Protocol for the study of hepatic bilirubin uptake in the isolated perfused rat liver

Marco Stebel University of Trieste https://orcid.org/0000-0002-8279-8065 Nevenka Medic University of Trieste https://orcid.org/0000-0002-1659-1461 Paola Pelizzo University of Trieste Paola Sist University of Trieste https://orcid.org/0000-0003-1626-5081 Federica Tramer University of Trieste https://orcid.org/0000-0003-4286-0191 Sabina Passamonti (  spassamonti@units.it ) University of Trieste https://orcid.org/0000-0001-7876-4666


Introduction
Bilirubin is the product of heme catabolism occurring in all cells, with the highest contribution coming from splenic macrophages involved in the turnover of senescent erythrocytes. Bilirubin (BR) circulates in the blood as a reversible complex with serum albumin. It is selectively taken up into the liver, metabolized and excreted into the bile as bilirubin glucuronide (BRG) 1 .
The membrane transporter(s) for hepatic BR uptake is still unknown 234 . In both rats and humans, no signi cant correlation between drug-induced inhibition of hepatic SLCO1B1 and 1B3 transporters (OATP1B1 and OATP1B) and hyperbilirubinemia was found 5 , possibly due the participation of several additional factors in the whole organism.
This points to the need of applying a simpli ed model to investigate BR hepatic disposition. Among the various in vivo and in vitro experimental models for the study of the liver function, the isolated and perfused rat liver (IPRL) presents the advantage of a fully intact architecture, preserving the critical feature of bile ow 6 , while enabling the control of the liver perfusion conditions, including the delivery of test compounds.
We have developed a protocol for the analysis of multiple single-pass events (up to 12) of sinusoidal BR uptake in the same preparation, by modifying the standard IPRL method 7 . This protocol can be implemented in a time window of approximately 25 min. The presence of BR and BRG was analyzed in the liver e uent by a uorometric method 8 , with some adaptations. The preparation was viable and intact, as shown by both standard viability parameters and by the occurrence of bilirubin glucuronide hopping 9 .

Page 3/21
This protocol offers the advantage of obtaining repeated observations in the same liver, thus reducing the number of animals. It offers therefore ample opportunity to screen molecules for their potential to inhibit hepatic BR and BRG membrane transporters.

TECHNICAL NOVELTIES
1. Perfusion of the liver without oxygen supplementation at physiologically low ow rate.
2. Repeated intra-portal albumin-free boli of BR alone or with inhibitors of sinusoidal membrane transporters, so increasing the number of single-pass events in the same preparation, with animal reduction.
3. Use of FITC-labelled bovine serum albumin for the assessment of the intra-hepatic micro-vascular integrity.
4. Direct analysis of BR and BR glucuronide in the hepatic venous e uent by a high-throughput uorometric assay. Dilute these concentrated solutions in DMSO, vortex and then add PBS to 1 mL, according to the scheme below, to obtain any of the 3 types of boli with the composition suitable for intra-portal injection (read Tips 1-2) ( Figure 1).

Albumin-FITC bolus
Prepare a solution of Albumin -FITC, consisting in 77.5 µM Bovine Serum Albumin (BSA), containing 3% w/w of BSA-FITC, according to the protocol described in 11  The other end of this tubing is be connected to the inlet catheter (read details below).

Assembly of the inlet catheter
The silicon tubing delivering the perfusion solution is to be connected to the 20 G winged & ported catheter for intra-portal perfusion. Manual delivery of repeated boli via the port of this catheter is not secure, since even small vibrations of the intra-portal catheter may injury the thin wall of the portal vein.
Thus, another port, to be used for boli delivery, is interposed between the intra-portal catheter and the terminal end of the silicon tubing. This port is assembled as follows: § Connect the end of the silicon tubing, stemming from the OUT-direction hole of the bubble trap, to a winged & ported catheter (20 G), modi ed by cutting away its cannula (see above Labware). The port of this truncated catheter is called "bolus delivery port", to distinguish it from the catheter port, which is not used. § Connect the tip of the truncated catheter to a short silicon tubing (OD 5 mm, 5 cm). § Insert the other end of the latter tubing into the luer of a truncated 18 G needle (see above Labware). § Mount a silicon tubing (OD 4 mm) sleeve on the truncated 18 G needle segment. § Fit this sleeve into the luer of the 20 G catheter for intraportal perfusion. This catheter is called "perfusion tubing terminal".
Assembly of the outlet catheter

