The goal of this protocol is to obtain high viability and high yield of hepatocytes and sinusoidal endothelial cells from liver. This is accomplished by perfusing the liver with a type IV collagenase solution via the portal vein, followed by differential centrifugation to obtain hepatocytes and sinusoidal endothelial cells.
This protocol demonstrates a method for obtaining high yield and viability for mouse hepatocytes and sinusoidal endothelial cells (SECs) suitable for culturing or for obtaining cell lysates. In this protocol, the portal vein is used as the site for catheterization, rather than the vena cava, as this limits contamination of other possible cell types in the final liver preparation. No special instrumentation is required throughout the procedure. A water bath is used as a source of heat to maintain the temperature of all the buffers and solutions. A standard peristaltic pump is used to drive the fluid, and a refrigerated table-top centrifuge is required for the centrifugation procedures. The only limitation of this technique is the placement of the catheter within the portal vein, which is challenging on some of the mice in the 18 – 25 g size range. An advantage of this technique is that only one vein is utilized for the perfusion and the access to the vein is quick, which minimizes ischemia and reperfusion of the liver that reduces hepatic cell viability. Another advantage to this protocol is that it is easy to distinguish live from dead hepatocytes by eyesight due to the difference in cellular density during the centrifugation steps. Cells from this protocol may be used in cell culture for any downstream application as well as processed for any biochemical assessment.
Collagenase perfusion of the liver to obtain hepatocytes has been performed since the early 1950s and has been continually improved upon1,2,3,4. A very nice review of many of the methodologies, techniques, and reagents used in liver cell purification has been compiled in Meyer et al5. Obtaining a satisfactory yield of highly viable hepatocytes and SECs is technically challenging. The mechanical forces that separate the cells including the quality of collagenase are some of the variables that are difficult to control. Due to hepatocyte sensitivity to mechanical forces, their viability is markedly reduced in sub-optimal conditions, prompting the need for a protocol describing optimal conditions for isolation. SECs appear not to be as sensitive to mechanical shearing. In situ perfusion with collagenase digestion is by far the best method to disrupt cell-cell junctions to obtain single cell preparations from liver and other organs such as intestine and spleen6. This protocol demonstrates a simple method for introducing perfusion buffer into the portal vein by using a retractable plastic catheter rather than an incision that could cause the portal vein to collapse, as described in Smedsrød et al7.
The goal of this manuscript is to demonstrate the most critical steps required for success in the liver profusion procedure. These steps include catheter placement, flow of the perfusion liquids, and handling of the tissue after digestion. Flow rates are adjusted above the natural rate of blood flow, but low enough to keep the Glisson's capsule intact. Once the liver is properly digested and cells are separated in solution, cell purification is relatively simple whether it is performed by differential centrifugation, plate adhesion, or by magnetic bead purification. Live hepatocytes have a higher density and are easily purified from the nonparenchymal cells (NPCs) and dead hepatocytes with slow speed centrifugation. For most applications, plate adhesion for separating Kupffer cells (KCs) and SECs is a common method8, though there are reports that it does not produce the best purity of SECs9. KCs have a tendency to adhere to solid rigid surfaces quickly and Petri dishes (standard polystyrene) are the most commonly used material for this procedure. SEC or KC purification with the use of antibody-conjugated magnetic beads is undoubtedly the best method for significant purification of these cells, though the procedure adds another 3 – 4 h onto this protocol and the overall yield is decreased9. This report demonstrates in detail the process for an optimal liver perfusion that typically yields a high number of viable cells.
All animal procedures outlined in this protocol have been approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Lincoln – Nebraska under protocol #1435.
1. Preparation
2. Animal Procedure
3. Liver Perfusion
4. Hepatocyte Purification
NOTE: Manipulation of the cells are performed in a sterile tissue culture hood to limit contamination if the cells are to be cultured in all subsequent steps.
