visible cells
Figure 5: Macroscopic results of rat liver decellularization under oscillating pressure conditions. Left: Rat liver decellularized via the hepatic artery. Right: Liver perfused via the portal vein. Both livers appear lucent, without visible remaining cells.
Histological Evaluation
Histological evaluation of decellularized liver matrices supported the macroscopic findings. Livers perfused via the hepatic artery showed no remaining cells, irrespective of whether they were perfused under oscillating pressure conditions. Livers perfused via the portal vein showed zones of remaining cells, even if they were perfused with the application of oscillating pressure conditions. However, the application of oscillating pressure conditions reduced the remaining cell mass in livers perfused via the portal vein. The ECM of the livers was conserved, without visible differences across all applied protocols.
Figure 6: H/E Staining of Decellularized Liver Matrices. A-P: Decellularized liver matrix perfused via the hepatic artery without oscillating pressure conditions. A+P: Decellularized liver matrix perfused via the hepatic artery under oscillating pressure conditions. PA-P: Decellularized liver matrix perfused via the portal vein without oscillating pressure conditions. PA+P: Decellularized liver matrix perfused via the portal vein under oscillating pressure conditions. Arrows: Remaining cell clusters.
DNA content
The DNA content per dry weight of liver matrix declined in all experimental groups compared with that in native livers, although differences were only statistically significant in the groups perfused under oscillating pressure conditions (PV+P; A+P). The lowest amount of DNA per dry weight was found in livers perfused via the hepatic artery under oscillating pressure conditions.
Figure 7: DNA Content Per Dry Weight of Liver Matrix (modified from Struecker et al. 14). Livers decellularized via the hepatic artery show less remaining DNA than those perfused via the portal vein do. Livers perfused under oscillating pressure conditions show less remaining DNA than those decellularized without these conditions do. Thus, livers perfused via the hepatic artery under oscillating pressure conditions show the least remaining DNA content.
Corrosion Casting
Corrosion casting of decellularized liver matrices confirmed that the microanatomy of livers (including the protein framework of the portal venous system, the arterial system and the biliary system) is conserved during decellularization.
Figure 8: Photograph of a Corrosion Casting of a Decellularized Rat Liver Matrix (modified from Struecker et al. 14). Despite removal of all cells, the microanatomy of the organ, including the portal venous (blue) and arterial (red) systems, are conserved. The remaining main branches of the biliary system are casted in yellow.
Table 1: Experimental Groups for Assessing the Effect of Oscillating Pressure Conditions and the Perfusion Route on Rat Liver Decellularization.
Although the presented technique for rat liver harvesting and decellularization is easily reproducible, there are certain critical steps to consider:
During preparation for liver harvesting, it is important to avoid severe bleeding because it will activate blood coagulation and may lead to blood clot formation within the liver. In our opinion, it is advantageous to incise the abdominal aorta directly before cannulation of the portal vein to avoid blood inflow via the hepatic artery during perfusion via the portal vein. Once the liver is cannulated and clearly perfused, blood clot formation is no longer an issue.
After liver harvesting and transportation to the perfusion device, the connection of livers to the perfusion device is a particularly critical step because the entry of gas bubbles into the perfusion circle must be avoided to prevent gas embolism. It is useful to prefill the three-way valve, which is connected to the liver cannula, with PBS using a syringe and a thin cannula directly before connecting the valve to the perfusion system.
After perfusion is established, whether all tube connectors, three-way valves and tubes are perfectly connected should be re-checked. If livers are perfused for longer durations (e.g., O/N), even a small error in connection can lead to air siphoning into the perfusion system, thus leading to fatal perfusion outcomes.
The presented data are limited by the fact that the optimal pressure and frequency of oscillating pressure changes for the best decellularization outcomes remain unclear. Furthermore, pressure-controlled perfusion may lead to better and more stable results than flow-controlled perfusion does. Pressure-controlled perfusion enables the application of one perfusion protocol to livers of different sizes and weights, with the same outcome, because the flow will automatically be adapted.
The hepatic artery and portal vein are significantly different in terms of size and physiologic flow rates. Thus, a constant perfusion rate (5 ml/min) will certainly result in different perfusion pressures within the two vascular systems. Similarly, a constant perfusion pressure will result in different perfusion rates. One method to compare the decellularization efficacy via the two perfusion routes would be to perform perfusion at physiologic perfusion pressures or perfusion rates via both routes. However, to make the results comparable in the current study, we chose a constant perfusion rate for both routes. Unfortunately, using our device, we were not able to measure the perfusion pressure within perfused livers.