Method and ethics
The surgical procedure is according to 7 , whereas the perfusion of the isolated liver with HBBS and PBS is new. Obtain the prior approval of animal experimentation from the competent bioethical authority. Justify the use of animals, by applying reasons for the 3R principles, e.g.: § There are no in vitro experimental models preserving the bile ow through the biliary system 6 (no available replacement). § The procedure causes no sufferance to the anaesthetized rat (maximum achievable re nement). § Repeated tests of BR (up to 12) can be performed in the same liver preparation, thus increasing the number of observations in the same liver, thus limiting the animal number (reduction by a factor of 10).
Surgery § Perform a horizontal laparotomy in the lower abdomen, by lifting the abdominal wall with the tweezers. Start the cut from the median point, continue towards each lateral side, and then turn vertically upwards, up to the lowest rib. Take care not to cut the diaphragm, the chest and its respiratory muscles at this stage. § Displace the intestine to the left side to expose the hepatic hilus, with the portal vein and the common bile duct. § Prepare a ligature around the inferior vena cava, above the insertion of the right renal vein, and a second one around the portal vein (about 1 cm distally from the hilus), using a two-strand sewing cotton thread. § At this stage, consider to prepare the common bile duct for cannulation, if planned in the experiment (not described in this protocol). § To prevent blood clotting, inject a bolus of heparin (500 U in 0.25 mL PBS) into the inferior vena cava by an insulin syringe. § Activate the peristaltic pump at 1 mL/min in order to ll the silicon tubing with the perfusion solution.
Allow it to leak dropwise. § Insert the inlet catheter (20 G winged & ported) into the portal vein and secure it by tightening the ligature around the portal vein. § Connect the inlet catheter to the "perfusion tubing terminal" (see above, Labware). § Perfuse the liver at a ow rate of 4 ml/min and at a constant temperature of 37 °C. § Incise the inferior vena cava caudally to the ligature, to allow blood and perfusion buffer to out ow from the liver, so to maintain normal intrahepatic pressure. § Perform a thoracotomy starting from the diaphragm up to the neck, along two lateral lines. § Insert the 18 G catheter through the right atrium into the inferior vena cava, secure it with a tight ligature. § Connect the screw cap on the PEEK tubing (OD 2 mm, ID 1.8 mm) to this catheter, to collect the liver perfusion e uent. § Close the inferior vena cava, above the tributary renal vein, by tightening the ligature, to prevent retrograde perfusion of the liver. § Accelerate the peristaltic pump to deliver the intra-portal perfusion HBSS solution at 6 mL/min for 20 min (110 mL), to exsanguinate the liver and prepare it for BR uptake tests (read Tip 6). § Switch the liver perfusion buffer from HBSS to PBS and perfuse until the e uent is clear (read Tip 7).

BR AND BRG UPTAKE TESTS
Sinusoidal administration of BR and BR uptake inhibitors § Prepare a set of at least 120 vials (1.5 ml) and 12 tubes (10 mL) for the collection of the e uent fractions. § Insert a 1 mL-syringe, lled with 0.2 mL BR solution, into the bolus delivery port on the tubing connected to the inlet catheter (read Tips 8-9). § Inject the BR bolus (0.2 mL) in 2 seconds and keep the syringe in place. § Collect 10 e uent fractions (0.4 mL each; 4 sec/fraction) into 1.5 ml vials (read Tip 10). § Remove the 1 mL-syringe from the portal cannula and continue the liver perfusion for 1 min (6 mL) before the next BR bolus and collect this inter-bolus fraction in a 15 mL-tube. § Inject the second BR bolus (0.2. mL), as described above. § Continue the liver perfusion for 1 min (6 mL), as above. § Repeat BR boli injections up to 12 times.

