This protocol details the use of a special intravenous catheter, standardized sterile disposable tubing, temperature control complemented by real-time monitoring, and an alarm system for two-step collagenase perfusion procedure to improve the consistency in the viability, yield, and functionality of isolated primary rat hepatocytes.
Primary hepatocytes are widely used in basic research on liver diseases and for toxicity testing in vitro. The two-step collagenase perfusion procedure for primary hepatocyte isolation is technically challenging, especially in portal vein cannulation. The procedure is also prone to occasional contamination and variations in perfusion conditions due to difficulties in the assembly, optimization, or maintenance of the perfusion setup. Here, a detailed protocol for an improved two-step collagenase perfusion procedure with multiparameter perfusion control is presented. Primary rat hepatocytes were successfully and reliably isolated by taking the necessary technical precautions at critical steps of the procedure, and by reducing the operational difficulty and mitigating the variability of perfusion parameters through the adoption of a special intravenous catheter, standardized sterile disposable tubing, temperature control, and real-time monitoring and alarm system. The isolated primary rat hepatocytes consistently exhibit high cell viability (85%-95%), yield (2-5 x 108 cells per 200-300 g rat) and functionality (albumin, urea and CYP activity). The procedure was complemented by an integrated perfusion system, which is compact enough to be set up in the laminar flow hood to ensure aseptic operation.
Primary hepatocytes are important tools for liver-related basic research, disease treatment, and application such as drug testing. The current gold standard for primary hepatocyte isolation is the two-step collagenase perfusion procedure1,2,3 introduced by Seglen in the 1970s4. However, this procedure is technically challenging and has a high failure rate when performed by novice surgeons. Even when a perfusion is considered successful, drastic differences in hepatocyte viability (typically 60%-95%) and yield (0.5-5 x 108 per 200-300 g rat) may be observed between isolations. This influences the quality and scale of downstream experiments. Apart from the technical procedure, the perfusion setup used for the isolation, either commercially available or custom built, is a contributing factor. Attention must be given to the assembly, optimization, and maintenance of the perfusion setup. The purpose of this protocol is to improve the success rate and stability between isolations of primary rat hepatocytes through multiparameter perfusion control of the technical procedure and perfusion setup of the two-step collagenase perfusion procedure.
From the technical aspect, the most difficult step in the procedure is the portal vein cannulation. As for the other steps, if good practice is observed and general precautions are taken, the stability of the isolation can be improved. Therefore, understanding of the reasoning for each step is important so that the surgeon could respond to various variables that may occur during the procedure.
Various protocols for the isolation of hepatocytes and liver non-parenchymal cells from rat and mouse have been published1,2,5,6,7,8,9. The perfusion setups used in these protocols had several disadvantages, which include the reuse of perfusion tubing, problems with temperature control, need for routine optimization of perfusion parameters, and/or usage of unsuitable type of intravenous (IV) catheter for portal vein cannulation. The reuse of perfusion tubing will increase the chances of contamination, especially if the tubing was not cleaned and disinfected properly. Reuse of tubing without routine replacement will also expose the perfusion setup to problems such as leaky tubing or connectors, clogged bubble trap and constricted tubing, all of which will substantially reduce the perfusate pressure and flow rate, thus, affecting liver digestion efficiency. Without a constant heat source in some setups for temperature control, pre-warmed buffers will cool down over time, leading to low collagenase activity and digestion. Although other setups utilize a jacketed glass condenser connected to a water circulator to warm the buffer, they are bulky and require careful cleaning. Temperature, pressure, and flow rate of buffer exiting the catheter must be measured and optimized before the start of isolation toensure stable perfusion condition. Even after optimization, the parameters could still change halfway during isolation due to the actions of the operator, thereby leading to suboptimal perfusion and digestion. Most types of IV catheter are not suitable for portal vein cannulation because they do not allow continuous perfusion during cannulation. They are unable to immediately inform the surgeon when the cannulation is successful. Furthermore, it is challenging to secure the portal vein on the soft catheter without deforming it.
Here, we address these problems using standardized disposable sterile tubing, a silicone heater jacket for precise and stable temperature control, real-time monitoring and alarm system with data storage and management and use of a special IV catheter, which allows continuous perfusion while puncturing portal vein during cannulation. To the best of our knowledge, we are the first group to combine all these features into an integrated perfusion system (IPS) that is compact, making it highly portable and able to be fit into a laminar flow hood to ensure aseptic operation.
All procedures and animal housing were carried out under protocol numbers R15-0027 and R19-0669 in accordance with the requirements of the Institutional Animal Care and Use Committee (IACUC) of the National University of Singapore.