Our next-generation perfusion devices will be smaller to minimize the amount of perfusion media necessary. Furthermore, pump devices will enable pressure-controlled perfusion.
The presented technique appears very useful for reproducible, homogenous rat liver decellularization. Whether the presented technique is appropriate for recellularization and organ maturation in vitro has to be further addressed. Whether oscillating pressure conditions have an impact on repopulated cells will be a subject of further experiments.
The presented technique is more sophisticated than other rat liver perfusion devices and shows several clear advantages: 1) The homogenous decellularization is reproducible, with good results when oscillating pressure conditions are applied. 2) The perfusion device is sealed and enables sterile decellularization, which is a prerequisite for matrix repopulation and long-term perfusion. 3) The presented protocol is shorter than other published protocols but is still very effective. 4) Selective perfusion of the portal vein and the hepatic artery renders selective cell repopulation via different systems possible, which may be an important tool.
The perfusion device will be applied in further decellularization experiments to refine and optimize rat liver decellularization. The effect of pressure-controlled perfusion will be experimentally evaluated and compared with flow-controlled perfusion. The addition of further elements (e.g., a “dialysis unit”, a heat exchanger, an oxygenator) for repopulation, culture and maturation experiments appears possible and reasonable to further develop the presented device.
The authors have nothing to disclose.
The authors would like to gratefully thank Steffen Lippert, Khalid Aliyev, Korinna Jöhrens and Katharina Struecker for their help during this project.
Self built arterial cannula | |||
Portex Non Sterile Polyethene Tubing | SIMS Portex | REF 800/110/100 | 0,28mm ID 0,61mm OD |
Portex Non Sterile Polyethene Tubing | SIMS Portex | REF 800/110/200 | 0,58mm ID 0,96m OD |
Venodrop Safe butterfly catheter | Fresenius Kabi | 3275851 | 21 G |
portal vein cannula | |||
Periphereal Venous Catheter | BD | 393224 | BD Venflon Pro 20G |
three-way stopcock | smiths medical | 888-101RE | |
surgery | |||
Cotton Sticks | Hecht-Assistent | 4302 | |
Cotton Pads | Shaoxing Zhengde Surgical dressing | 13H118-03 | |
Gauze Bandage | Hubei Haige Medical Instruments | 14388 | |
Ringer Solution | Fresenius Kabi | 13 HKP022 | 1000ml |
10ml Syringe | Braun | 4606108V | 10ml/ Luer Solo |
5ml Syringe | Braun | 4606061V | 5ml /Luer Solo |
Suture (Silk 6/0) | Resorba | H1F | LOT 105001.81 |
medical drape | Shaoxing Zhengde Surgical dressing | D0613011 | |
surgical instruments | |||
needle holder | Geuder | 17570 | |
micro-forceps | Inox-Electronic | 91150-20 | |
micro-scissors | Martin | 11-740-11 | |
micro-forceps | S&T | 112314 | |
Clamp | Aesculap | BH111R | |
scissors | F S T | 14501-14 | |
surgical forceps | Aesculap | BD 557 | |
Decellularisation | |||
Respirator | Resmed | 14.24.11.0004 | SmartAIR ST |
Perfusion Device | Charite, medical engineering laboratory | custome-made device | decellularisation device |
peristaltic pump ismatec reglo ICC | IDEX | ISM4408 | 4-channel |
heidelberger extension 75 cm | Fresenius Kabi | 2873 | 75 cm |
MS/CA pump-segment | IDEX | IS 3510 | MS/CA/click'n'go/POM-C |
CA 2-stopper tube | Pharmed | BPT NSF-51 | |
bubble trap | custome-made item | ||
Luer Lock hose connector | Neolab | No. 02-1887 | |
Detergents | |||
SDS pellets | Carl Roth | CN30.4 | 2,5 kg |
Triton X-100 | Carl Roth | 3051.1 | 10l |
PBS | Gibco | 14190-094 | DPBS |
staining | |||
Eosin 1% | Morphisto | 10177 | |
Mayer hematoxylin | AppliChem | A4840 | |
gomori staining | Morphisto | 11104 | |
AlcainBlue-PAS staining | Morphisto | 11388 | |
Direct Red 80 | Sigma Aldrich | 365548 |