A method of obtaining a suspension of intact parenchymal cells from adult rat liver

The effect of lipid-hydrolysing enzymes and lipid-soluble drugs on the permeability of the plasma membranes of several types of normal and malignant cells is currently under investigation. It is essential that the cells used in this study are undamaged. Howard, Christensen, Gibbs & Pesch (1967) describe a method of obtaining suspen sions of rat liver cells by first perfusing the organ with collagenase and hyaluronidase, incubating the tissue with the 2 enzymes and then gently sucking the tissue up and down a wide-bore pipette. We have obtained only very low yields of intact cells by this method.

The preparation of cells (of which 0·I–I % are capable of growth in vitro) from the livers of 5-to 8-day-old rats by incubation with trypsin and collagenase is reported by Coon (1969).

We describe below a method in which adult rat liver was perfused in situ and then incubated in vitro with a solution of trypsin, hyaluronidase and collagenase. Cells which had separated from the tissue mass were collected. The properties of these cells were compared with those of cells isolated by other methods.

Burroughs Wellcome, Beckenham, England. Sterile Hanks solution: Oxoid, London, England. Calf serum: Oxoid. Bovine serum albumin (BSA) Cohn Fraction V: Koch-Lights Laboratorie Ltd. Tetraphenyl boron: BDH Chemicals, Poole, Dorset, England. iV-2-hydroxyethylpiperazine iV-2-ethanesulphonic acid, A grade (HEPES): Calbiochem, Los Angeles, Calif., U.S.A. Hyaluronidase (ex ovine testes): Koch-Lights Laboratories Ltd. Collagenase, Form I: Koch-Lights Laboratories Ltd. Trypsin: Difco 1:250, Difco Laboratories, Inc., Detroit, Michigan, U.S.A.

137 raM NaCl, 2 · 68 mM KC1, 8 · 10 mM Na₂HPO₄, 1 · 47 HIM KH₂PO₄, 0 · 456 mM CaCl₂, 0 · 491 mM MgCl₂.

PBS/A. 137 mM NaCl, 2 · 68 mM KC1, 8 ·10 mM Na₂HPO₄, 1 · 47 mM KH₂PO₄.

Trypsin (1/9, vv/v) was dissolved in buffer (1 · 37 M NaCl, 5 · 36 mM KC1, 0 · 42 mM Na₂HPO₄, 0 · 44 mM KH₂PO₄, 2 · 78 mM glucose).

After centrifugation at 10000 g for 1 h at 4 °C, aliquots of the supernatant were stored in ampoules at—40 °C. Solution E. Trypsin (0 · 1%) dissolved in PBS/A containing 0 · 3% collagenase and 0 · 3% hyaluromdase.

Citric acid (27 mM) was dissolved in Locke’s solution. The solution was adjusted to pK 7 with 2N NaOH. (Anderson, 1953.)

Eagle’s basic culture medium containing 28 mM HEPES was adjusted to pH 7 · 3 with 2 N NaOH.

Lister Hooded rat, 150–250 g, diet 41 B: Dixons.

Adult hooded rats were killed by a blow on the head and exsangui nation. The livers were perfused in situ at room temperature at a pressure of 45 cm of water (4–412 kN m^{− 2}) through the hepatic portal vein within 5 min of death.

Cells obtained from liver by the various methods were separated from broken cells and debris by filtering through surgical gauze 05-mm mesh and centrifuging the filtrate at 300 g for 1-2 min at 4 °C. The supernatant was discarded and the cells were washed 3 times by suspension and centrifugation in PBS containing 2 % BSA and were finally re suspended in this solution.

Cells were counted in an improved Neubauer Counting Chamber (Hawksley).

A small piece of the liver from each experiment was sus pended in PBS and homogenized by several strokes in a Potter-Elvehjem homogenizer.

L-Aspartate: 2-oxoglutarate aminotransferase (glutamic oxaloacetic transaminase, GOT) was estimated by the methods of Karmen, Wroblewski & La Due (1955) as modified by Boehringer (Biochemica Test Combinations, Boehringer (London) Ltd.) for estimating these enzymes in serum. Lactate dehydrogenase (LDH) was estimated by the method of Wroblewski & La Due (1955). Activity is measured as the change in optical density at 340 ran when reduced nicotinamide adenine dinucleotide (NADH) is oxidized by an enzyme reaction linked to the GOT reaction or, in the case of LDH, by the enzyme itself. One molecule of oxidized NADH is equivalent to one molecule of substrate transformed. All measurements were made with a Hilger-Gilford recording spectrophotometer.