5. SEC Purification
A demonstrational, refined method for liver perfusion and purification/enrichment of hepatocytes and SECs is presented here that provides detailed tips for optimal cell purification/enrichment similar to other printed reports compiled in this referenced review 5. The critical steps which determine success for cellular purification occur in the route and technical details of the collagenase perfusion procedure and are outlined in this protocol. The set-up for the apparatus is fairly simple and cost effective with standard laboratory equipment, in contrast to other systems that have been published11 (Figure 1). In this set-up, the tray holding the mouse is sitting on the water bath with some of the excess tubing in the water bath to maintain the temperature of the liquids perfusing within the liver. The apparatus as shown here may be used for mice as well as rats, with a slightly different geometry for rat liver perfusion12 (Figure 2).
In this procedure, the liver is perfused via the portal vein instead of the vena cava (which is another popular perfusion route) due to its ease of access within the abdomen and that the vein feeds directly into the liver. The viewer should be aware that the portal vein has several small branches which may short-circuit the perfusate, and the catheter should be placed past these branches for optimal success13 (Figure 3). Once the portal vein is identified, curved forceps are used to draw a surgical suture or polyester thread beneath the portal vein between the liver and the superior pancreatic-duodenal vein. The forceps are also used to straighten out the portal vein which is under positive blood pressure (Figure 4A). The needle within the catheter should be placed parallel and next to the vein (Figure 4B) and the bevel should be gently inserted into the lumen of the vein (Figure 4C). Proper placement will be indicated by blood appearing along the catheter. Once the needle is retracted the catheter should be stabilized with a thread tied down upon it and the blood pressure will force blood up through the catheter. It is recommended that the catheter be connected to the pump tubing before blood spills out (Figure 4D). Once the pump tubing is connected, immediately cut a major blood vessel such as the vena cava or aorta abdominalis for blood/fluid drainage (Figure 5A). It is usually necessary to cut the side of the abdominal wall of the mouse for sufficient drainage, as a buildup of blood in the abdomen makes it difficult to visualize the perfusion process, and blood may carry over into the cell purification (Figure 5B). Once the PBS begins to perfuse the liver, the liver will blanch with the absence of blood (Figure 5C). If only a few of the lobes blanch, then the likely cause is that the catheter has been placed too far inside the liver and will need to be backed out slowly. Once placement of the catheter is confirmed, use water-resistant masking tape to secure the tubing in place. General laboratory tape is usually not sufficient. To confirm proper placement of the catheter, squeeze the cut blood vessel to cease drainage and observe the increase pressure within the liver. Doing this will aid in flushing blood out of the liver (Figure 5D). This should also be done when the collagenase solution initially goes into the liver to ensure that all lobes are exposed to collagenase. After the collagenase solution runs out, a good collagenase digestion is indicative by separation of the Glisson's capsule from the parenchyma.
The general procedure for cellular purification is outlined in Figure 6 and Figure 7 in which hepatocytes are harvested early on in the procedure during the low speed spins and are very pure after 4 washes in BSA-containing buffers (Figure 8). SECs are co-purified with KCs on a PVP gradient (Figure 9A) and then separated by selective adhesion on a collagen-coated polystyrene Petri dish. The purity of the SECs using this method is 83 – 90% pure and largely depends on the overall effectiveness of the collagenase digestion (Figure 9B). Quantitative measurements using light microscopy (as both hepatocytes and SECs have distinguishing features from other cells) show that in a representative preparation, hepatocytes are nearly 100% pure and SECs are just over 89% pure (Table 1). Alternatively, a higher enrichment of SECs may be obtained by magnetic column separation after the PVP gradient, although that is beyond the scope of this protocol. More information about magnetic separation of SECs and KCs may be found in Meyer et al.9 and Liu et al.14 It should be remembered that viability of hepatocytes rapidly decreases in an insufficiently digested liver, but this is not necessary true for the NPCs. Insufficiently digested livers also produce more cellular debris, which is not entirely eliminated in the centrifugation steps.
Figure 1: Schematic representation of the perfusion suite. The flasks are held in place by clamps (not shown); the tubing containing fluid is swapped from the 1L flask to the 125 mL flask during the procedure is represented by the red dotted line. Note, that the oxygen feed does not bubble directly into the solutions in the flasks. The corks should also be notched to allow a closed-circuit fluid flow with the tubing inserted within the notch. This is critical during warm-up of the tubing and for ejection of all air within the tubing. The air bleed is composed of the T-connector which is connected to Tygon tubing and pinch clamps for quick open and closure of the system. Please click here to view a larger version of this figure.
