The protocol describes a straightforward method of resectioning an intact mouse liver for metabolic studies through portal vein perfusion.
Metabolic diseases such as diabetes, pre-diabetes, non-alcoholic fatty liver disease (NAFLD), and nonalcoholic steatohepatitis (NASH) are becoming increasingly common. Ex vivo liver perfusions allow for a comprehensive analysis of liver metabolism using nuclear magnetic resonance (NMR), in nutritional conditions that can be rigorously controlled. As in silico simulations remain a primarily theoretical means of assessing hormone actions and the effects of pharmaceutical intervention, the perfused liver remains one of the most valuable test beds for understanding hepatic metabolism. As these studies guide basic insights into hepatic physiology, results must be accurate and reproducible. The greatest factor in the reproducibility of ex vivo hepatic perfusion is the quality of surgery. Therefore, we have introduced an organized and streamlined method to perform ex vivo mouse liver perfusions in the context of in situ NMR experiments. We also describe a unique application and discuss common issues encountered in these studies. The overall purpose is to provide an uncomplicated guide to a technique we have refined over several years that we deem the golden standard for obtaining reproducible results in hepatic resections and perfusions in the context of in situ NMR experiments. The distance to the center of the field for the magnet as well as the inaccessibility of the tissue to intervention during the NMR experiment makes our methods novel.
Ex vivo perfusions are crucial in the study of hepatic metabolism, and perfusion through the portal vein is the standard for these studies. In order to study hepatic metabolism in isolation, the liver must be resected from the body to avoid complications arising from metabolism in other organs (i.e., whole-body metabolism) and to exert control over hormone availability (insulin, glucagon, etc.). This approach can be essential for understanding the effects of diseases such as diabetes, NAFLD, and NASH on hepatic metabolism as well as mechanisms of drug action. This article serves as a guide to hepatic resection and perfusion. We have developed a streamlined procedure to perform these metabolic liver studies with sufficient rigor and reproducibility. If the surgery is not performed correctly, there is pronounced variability in the metabolic data obtained. We describe an organized method to perform portal vein catheterization and liver resection in the context of metabolic studies in situ in a nuclear magnetic resonance (NMR) spectrometer, as described in the literature1,2,3,4,5.
Currently, there is no literature describing an ex vivo hepatic perfusion using a glass column within an NMR. Nor is there a video or text publication providing a clear example of how to perform the procedure with the mouse liver, specifically, demonstrating how to catheterize the portal vein, resect a liver, transfer, and hang the liver onto a glass column. As the genetically modified mouse is ubiquitously used for studying liver metabolism, this is an essential procedure that deserves a complete description. Liver perfusion surgeries are not new, but this article is a gold standard method accompanied by a video demonstrating the technical excellence described in this paper to aid everyone interested in this procedure. The method presented here would be best applied to real-time metabolism to detect the function and turnover of metabolites in disease models.
This method uses a 100 cm water-jacketed glass column, which allows the liver to hang at the bottom of the cannula encapsulated by perfusate inside an NMR tube. Heated water in the glass jacket is used to control perfusate temperature. A thin layer oxygenator is pressurized with 95%/5% O2/CO2 for pH control. By using three separate pumps, the perfusate column height is set, which provides constant pressure to the liver. Flow rates are not controlled beyond the application of constant pressure (Figure 1). To confirm the liver is appropriately functioning, oxygen measurements are taken along with flow rates. In our hands, this set of preconditions leads to highly repeatable NMR experiments for the assessment of liver metabolic function.
Experiments involving mice were handled in compliance with the University of Florida Institutional Animal Care and Use Committee (protocol number #201909320). The mouse strain used was C57BL/6J; all mice were male. This method is generally applicable for studies using other standard mouse strains as well. This surgery is optimally performed by two individuals working together.
1. Initial set-up
2. Pre-surgery set-up
3. Perfusate column set-up
4. Anesthetization of the mouse
5. Celiotomy
6. Cannulation of portal vein
7. Resection of liver post portal vein cannulation
8. Hanging the liver from the column
9. Flow measurement
10. Oxygen Measurement
NOTE: Oxygen meter measurements were set up according to the manufacturer's instructions15.