The cells were fixed, dehydrated and embedded on a o-8-/tm Millipore filter. Cells were allowed to sediment on the filter to form a monolayer. The cells on the filter disk were transferred to 3 % glutaraldehyde in Rhodin & Zetterqvist buffer pH 7 · 4 (Zetterqvist, 1956), fixed for 1 h at o °C, washed twice and left overnight at 2 °C in the ame buffer. They were post-fixed by transferring the disk to a 1 % solution of OsO₄ in Rhodin & Zetter qvist buffer for 45 min at o °C. After washing twice with the same buffer the cells were de hydrated for 5 min in each of 20%, 40%, 70%, 90% and 3 changes of 100% isopropanol. They were then left for 30 min in 50 % Araldite in toluene and embedded in Araldite. The layer of cells on the filter disks was sectioned, stained with lead citrate (Reynolds, 1963) and observed in a Philips EM 200 electron microscope.

Cells were prepared from adult rat liver by the following methods, all of which entail per fusion of the organ.

A suspension of rat liver cells was prepared by the method of Anderson 0953) which removes calcium. The liver was perfused with Locke’s solution containing 27 mM citrate, finely chopped with scissors and placed in a Petri dish. The cells were teased from the tissue with a silicone rubber bung.

The liver was perfused with 10 ml of Hanks’s solution (without Ca²⁺) containing 0 · 16% hyaluronidase and 0 · 10 % collagenase by a method similar to that of Howard et al. (1967). It was then finely chopped with scissors and incubated for 1 h at 37 °C in enzyme solution (3 ml/g wet weight of liver) in an atmosphere of 5 % CO₂ in O₂. The cells were separated either by swirling the medium, sucking up and down using a wide-bore Pasteur pipette, or by teasing with a silicone rubber bung.

A cell suspension was prepared using the method of McLimans et al. (1966). The liver was perfused with a Ca²⁺- and Mg²⁺-free salt solution then incubated at 37 °C for 30 min in the same solution. It was finely chopped and transferred to a glass homogenizer and the cells separated using a ‘pestle’ consisting of a loose-fitting rubber bung on a metal rod.

Method 4: tetraphenyl boron. The liver was perfused with a NaCl/sucrose solution in phos phate buffer pH 7 · 8 containing 10^{− 3} M tetraphenyl boron (Rappaport & Howze, 1966) designed to remove potassium. The cells were separated from the chopped tissue (placed in a Petri dish) by teasing with a silicone rubber bung.

The liver was gently perfused at room temperature with 10 ml of a solution of 0 · 1 % trypsin in PBS/A followed by 10 ml of a solution of 0 · 3% collagenase and 0-3% hyaluronidase in 0· 1% trypsin in PBS/A (solution E). The liver was removed from the rat and chopped finely with scissors to give pieces approximately 1 mm³. The chopped liver was transferred to an Erlenmeyer flask and incubated for 15 min at 25 °C with 20 ml of solution E with occasional swirling. The supernatant was filtered through surgical gauze into a flask containing about 2 ml calf serum (to inactivate the trypsin). Trypsin solution (10 ml) was added to the remaining pieces of liver and after 5 min, with occasional swirling, the supernatant was added to the previous one. This was repeated twice. The cells were washed and resuspended in culture medium containing 10% (v/v) calf serum.

Cells were stained with 0 · 05 % eosin in PBS. Cells which excluded the dye were considered viable (Schrek, 1936).

To make them readily accessible to substrates and cofactors, the enzymes were released by adding an equal volume of 1 % Triton X-100 to the cell suspension. The enzymes were released from the homogenates by the same method in order to compare the values for their activities with those of the cell preparations.

Phase-contrast and electron microscopy.

The yields of cells in suspensions prepared from rat liver by the 5 methods (1–5) varied from 2% to 25 % (w/w) of whole liver (Table 1). Two per cent of cells prepared by method 2 and 100% of those prepared by method 5 excluded eosin; all cells pre pared by the other methods were permeable to the dye.

Lactate dehydrogenase was readily lost from cells isolated by methods 1-4 and only those cells excluding eosin contained an amount of LDH similar to that in liver homo genate (Table 2). The ratio LDH/GOT for the trypsin method 5 was 2-100 and for the liver homogenate 2-600. The values for the other methods were 0 · 037–0 · 104 (Table 2).

Cells prepared by methods 1–5 appeared undamaged when viewed by phase contrast microscopy (Figs. 1–5). However, cells prepared by methods 1, 3 and 4 were seen to have damaged membranes and grossly swollen organelles when observed with the electron microscope (Figs. 6–8). In electron micrographs, cells fixed immedi ately after preparation (method 5) and suspended in Eagle’s medium containing BSA (Figs. 9, 10) had intact plasma membranes, predominantly normal mitochondria and endoplasmic reticulum. Undamaged cells from the trypsin method were bi refringent under phase contrast (Fig. 5).