HEPATIC VENOUS EFFLUENT ANALYSIS Bilirubin and Bilirubin glucuronide
Perform the uorometric analysis of BR by means of the Help-UnaG fusion protein (HUG) 8 , after uorescence calibration with a BR standard solution. Check for interference of BR-speci c uorescence by the compounds added with BR. Analyze BRG as BR equivalents, after hydrolysis by β-Glucuronidase.

Analysis of assay interferences
When planning to co-inject BR with other compounds via intra-portal boli, perform a prior check for their possible interference with the BR uorometric assay (read Tip 12). § Test the compounds at 1/10 their concentration in the bolus, to take account of their dilution in the sinusoids (read Tip 13). § Use the concentrated solutions of compounds in DMSO, as described above (section bolus preparation): E17G, Indomethacin and Ketoprofen (50 mM); C3G, M3G, P3G, TC (100 mM); Pravastatin (300 mM). § Dilute 10 µL of these concentrated solutions in 5 mL of standard BR as in the table below, to obtain BR standard solution containing 0.1 mM E17G or Indomethacin or Ketoprofen, 0.2 mM C3G or M3G or P3G or TC, and 0.6 mM Pravastatin, as in Figure 3.
2. When using mixed boli, prepare the bolus solution as late as possible before injection, to avoid possible chemical reactions (e.g., electron transfer) or interactions (e.g., aggregation or complexation) between BR and compounds. 8. Mind not to use the catheter port, but only the "bolus delivery port" (see Liver perfusion set-up). Fill the latter with 0.1 mL perfusion buffer before inserting the 1 mL-syringe, to avoid injecting an air bubble before the bolus. 9. When injecting repeated boli, do not pre ll the syringe with multiple bolus volumes (multiples of 0.2 mL), to avoid repeated mixing of the perfusion uid with the syringe contents, leading to imprecise delivery of the compounds. 10. Bilirubin diluted in PBS to nM concentration is chemically unstable. Thus, transfer samples of the venous e uent to the HUG solution immediately after collection.
11. PBS, but not HBSS, is suitable for applying the analysis of BR in the perfusion e uent 8 . 12. Systematically check the interference of compounds delivered as intra-portal boli on the BR assay.
Some compounds may interfere with uorescence absorption and/or emission of the HUG-BR complex.
13. If the direct analysis of a given compound in the venous e uent of the liver is not possible, then indirectly estimate the highest boundary of the compound dilution by injecting a bolus of Albumin-FITC and analyze its concentration in the e uent fractions. Consider that a compound that is e ciently taken up may be more diluted than BSA-FITC.
14. Optimal BSA bolus is 77.5 µM BSA with 3% BSA-FITC, while lower [BSA] may result in unspeci c adsorption of BSA-FITC and underestimation of recovery.
15. For implementing this protocol, three operators are needed: one expert in rat surgery, who performs liver perfusion and intra-portal delivery; one assistant to surgery and managing sample collection; one assistant to sample collection and immediate transfer to the BR assay.
Time Taken § Rat anesthesia, laparatomy, exposure of blood vessels and preparation of portal vein cannulation: 20 min. § Isolation of the liver (from cannulation of the portal vein to cannulation of the inferior cava vein): 15 min. § Exsanguination of the isolated liver: 25 min § Uptake tests (12 sequential boli of BR): 25 min Anticipated Results § Undetectable lipoperoxidation products, due to the absence of oxygen supplementation. § Liver viability maintained for up to 70 min after isolation from the circulation and onset of perfusion. § Extensive uptake of BR, with no or very low (nM) concentrations found in the venous hepatic e uent. § Detection of very low (nM) concentrations of BRG in the venous hepatic e uent. § Intra-portal co-injection of BR with compounds targeting sinusoidal membrane transporters may result in transiently increased concentrations of BR or BRG in the venous e uent from the liver (peak-or waveshaped). § Possibility to test the effects of up to 3 different compounds, known to inhibit hepatic sinusoidal transporters, in the same liver preparation.
Preparation of standard intra-portal boli.  Preparation bilirubin standard solutions with potentially interfering compounds.