Liver function is primarily assessed by oxygen consumption and flow rate. A flow rate of 4-8 mL/min and oxygen consumption of 1 µmol/min.g is typical. These measures will vary depending on specific experimental conditions and biological differences.
The exact amount of isoflurane used will depend on the type of anesthesia system being used as well as the environment and age/weight of the mouse. During the surgery, the isoflurane and delivery gas do not change, although some changes may be necessary depending on the specifics of the surgical area (e.g., background noise)10. When heparin is injected deep subcutaneously, the onset of action could be delayed by up to 20-40 min. A 10 min waiting period post heparin administration ensures onset of action16. Lidocaine has a 2 min onset of action11.
When inserting the catheter, keep the bevel pointed up, and enter at no more than 15° angle from the portal vein. Both sutures have two knots. The first suture must be tied past the catheter tapper. If cannulation of the portal vein is successful, the liver bleaches from the flush. As the surgeon is resecting the liver, the assistant clears the resected contents with a cotton-tipped applicator. To prevent contaminating the liver and prevent nicks to the lobes, do not cut through the stomach. Do not apply too much tension to the portal vein or liver when holding the catheter to prevent dislodging or tearing the portal vein.
The perfusion hardware setup requires extensive attention to detail (Figure 1). Heparin injections (Figure 2) are essential to the experiment. If blood coagulates, it will occlude the catheter that is inserted into the portal vein, preventing flow. The lidocaine injection (Figure 3) is to aid in desensitizing the area for pain relief. Table 1 provides a simple dosing chart for heparin and lidocaine with saline. The celiotomy and suturing (Figure 4) are essential for a successful portal vein catheterization, liver resection, and successful transfer to the perfusion rig. Flow rate and oxygen consumption measurements are vital to monitoring the liver's health and function (Figure 6). There is often a slight difference in O2 consumption between fed and fasted livers, which we attribute to increased energy demands imposed by gluconeogenesis in the fasted liver.
Figure 1. Perfusion column and pumps. A. A 100 cm water-jacketed glass column in which the liver hangs at the bottom. B. The water pump circulates water through the glass column and heats the perfusate. C. Glass thin layer oxygenator is pressurized with 95%/5% O2/CO2 oxygenating the perfusate. D. The ball-bearing pump circulates perfusate from the water bath into the thin layer oxygenator and the glass column. E. The ball-bearing pump circulates perfusate being delivered into the column from the delivery pump keeping perfusate oxygenated and maintaining a flow of 8 mL/min. F. The ball-bearing pump removes efferent perfusate from the NMR tube. G. Weighing scale to weigh perfusate from NMR tube to obtain flow rate of the liver. Please click here to view a larger version of this figure.
Figure 2. Heparin injection. Deep subcutaneous injection of heparin is given in the lower abdominal fat layer of the mouse. It is important when picking up the mouse, to pull the skin tight to allow the needle to penetrate the skin with ease. Please click here to view a larger version of this figure.
Figure 3. Lidocaine injection. The mouse is placed in the supine position on the surgical platform with its paws taped down and the nose in the nose cone. Lidocaine is administered subcutaneously in the iliac crest region. Please click here to view a larger version of this figure.
Figure 4. Celiotomy and suture. The celiotomy exposes the internal organs, and a hemostat pulls traction through the xiphoid process to help open the incision further. Two sutures are placed around the portal vein, the catheter is inserted, and the sutures are tied. Please click here to view a larger version of this figure.
Figure 5. NMR tube. The liver removed from the body along with the catheter which is then attached to the silicon tubing attached to the glass column. The liver is hung from the column and encapsulated by the NMR tube. A 20 mm NMR tube is then carefully screwed onto the column encapsulating the liver. Please click here to view a larger version of this figure.