Figure 2: The perfusion suite during mouse liver perfusion. The tray with the mouse is placed on the corner of the water bath. The oxygen is disconnected from the collagenase solution as that was already oxygenated during the warm-up. Location of items are A) oxygen lines, B) 1 L flask containing PBS, C) 125 mL flask containing buffer 2 with collagenase, D) pump, E) pump controls, F) bubble trap, G) fluid line running from the flask, through the pump and to the mouse. Please click here to view a larger version of this figure.
Figure 3: Image of the vasculature of the mouse abdomen. Insertion of the catheter should be just below the gastric and pancreatic vein junctions coming off the portal vein (green dot) and the tip should be placed near the left and right hepatic portal veins (yellow dot) which forms a fork into the main lobes of the liver. Once correct placement is achieved, the catheter should be stabilized with thread using a simple overhand knot. K = kidney, L = liver, SI = small intestine. Please click here to view a larger version of this figure.
Figure 4: Catheter placement in the portal vein. (A) The forceps are used to ensure blood engorgement of the portal vein and to straighten out the vein for catheter placement. (B) The catheter is lined up in parallel with the portal vein with the bevel up. (C) The bevel of the needle within the catheter is inserted into the vein, not through the vein. (D) The needle of the catheter is retracted and blood will backflow through the catheter as indicated by the tip of the forceps. Please click here to view a larger version of this figure.
Figure 5: Proper liver perfusion. (A) After catheter placement, the female Luer end of the catheter is connecting to the male Luer end (cut from a 1 mL syringe) of the pump tubing. (B) After cutting one of the major descending blood vessels, blood and PBS are drained from the abdomen by slicing the side of mouse with scissors. (C) The liver should blanch while the blood is flushed out and (D) will swell when under pressure by squeezing shut the cut blood vessel with forceps. Please click here to view a larger version of this figure.
Figure 6: Schematic outline of hepatocyte purification and NPC isolation. Most of the centrifugation steps aim to remove live and dead hepatocytes from the non-parenchymal cells. Please click here to view a larger version of this figure.
Figure 7: Schematic outline of SEC enrichment and purification. NPCs are separated by the PVP gradient followed by SEC separation by short adhesion to a standard polystyrene plate. Please click here to view a larger version of this figure.
Figure 8: Purified hepatocytes. Hepatocytes were plated on collagen coated tissue culture plates with DMEM + 8% FBS + Pen/Strep and incubated for 6 h followed by image collection by a microscope at 400X. Bar equals 20 µm. Please click here to view a larger version of this figure.
Figure 9: SEC purification. (A) SECs and KCs were collected from the 25/50% PVP interface (yellow arrows) and washed with serum-free medium. (B) The SECs were separated from the KCs by selective adhesion on standard Petri dishes, washed, and plated on collagen-coated tissue culture dishes in RPMI + 5% FBS. Images were taken by an EVOS inverted microscope at 400X. Bar equals 20 µm. Please click here to view a larger version of this figure.
Figure 10: Scanning electron microscopy of mouse liver anatomy. A mouse liver was perfused with PBS followed by fixative (100 mM sodium cacodylate, 12 mL 25% glutaraldehyde, 15.6 mL 16% paraformaldehyde, 2.65 mM calcium chloride, 180.2 mM sucrose, mixed in 100 mL solution) at a rate of 1 mL/min for 4 min. Tissues were sliced and prepared for scanning electron microscopy. Images were collected on a Field Emission SEM at 10,000X. Yellow arrows indicate the microvilli between the hepatocytes. Bar equals 5 µm. Please click here to view a larger version of this figure.
Figure 11: Visualization of dead versus live hepatocytes. After 2 washes with Buffer 1, pelleted hepatocytes may be assessed for live/dead ratio by eye. A lighter pellet (left) indicates that at least half the hepatocytes are dead. The darker pellet on the right is more firmly pelleted to the bottom of the tube and is more than 90% live. Please click here to view a larger version of this figure.
Table 1: Cell purification | ||
Cell Type | % target cells | % other cells |
Hepatocytes | 99 | 1 |
Sinsusoidal endothelial cells | 89.1 | 10.9 |
Table 1: Assessment of cell purity by light microscopy.