1. Preparation of solutions and surgical instruments
2. Setting up the IPS (see Figure 1)
Figure 2: Self-test interface. Please click here to view a larger version of this figure.
Figure 3: Operation interface. Please click here to view a larger version of this figure.
Figure 4: Popup window prompting for filename and username. Please click here to view a larger version of this figure.
3. Animal procedure
4. Liver perfusion and digestion
5. Hepatocyte isolation
6. Hepatocyte culture
A surgeon could tell whether liver perfusion is going on smoothly by observing the outcome after certain steps. The first outcome can be observed upon cannulation, cutting of the infrahepatic IVC, and restoring the perfusion flow rate. The liver should have completely changed color from dark red to brown, while maintaining its volume. If the liver looks slightly deflated and has a reddish tint or blotches of red, it means that the perfusion flow rate was set wrongly (too low), or the portal vein was not cannulated correctly. If the liver not only turns brown but becomes bloated and stiff, it means that the flow rate was set wrongly (too high), or the infrahepatic IVC was not cut properly. If only a few lobes are not being perfused well, it means that the catheter was inserted too deep and is only perfusing one branch of the portal vein. The second outcome can be observed after resection and clipping of the infrahepatic IVC. If the outflow from the suprahepatic IVC is slow and weak, it might mean that the catheter had detached during resection, or that the infrahepatic IVC was not clipped properly. The third outcome can be observed during digestion. On the brown liver, translucent spots/network should appear and enlarge. The liver should lose its springy consistency and become soft and mushy. We optimize our collagenase concentration so that this state can be reached within 12 min. When trying out a new lot of collagenase, digestion time should be extended until this state is reached. The fourth outcome can be observed when cells are released from the liver. When the cells are released, the media should look cloudy. If a lot of grainy texture is observed, it could mean that the liver was insufficiently digested. If the digestion is complete, only white weblike vessels should be left. Even if the digestion is partial (Figure 6), it is still possible to acquire good cells.
Our protocol described here consistently generate higher cell yield of up to 5 × 108 hepatocytes per isolation from rats weighing 200-300 g and cell viability between 88%-94.8% as determined by trypan blue counting12,13,14,15,16 (Table 2). Hepatocyte purity obtained as determined by immunostaining of albumin was 96.8 ± 2.0%11 (Figure 7). Hepatocytes in sandwich culture formed distinct bile canaliculi and had good cell-cell contact (Figure 8). The hepatocytes have high functionality as shown by albumin, urea, and CYP assays10 (Table 3).
Figure 1: Schematic diagram of the IPS. The perfusate starts in the perfusate bag or bottle (1, 2). Two flow regulators (3, 4) work together to control the passage of perfusate. Perfusate proceeds to a bubble trap (5). Subsequent processes are controlled by the LCD touch screen (6) of the main system. Perfusate is then pulled by a peristaltic pump (7) and passes through a silicone jacketed electronic heater (8), which heats up the perfusate to the desired temperature. The silicone heater jacket (8) is connected to the main system by a connector (9). When necessary, a disposable pressure sensor (10) can be connected to the main system by a connector (11), to allow measurement of the perfusion pressure. The perfusate then passes through a perfusion monitor (12), which measures the temperature of the perfusate and detects the presence of bubbles. The monitor can also determine whether the perfusate has run out. Monitor readings are transmitted to the main system via LoRa. The perfusate then proceeds to a bubble filter (13) and enters the liver through the portal vein cannula (14). Please click here to view a larger version of this figure.
Figure 5: Schematic diagram of the two-step collagenase perfusion for primary rat hepatocyte isolation. (A) Perfusion of calcium-free buffer. The roller clamp is completely loosened for inlet 1 and completely tightened for inlet 2. (B) Liver digestion with collagenase buffer. Resected liver is transferred to the top of a beaker to allow buffer recirculation. The roller clamp is completely loosed for inlet 2 and completely tightened for inlet 1. Please click here to view a larger version of this figure.
Figure 6: Liver after collagenase digestion and hepatocyte isolation. Shown is a partially digested liver lobe. Undigested parts of the liver remained brown. In digested parts of the liver, only white vessels were left behind. Please click here to view a larger version of this figure.
Figure 7: Immunostaining of hepatocytic marker for quantification of hepatocyte purity. (A) Sandwich culture of unpurified liver cell suspension (before differential centrifugation; left) and purified hepatocyte suspension (after differential centrifugation; right) seeded at low cell density. (B) Representative images show cells stained with DAPI (nucleus; blue) and antibody against albumin (hepatocyte-specific marker; green). (C) Quantification of hepatocyte purity. The purity was determined by counting albumin positive fluorescent stained cells in relation to the total cell number visualized by DAPI nuclear staining. Images in false colors were shown. Albumin (green), nucleus of cells with albumin (blue), nucleus of cells without albumin (red). Scale bars 100 µm. Please click here to view a larger version of this figure.