Figure 6. Oxygen consumption and flow rate. Representative data from comparing hepatic oxygen consumption and flow rate measurements between fed and fasted livers. N = 3 and error bars are standard deviation. Please click here to view a larger version of this figure.
Heparin 1000 units/mL | Saline 0.9% | Total |
0.01 mL | 0.19 mL | 0.2 mL |
Lidocaine 2% | Saline 0.9% | Total |
0.2 mL | 0.6 mL | 0.8 mL |
Table 1. Heparin and lidocaine dose with saline. The table displays the concentration of heparin and lidocaine and the dose of each pharmaceutical with saline.
This surgical procedure is challenging and requires extensive practice to achieve reproducible results. Isoflurane and carrier gas should be adjusted as needed to maintain the viability of the animal through as much of the surgical procedure as possible. Environment, time of day, age, weight, and several other factors will affect anesthesia. Weight, diet, strain of mice and age could affect surgery as fat buildup can interfere with visualizing the portal vein. When taping the paws down, care must be taken to not apply any strain on the neck that may result in suffocation. Furthermore, the fattier the mouse the tighter the suture will need to be around the catheter to counteract the decreased coefficient of friction between the catheter and vein induced by the lipids. The onset of action time for heparin is essential as excessive exposure to isoflurane produces artifacts in organs17. Administration of heparin and lidocaine injections require a 23G needle 19.05 mm length, ensuring no trauma to internal organs during the injection. The suture loop must be over the taper of the catheter, or it will occlude the vein when tightened. Once the catheter is inserted there is usually a backflow of blood out of the catheter from the pressure, which is a positive sign of correct placement. The catheter may be pulled upwards for visual confirmation that the catheter tip is far enough past the first suture. Rolling the wrists and shoulders will ensure that the suture will not slide off the tapered tip when the assistant ties the suture. When transferring the liver from the silicone perfusion tubing to the column a bead of perfusate is left on top of the catheter. The perfusate bead will prevent any air bubbles from entering the catheter and going into the liver. The meniscus of perfusate on top of the catheter provides sufficient volume for the liver to function until its connected. To avoid torsion of the liver and the portal vein, the NMR tube is slowly screwed on. Additionally, the perfusion system used in this experiment does not require a pressure valve. The flow rates of the pumps are maintained such that glass perfusions column contains perfusate at a height of ~12 cm. The liver takes up Krebs buffer by gravity, negating any pressure difference between the two pumps with no effect on the flow rate of the liver. Since the liver is perfused through gravity the amount of perfusate taken up by the liver is set by the liver's natural biological activity. No data was collected for perfusion pressure as portal vein pressure is not measurable in this system.
The first incision of the celiotomy is shallow to create an opening, and subsequent cuts are deeper to avoid nicking the lobes of the liver. The overall length of the incision for the celiotomy is 3 cm for mice of this age and size but will change based on strain, age, and weight. Although the study described uses mice 9-13 weeks old, older, or younger mice can be studied as well as rats. The size of the NMR tube and the portal vein catheter would need to be changed based on the anatomical size of the liver and portal vein for the study of concern. If the portal vein is not straight, a cotton-tipped applicator can aid in manipulating the vein when inserting the catheter. Although the bile duct is not removed, if there is an experimental need for its absence the duct can be removed with fine tweezers. Excessive manipulation of the portal vein when placing sutures will cause constriction making catheter placement more difficult. Any fur that sticks to the liver is rinsed off with perfusate before press fitting the catheter to the column and encapsulating it within the NMR tube. The time frame of 30 min for perfusion can be changed from 20 min to 60 min, but all reliable data will be collected after the initial 10 min of perfusion.
A marker of successful hepatic tissue resection is after the perfusion the liver has no deformities or any other integrity issues. It is homogeneously pale yellow throughout. If the tissue was injured during surgery such as a nick it would have dark yellow spots around it. Also, if the liver was damaged from the perfusion, it would not perfuse. If the tissue experienced poor perfusion of the Krebs buffer, there would be dark yellow streaks throughout the organ as a result of starvation leading to tissue death. Another method of monitoring liver health is hepatic oxygen consumption (Figure 6). It has been shown that mice livers contain higher lipid and glycogen amounts but have similar total protein amounts, so it was expected that the hepatic oxygen consumption when normalized to liver mass would have a similar value. A third method is the NMR data of the real-time data of metabolic turnover.