This protocol highlights the tools that are available and that will afford the user a high rate of success in this procedure. A successful perfusion is fundamental for any downstream application when working with primary cells.
There are a few critical steps of the procedure that determine success. First, placement of the catheter tip within the portal vein must be correct. If placed too far inside the liver, only the minor lobes will be perfused. The tip should be just below the right and left branches of the portal vein. The advantage of using a polymer-composite catheter over a needle is that the barb of the needle is more likely to tear the vein during the perfusion than a catheter. Second, the collagenase quality and quantity determine the digestion efficiency of the liver. In this procedure, pre-qualified collagenase Type 4 was used and it is resuspended at 0.5 mg/mL in Buffer 2 if the collagenase activity to be assayed is greater than 900 IU.
At the end of the collagenase perfusion, the liver should retain a somewhat dark tan to brown color. If it is a light tan color, then most of the hepatocytes are dead. The liver should be falling apart when cut from the mouse. In the best conditions, the liver may be scooped out of the body cavity of the mouse with a small spoon. Livers in this condition always give very high cell viability. If it takes much effort to shake the cells apart, if the liver is still firm during extraction, or if there is a lot of force required to move the hepatocytes through the filters, then the hepatocytes will have a lower viability. Hepatocytes are covered with microvilli, which allows them to have a very high surface area (Figure 10). Incomplete digestion retains cell-cell junctions between hepatocytes, and mechanical shearing will tear the plasma membrane apart. In situ perfusion with collagenase is the best method to break apart the cells and maintain high viability. In Figure 11, two hepatocyte preparations are made in which the cell pellets were compared after a few washes in Buffer 1. The pellet on the left is lighter and contains a cell viability of about 50%. The pellet on the right is darker and has a viability of 92%. Similarly, when the liquid is poured from the 50 mL conical, the darker pellet is stationary within the tube, in contrast to the lighter pellet, which will slide in the tube as the liquid is poured off. Lower cell viability or inefficient perfusions may occur when the liver has high levels of sclerosis.
Since live hepatocytes have a higher density than dead hepatocytes, the centrifugation procedures will result in a preparation that is highly pure in viable hepatocytes15. If there is a significant amount of dead hepatocytes (which often occurs when conditions are suboptimal), live cells may be further enriched and separated from dead cells using PVP gradients. Additionally, since live hepatocytes pellet faster than dead hepatocytes, aspiration of the top 3rd of the pellet will also increase the ratio of live to dead cells in the pellet. These procedures are quick and easy methods to separate live from dead hepatocytes and cell debris when needed10,16. This is particularly useful if the sample is precious, and only a few million cells are required for the experiment.
In conclusion, this is an easy and efficient method for harvesting hepatocytes and SECs from the liver. At current prices, the cost of performing this procedure including all reagents and disposables is under 75 USD per preparation. If multiple mice are desired, it is best to proceed with the hepatocyte purification and keep the NPC fraction on ice until all the mice have been processed. NPCs are generally stable on ice for at least 5 h, but longer times have not been tested in this laboratory.
The authors have nothing to disclose.
Funding is provided in part by the NIH from grant R01HL130864.