Figure 8: Phase contrast image of isolated primary rat hepatocytes from two-step collagenase perfusion in sandwich culture at day 3. Scale bar 50 µm. Please click here to view a larger version of this figure.
Name of Buffers/Media | Final concentration | Amount | Comments/Description |
Hepatocyte culture media | |||
William’s E media | 500 mL | Add all reagents into William's E media. Filter through 0.22 µm filter. Store at 4 °C. | |
BSA | 1 mg/mL | 0.5 g | |
Penicillin-Streptomycin (100X) | 1X | 5 mL | |
Insulin (20 mg/ml) | 0.5 µg/mL | 12.5 µL | |
Dexamethasone (1 mM) | 100 nM | 50 µL | |
Linoleic acid (7.5 µg/mL) | 50 ng/mL | 3.33 µL | |
Glutamax (100X) | 1X | 5 mL | |
Collagenase buffer | |||
NaCl | 0.100 g | Dissolve in 380 mL of ultrapure water. Adjust to pH 7.49 – 7.51 with NaOH. Top up to 400 mL with ultrapure water. Filter through 0.22 µm filter. Store at 4 °C. | |
KCl | 0.789 g | ||
Collagenase Type IV | 80 – 200 U/mL | ||
CaCl2·2H2O | 0.140 g | ||
HEPES | 4.800 g | ||
Calcium-free buffer | |||
KH2PO4 | 0.0815 g | Dissolve in 900 mL of ultrapure water. Adjust to pH 7.49-7.51 with HCl. Top up to 1 L with ultrapure water. Filter through 0.22 µm filter. | |
NaHCO3 | 1.0500 g | ||
NaCl | 3.4480 g | ||
KCl | 0.1750 g | ||
DMEM for cell isolation | |||
DMEM | 1 sachet | Top up to 1 L with ultrapure water. Filter through 0.22 µm filter. Store at 4 °C. | |
NaHCO3 | 1.5 g | ||
Penicillin-Streptomycin (100X) | 1X | 10 mL | |
Collagen overlay (1 mL) | |||
0.1 M NaOH | 1 part | 20.8 µL | Prepare fresh. Add collagen last. |
10X PBS | 1 part | 20.8 µL | |
1X PBS | 6 parts | 125 µL | |
Hepatocyte culture media | 32 parts | 667 µL | |
Type I bovine collagen | 8 parts | 167 µL | |
Collagen coating solution (1 mL) | |||
Type I bovine collagen | 1 part | 0.5 mL | Prepare fresh. |
0.01 N HCl | 1 part | 0.5 mL |
Table 1: Recipes for buffers and solutions.
Hepatocyte viability (%) | Hepatocyte yield (x 108 cells) | Viable hepatocyte yield (x 108 cells) |
91.4 ± 3.4 | 4.39 ± 1.58 | 4.04 ± 1.48 |
Table 2: Cell viability and yield. Values ± SD from 18 different experiments are shown.
Urea (µg/million cells/day) | Albumin (µg/million cells/day) | CYP1A2 activity (µg/million cells/day) | CYP2B1/2 activity (µg/million cells/day) | CYP3B2 activity (µg/million cells/day) | |
Day 1 | 124 ± 2.6 | 39.0 ± 3.6 | |||
Day 3 | 65.7 ± 3.8 | 516.0 ± 95.1 | 2249 ± 56 | 188 ± 49 | 93.7 ± 0.6 |
Table 3: Functional levels of isolated primary rat hepatocyte in sandwich culture. Values ± SD of urea, albumin, CYP1A2, CYP2B1/2, and CYP3B2 assays from three different experiments are shown.