The method's main limitation is terminal surgery itself. There is a substantial cost in mice, equipment, time, and personnel. Therefore, utmost care must be exercised when performing these procedures and collecting data. Biological variation within the mouse model can generate difficulty in surgery. Moreover, it is imperative to avoid optical enhancements. No optical enhancements are needed as all anatomy is visible to the naked eye. Optical enhancements increase the potential for errors to occur as the surgeon and assistant have a limited field of view, leading to knicks in the liver or unintended tension on the portal vein, causing a failure if the catheter pulls out. Proper implementation of these methods will result in > 95% surgery success rate in the C57BL/6J mouse. Another limitation to consider is the 10 min period required for the liver to reach a steady state. It is not a limitation for the study described here, nor in many others, but for any experiment warranting the initial 10 min of data this method will not suffice. The lack of the complex hormonal signature associated with whole-body metabolism also serves as a limitation, though glucagon, insulin, etc., and any combination thereof, can be added back to the perfusate.
There are several potential future applications for this technique. As more pharmaceuticals are developed for the treatment of NASH, standard methods for assessing liver energy metabolism could find wide application. As NASH is strongly associated with liver cancer, models of these cancers are also subjects for study.
The authors have nothing to disclose.
This work was supported by funding from the National Institutes of Health (R01-DK105346, P41-GM122698, 5U2C-DK119889). A portion of this work was performed in the McKnight Brain Institute at the National High Magnetic Field Laboratory's Advanced Magnetic Resonance Imaging and Spectroscopy (AMRIS) Facility, which is supported by the National Science Foundation Cooperative Agreement No. DMR-1644779 and the State of Florida.
1 mL Luer-Lock Single Use Sterile Disposable Syringe | N/A | N/A | Non-specific Brand |
100 cm Water Jacketed Glass Column | N/A | N/A | Custom Made |
2-0 Silk Suture | Braintree Scientific | N/A | |
22 Gauge Catherter 1 in. Without Safety | Terumo | SRFF2225 | |
23 G 0.75 in. Hypodemeric Needles | Exel International | 26407 | |
27 G 1.5 in. Hypodemeric Needles | Exel International | 26426 | |
4×4 in. Surgical Platform | N/A | N/A | Custom Made |
70% Alcohol Wipe | N/A | N/A | Non-specific Brand |
Circulating Water Bath | MS Lauda | N/A | Model no longer manufactured |
Cotton Tip Applicator | N/A | N/A | Non-specific Brand |
Delicate Operating Scissors; Straight; Sharp-Sharp; 30mm Blade Length; 4 3/4 " | Roboz | RS-6702 | |
Dumont #5/45 Forceps | Fine Scientific Tools | 11251-35 | |
Dumont #7 – Fine Forceps | Fine Scientific Tools | 11274-20 | |
Hemostats | Fine Scientific Tools | 13015-14 | |
Heparin Sodium Injectable 1000 units/mL | RX Generics | 71288-0402-02 | |
Isoflurane | Patterson Veterinary | 14043-0704-06 | |
Lidocaine HCl 2% | VEDCO Inc. | 50989-0417-12 | |
Membrane-Thin-Layer Oxygenator | Radnoti | N/A | |
Metzenbaum Scissors; Curved; Blunt; 27 mm Blade Length; 5 " | Roboz | RS-6013 | |
Oxygen Meter System | Hanstech Instruments Ltd. | N/A | |
Saline 0.9% Solution | N/A | N/A | Saline is made in lab |
Scale | N/A | N/A | Non-specific Brand |
Variable Speed Analog Console Pump Systems | Cole Palmer | N/A | Models are custom per application |
Weigh boats | N/A | N/A | Non-specific Brand |