1 L Erlenmeyer flask | Fisher Scientific | S63274 | |
250 mL Erlenmeyer flask | Fisher Scientific | S63271 | |
silicone tubing | Cole-Parmer | 96400-14 | This tubing runs from the flasks through the pump to the T connector and then to the 1.0 mL syringe that is connected to the catheter. |
Tygon tubing | Fisher Scientific | R3603 | Used as an adaptor between 96400-14 and pipettes and T connector. This may also be used for the oxygen tubing. |
T-connectors | Cole-Parmer | EW-06294-82 | |
Quick dissconnects | Fisher Scientific | 6150-0010 | |
Pinch Clamps | Fisher Scientific | 6165-0002 | |
Masterflex L/S Variable speed Pump, model 7553-70 | Cole-Parmer | EW-07559-00 | Periplastic pump with variable speed |
Pump head, model 7014-20 | Cole-Parmer | EW-07014-20 | |
glass graduated 1.0 mL pipettes | Fisher Scientific | 13-678 | |
curved non-serrated scissors | Fine Science Tools | 14069-12 | |
Dumont forceps | Fine Science Tools | 11252-20 | |
Curved forceps | Fine Science Tools | 13009-12 | |
10 mL syringe | Fisher Scientific | 03-377-23 | Only barrel of syringe will be needed |
sterlized spoon | Home supply store | ||
Cotton ball(s) | Home supply store | ||
Polyester sewing thread | Home supply store | ||
Masking tape | Home supply store | ||
thumb tacks | Home supply store | ||
styrofoam pad | 50 mL conical rack | ||
cookie/baking sheet | Home supply store | ||
Absorbant underpads | Fisher Scientific | 14-206-64 | |
19 L water bath | Fisher Scientific | TSCOL19 | |
BD Insyte Autogaurd Shielded IV Catheter 24 guage | Becton Dickinson | 381412 | Plastic cathetar with retractable needle |
Crystallizing dishes 100×50 | VWR | 89000-290 | |
Polystyrene petri dishes | Sigma Aldrich | P5481-500EA | |
50 mL conical tubes | Fisher Scientific | 12-565-270 | |
graduated pipettes (5 mL) | Fisher Scientific | 170355 | |
graduated pipettes (25 mL) | Fisher Scientific | 170357 | |
EasyStrainer 100 μM | Greiner bio-one | 542000 | 100 μm filter |
EasyStrainer 40 μM | Greiner bio-one | 542040 | 40μm filter |
Sterile transfer pipettes | Fisher Scientific | 13-711-20 | |
Refrigerated swinging bucket centrifuge | Sorvall Legend XTR | 75-217-406 | Centrifuge with swinging bucket rotar |
Galaxy 170R tissue culture incubator | Eppendorf | CO170R-120-0000 | Humidified tissue culture incubator |
Reagents | |||
Name | Compound | Grams (g/L) | Millimolar (mM) |
Buffer 1, pH 7.4 | NaCl | 8.3 | 142 |
KCl | 0.5 | 6.7 | |
HEPES | 2.4 | 10 | |
BSA | 15 | 0.226 | |
Buffer 2, pH 7.4 | NaCl | 3.9 | 66.74 |
KCl | 0.5 | 6.71 | |
CaCl2 | 0.7 | 6.31 | |
HEPES | 24 | 100 | |
BSA | 15 | 0.226 | |
Phenol Red | 0.01 | 0.03 | |
Buffer 3, pH 7.4 | NaCl | 8 | 137 |
KCl | 0.35 | 4.7 | |
MgSO4 | 0.08 | 0.66 | |
CaCl2 | 0.18 | 1.62 | |
HEPES | 2.4 | 10 | |
BSA | 15 | 0.226 | |
PBS, pH 7.4 | NaCl | 8 | 137 |
KCl | 0.2 | 2.7 | |
Na2HPO4-7H2O | 1.15 | 4.3 | |
KH2PO4 | 0.2 | 1.4 | |
Other reagents | |||
Name | Company | Catalog Number | Comments |
Percoll (PVP solution) | GE Healthcare | 288555 | |
Collagenase Type IV | Sigma Aldrich | C5138 | |
Isoflurane | Abcam | ab144581 | |
Hepatocyte Growth Medium | |||
DMEM | Gibco | 11965118 | |
10% by volume fetal bovine serum | Gibco | 10437010 | |
Pen/Strep (2x) | Gibco | 15140148 | |
Long-term Hepatocyte Growth Medium | |||
DMEM | |||
10% by volume fetal bovine serum | Gibco | 10437010 | |
Pen/Strep (2x) | Gibco | 15140148 | |
Glucagon | Sigma | G3157-2mg | 14 ng/mL |
Insulin | Sigma | I9278-5mL | 0.5 U/mL |
Hydrocortisone | Sigma | H0888-1g | 7.5 mg/mL |
Epidermal Growth Factor | BD Biosciences | 354001-100ug | 20 ng/mL |
LSEC medium | |||
RPMI | Gibco | 11875119 | |
10% by volume fetal bovine serum | Gibco | 10437010 | |
Pen/Strep (2x) | Gibco | 15140148 |