There are a few points that are particularly important to observe for two-step collagenase perfusion procedure in general. Firstly, special care must be given when resecting the liver. Ensure that the gastrointestinal tract is not damaged as leakage of the contents will result in bacterial contamination. In addition, avoid damaging the Glisson's capsule, which covers the surface of the liver during the animal procedure. If the tear is large enough, it might allow premature release of disassociated hepatocytes into the collagenase buffer. Secondly, the ability of collagenase to digest the liver while ensuring good hepatocyte viability may vary by lot number. The amount of collagenase used must be optimized for every new lot of collagenase5. It is advisable to test a few lots of collagenase and select the best lot before buying in bulk. Ensure that the water is ultrapure, and the pH is correct. Impurities in the water such as ferrous and ferric ions and the wrong pH could affect collagenase activity, which will then drastically affect the quality of the isolated cells. Although collagenase buffer recirculation in this protocol can be optional, it is not advisable because the volume of collagenase buffer needed will increase to >400 mL. Lastly, perfusion pressure and flow rate should be in an appropriate range. High pressure and flow rate will lead to liver cell death while low pressure and flow rate will result in insufficient buffer flow through the smaller vessels10. During calcium-free buffer perfusion, the pressure will initially rise. However, upon collagenase buffer perfusion, the pressure will then slowly drop over time since digestion by collagenase reduces flow resistance. If the pressure drop was not observed, the digestion time can be extended as required. Since the IPS has the optional pressure sensor, it is possible to vary the flow rate of the perfusate to compensate for pressure variations. On another note, although flushing the circuit with an oxygenator was used in some protocols2,5, we found that it was not critical for hepatocyte viability and yield. Thus, our device does not include an oxygenator.
In contrast to the common practice of liver digestion before resection, this protocol involves resection of the liver prior to digestion. The approach in this protocol holds several advantages compared to the common practice. Firstly, over-digestion of the liver can be avoided. Depending on the surgeon's skill, liver resection may take anywhere from 5-10 min. Following the common practice, even though the perfusion has been stopped, this still equates to an additional 5-10 min where the liver is exposed to collagenase. Over-digestion may decrease cell viability and reduce hepatocyte attachment and function. Secondly, this approach could potentially reduce hepatocyte loss. The Glisson's capsule is less likely to tear when the liver is still springy before digestion, compared to when it is soft and mushy after digestion. Finally, resection of the liver prior to digestion enables the possibility of collagenase recirculation. Collagenase recirculation eliminates the occurrence of failed isolations due to insufficient volume of collagenase being prepared. Our approach allows digestion time to be extended, as necessary. It also has a minor advantage of reducing the amount of collagenase needed per isolation. The main disadvantage of this approach is that it introduces a new problem of catheter detachment halfway during perfusion, but which could be addressed with a few technical precautions. Another difference is that this protocol does not involve the use of density gradient such as Percoll. By removing dead cells, Percoll has been reported to increase the final viability of cells with some loss of yield, from 20%-50%16. Our cell viability is high because we test and select collagenase lots. From our experience, some collagenase lots consistently give lower viability. Without density centrifugation, it will be necessary to avoid bad collagenase lots through prior testing. For those who do not have such liberty and must consider density gradient, care should be given during Percoll dilution to avoid selection of hepatocyte subpopulations with a slightly different metabolic profile17,18.
Our approach in using an IPS for primary rat hepatocyte isolation resolves several issues with existing perfusion setups1,2,5,6,7,8,9. Usage of disposable sterile tubing saves time by removing the need to clean and disinfect perfusion tubing before and after each isolation. It reduces fire hazard since the perfusion tubing need not to be filled with flammable 70% ethanol when not in use to maintain sterility. The risk of contamination due to improper disinfection and storage can also be eliminated. Most importantly, the need for routine replacement of old tubing can be circumvented; this process is essential but is often complicated and time consuming. The addition of a bubble filter to the tubing near the catheter allows bubbles to escape before reaching the catheter and prevents air from passing through when the perfusion buffer runs out. Usage of a silicone heater jacket as the temperature control system allows precise and stable maintenance of perfusate temperature during digestion, in contrast with pre-warmed buffers which will significantly cool down over time. In addition, the silicone heater jacket is compact and can be easily maintained, unlike a glass jacket coupled to a water bath circulator. Usage of monitoring sensors for temperature and pressure replaces the optimization step before isolation. Temperature variation may occur during or between experiments. Pressure (and thus flow rate) may also change due to issues with perfusion tubing. Placement of sensors near the cannula allows more accurate reading of the temperature of buffer perfusing into the liver. Sensor alarms can remind the user to take corrective action. The recorded log also allows for troubleshooting. Usage of a special IV catheter for cannulation simplifies cannulation of the portal vein. Unlike other types of IV catheter, the catheter described here allows fluid communication between the cannula and the tubing in both its puncturing and retracted needle position due to a hole at the side of the needle. Thus, surgeons could use color change of the liver as a sign of successful cannulation. Upon successful cannulation, the needle tip can be retracted behind the catheter tip to prevent the re-piercing of the portal vein. The retracted needle could also act as a foundation for the ligature to secure the portal vein onto the soft cannula and prevent it from being deformed. This catheter can be operated with one hand; therefore, the non-dominant hand can be used to support the portal vein with forceps. The compact integrated design allows the whole perfusion device to be put into the laminar hood, reducing the risk of contamination, and saving space. This is especially important if the animal procedure is to be performed in a dedicated animal facility where space availability can be a challenge.