Cells isolated by method 5 and suspended in Eagle’s culture medium buffered with HEPES and containing 2% BSA did not remain impermeable to eosin for more than 1 h. Within an hour of preparation blebs were observed in cells maintained at room temperature (Fig. 5) and the enzymes were detected in the medium. However, the cells survived for a longer time in this medium than in PBS or Hanks’s solution.

Liver cells in vivo are firmly joined by special areas of contact known as junctions of which 3 types have been recognized: tight junctions, intermediate junctions and desmosomes (Farquhar & Palade, 1963). Electron microscopy showed that the plasma membrane was altered at these junctions. They also have special permeability pro perties (Kanno & Loewenstein, 1966; Loewenstein, 1966). The problem of isolating liver cells in suspension is essentially one of separating these junctions in such a way that any breaks in the cell membranes are sealed off, so that the cells become impermeable.

In method 5 perfusion of the liver with solution E was an essential to subsequent cell separation; without it yields were extremely low. After perfusion the liver was cut into 1-mm cubes and the cells were separated by mechanical agitation from the cut edges. There was an optimum cube size of about 1 mm³ below which too many cells were damaged; the use of cubes larger than 1 mm³ drastically reduced the yield of cells. In Fig. 11 double membranes are seen at various areas on the cell surface; they may represent unbroken junction areas, the isolated cell being obtained intact through loss of surrounding damaged cells by tryptic digestion. Loewenstein & Kanno (1967) have shown that cells respond to damage of neighbouring cells by sealing themselves off. Therefore cutting the liver into cubes may facilitate separation of cells from the edges.

It is not known how trypsin acts on the intercellular substance or cell membrane to facilitate separation of the cells. It may act on the membranes as a whole (trypsin reduces the cross-section of isolated membranes from 8 to 6 nm (Beneditti & Emmelot, 1968)); and it may also act on intercellular protein. Trypsin apparently cannot pene trate the intact cell but immediately a cell is damaged trypsin digests it.

In all 5 methods mechanical treatment, such as teasing or homogenization, was needed after perfusion with the various agents that facilitate cell separation. The highest yields of undamaged cells were obtained by method 5, which entailed gentle swirling of the liver cubes in the enzyme solution; and undamaged cells were obtained by use of collagenase and hyaluronidase in method 2 only when the minimum of mechanical means were used; e.g. suction up and down in a wide-bore pipette, as described by Howard et al. (1967).

Eosin is one of the dyes which living cells exclude and dead or dying cells take up. There is much controversy about the criteria for ‘live’ cells and some workers regard all cells as dead unless they are able to multiply in culture (Hoskins, Meynell & Sanders, 1956; Tennant, 1964). We have been able to maintain the cells obtained by method 5 in culture for 72 h (unpublished results).

We found that cells obtained by the citrate method 1 contained less than 5% of the lactate dehydrogenase activity expected on the basis of enzyme activity in total liver homogenate, though the content of GOT sometimes approached 100%. Loss of LDH was correlated with uptake of eosin and was used in later experiments for a quantitative assessment of the condition of the cells. GOT, although originally thought to be in the ‘soluble’ fraction of the cell (Dixon & Webb, 1958), is now known to be firmly bound in the mitochondria and is a mixture of 2 or more isoenzymes, one of which accounts for about 70% of the total GOT activity (Boyd, 1966; Schmidt & Schmidt, 1967; Schnaitman & Greenawalt, 1968). Lactate dehydrogenase has been shown to occur in the ‘soluble’ fraction after differential centrifugation. The high value of 2-6 for LDH/GOT in the liver homogenate as compared with that of the cells from method 5 (Table 2) might be accounted for by the fact that cells prepared by the trypsin method are very fragile and a few break down and leak during separation.

However, enzyme ratios of whole liver homogenate might be different from those of parenchymal cells because Kupffer and endothelial cells are present.

In parenchymal cells obtained by method 5 there was a good correlation between birefringent appearance under phase contrast (Fig. 5), good morphological detail under the electron microscope (Figs. 9, 10), exclusion of eosin and maximum content of enzymes. The cells could be maintained in suspension for up to 1 h in a medium buffered with HEPES and containing 2% BSA. In PBS the cells formed blebs after about 10 min. Yields are low, not more than 2–4% of total liver (0 · 1–0 · 2 ml packed cells), but this is sufficient for the study of several aspects of the permeability of the mammalian cell.

The authors wish to thank Dr A. M. Lawn for the use of the electron microscope, for help in preparing sections and for helpful discussion. They would also like to thank Miss Phillipa Twelvetrees for technical assistance. Support of this work by the British Empire Cancer Campaign for Research is gratefully acknowledged.