In comparison to previously reported rat hepatocyte isolation protocols using the two-step collagenase perfusion procedure1, the conditions described consistently generate higher cell yield of up to 5 x 108 hepatocytes per isolation, cell viability between 88%-94.8% and good hepatocyte-specific function.
This protocol can be concurrently used to isolate other non-parenchymal liver cells such as Kupffer cells, liver endothelial cells and hepatic stellate cells from the same rat. This protocol can also be modified for the isolation of liver cells from mice by using a smaller gauge IV catheter for cannulation and decreasing the perfusion flow rate to a range suitable for mouse species6. The perfusion procedure for mice can be simplified by resecting the liver after digestion.
The IPS device used in this experiment can be also used for ex vivo liver perfusion application such as experiments on liver transplantation as well as decellularization of rodent and discarded human livers19,20. The IPS has an advantage where the liver must be perfused for long periods because the IPS can monitor perfusion conditions in real-time and allow discontinuation of the perfusion upon completion either automatically or manually by remote control.
The authors have nothing to disclose.
This work is supported in part by MOE ARC (MOE2017-T2-1-149); NUHS Innovation Seed Grant 2017 (NUHSRO/2017/051/InnovSeed/02); Mechanobiology Institute of Singapore (R-714-106-004-135); and Institute of Bioengineering and Nanotechnology, Biomedical Research Council, Agency for Science, Technology and Research (A*STAR) (Project Numbers IAF-PP H18/01/a0/014, IAF-PP H18/01/a0/K14 and MedCaP-LOA-18-02) funding to Hanry Yu. Ng Chan Way is a research scholar of the National University of Singapore. We would like to thank Confocal Microscopy Unit & Flow Cytometry Unit of the National University of Singapore for help and advice in hepatocyte purity analysis.
Material/Equipment | |||
1 mL syringe | Nipro | ||
27G needle | Nipro | ||
Black braided silk non-absorbable, non-sterile surgical suture | Look | SP117 | |
Bochem 18/10 stainless steel forceps, sharp tip contain bent round tip | Bochem | 10333511 | |
Disposable Perfusion Set | Vasinfuse | BPF-112 | |
Floating circular 1.5 mL microcentrifuge tube rack | Sigma-Aldrich | R3133 | |
German Standard Tissue Forceps, Serrated / 1×2 teeth , 14.5cm | Walentech | ||
Greiner Cellstar aspirating pipette | Merck | GN710183 | |
Haemocytometer | |||
Integrated Perfusion System | Vasinfuse | IPS-001 | |
Iris Scissors curved, stainless, 11cm | Optimal Medical Products Pte Ltd | CVD | |
Light microscope with 10X lens | Olympus | ||
Mesh Sheet 100µM Nylon | Spectra-Teknic(s) Pte Ltd | 06630-75 | |
Operating Scissors, BL/BL, 13cm | Optimal Medical Products Pte Ltd | STR – BL/BL | |
Operating Scissors, SH/BL, 13cm | Optimal Medical Products Pte Ltd | STR – SH/BL | |
Reverse force hemostatic clip | Shanghai Jin Zhong Pte Ltd | XEC230 | |
Water bath | Grant | ||
Reagents/Chemicals | |||
10X Phosphate buffered saline (PBS) | Sigma-Aldrich | ||
Bovine serum albumin (BSA) | Sigma-Aldrich | A9056 | |
CaCl2·2H2O | Merck | 137101 | |
Collagenase Type IV | Gibco | 17104019 | |
Dexamethasone | TCI | D1961 | |
DMEM | Gibco | 31600-034 | |
Glutamax | Gibco | 35050061 | |
HEPES | Invitrogen | 11344-041 | |
Insulin | Sigma-Aldrich | 1-9278 | |
KCl | VWR | VWRC26764.298 | |
KH2PO4 | Sigma-Aldrich | P5379 | |
Linoleic acid | Sigma-Aldrich | L9530 | |
NaCl | Sigma-Aldrich | S5886 | |
NaHCO3 | Sigma-Aldrich | S8875 | |
NaOH | Merck | 106462 | |
Penicillin-Streptomycin | Sigma-Aldrich | P4333 | |
Type I bovine collagen | Advanced BioMatrix | 5005-100ml | |
William’s E Media | Sigma-Aldrich | W1878 |