Transfer of the Hepatocyte Receptor for Serum AsialoGlycoproteins to the Plasma Membrane of a Fibroblast ACQUISITION OF THE HEPATOCYTE RECEPTOR FUNCTIONS BY MOUSE L-CELLS*

Rat liver hepatocytes have oriented externally on their plasma membrane a glycoprotein receptor which can recognize, bind, interiorize, and degrade serum glycoproteins from which terminal sialic acid residues have been removed. Mouse L-cells, a tissue culture cell line, like most cells other than hepatocytes, do not have this receptor. A membrane vesicle fraction containing right side out vesicles was prepared from rat liver by sucrose gradient fractionation techniques. These vesicles were enriched 8to lo-fold over the homogenate in their ability to bind ‘251-asialo-orosomucoid. However, they cannot degrade the iodinated asialo-orosomucoid. The membrane vesicles derived from rat liver were reacted with mouse L-cells in the presence of polyethylene glycol. When L-cells and ‘251-labeled membrane vesicles were subjected to two cycles of polyethylene glycol-mediated fusion, the iodinated membrane vesicles became stably associated with the L-cell membrane. That is, the turnover behavior of the fused liver membrane was indistinguishable from the plasma membrane of the L-cell. After fusion with liver membrane vesicles, the L-cell at 37°C was able to bind, interiorize, and degrade ‘251-asialo-orosomucoid which was given to the cells in the culture medium. Although the iodinated asialo-orosomucoid was degraded to acidsoluble material, the receptors inserted into the L-cell membrane maintained their ability to function through a second cycle of binding, interiorization, and degradation of ‘251-asialo-orosomucoid. These experiments show that it is possible to impart to a cell a complex physiological process by inserting into the plasma membrane of that cell the requisite receptor that initiates the sequence of events in the process.


Rat liver hepatocytes
have oriented externally on their plasma membrane a glycoprotein receptor which can recognize, bind, interiorize, and degrade serum glycoproteins from which terminal sialic acid residues have been removed.
Mouse L-cells, a tissue culture cell line, like most cells other than hepatocytes, do not have this receptor.
A membrane vesicle fraction containing right side out vesicles was prepared from rat liver by sucrose gradient fractionation techniques. These vesicles were enriched 8-to lo-fold over the homogenate in their ability to bind '251-asialo-orosomucoid. However, they cannot degrade the iodinated asialo-orosomucoid. The membrane vesicles derived from rat liver were reacted with mouse L-cells in the presence of polyethylene glycol.
When L-cells and '251-labeled membrane vesicles were subjected to two cycles of polyethylene glycol-mediated fusion, the iodinated membrane vesicles became stably associated with the L-cell membrane. That is, the turnover behavior of the fused liver membrane was indistinguishable from the plasma membrane of the L-cell. After fusion with liver membrane vesicles, the L-cell at 37°C was able to bind, interiorize, and degrade '251-asialo-orosomucoid which was given to the cells in the culture medium.
Although the iodinated asialo-orosomucoid was degraded to acidsoluble material, the receptors inserted into the L-cell membrane maintained their ability to function through a second cycle of binding, interiorization, and degradation of '251-asialo-orosomucoid. These experiments show that it is possible to impart to a cell a complex physiological process by inserting into the plasma membrane of that cell the requisite receptor that initiates the sequence of events in the process.
In animal cells, protein and glycoprotein "receptors" that are externally oriented in the plasma membrane are believed to provide the initial steps in complex systems that are responsible for the cell's ability to recognize and respond to changes in the environment (l-3 to the cell have been identified in recent years. These include receptors for polypeptide hormones such as insulin (4) and glucagon (5), receptors for lysosomal enzymes (6,7), receptors for normal serum components such as low density lipoproteins (3,8), and receptors for altered serum components such as a receptor on liver hepatocytes that can recognize 'and bind serum glycoproteins that have had their terminal sialic acid residues removed, thus exposing penultimate galactose residues (9-11). The mechanisms by which receptors which are externally oriented on the plasma membrane function to bring about a complex physiological response in the cell is, for the most part, not well understood.
Several properties of these receptor proteins have made it difficult to analyze their mechanisms of action. Many of the receptor proteins are present in the membrane at very low concentrations. Further, they are often not confined exclusively to the plasma membrane, but are found in other intracellular membrane systems of the cell (12, 13); the presence of the same receptor in two different membrane compartments complicates enormously mechanistic and biogenetic studies of these membrane receptor proteins (14).
One way to obtain insight into the mechanism of receptor action would be to "induce" a cell to make and insert into its plasma membrane a receptor that it did not previously have. But, it has been very difficult to do this type of experiment rigorously in animal cells. However, in the present study, we use an alternative approach and show that the specific hepatocyte receptor that can recognize serum asialoglycoproteins can be inserted into the plasma membrane of mouse L-cells by cell fusion techniques.
The receptor confers upon the mouse L-cell a complex of functions that it did not previously have, i.e. the ability to recognize, bind, interiorize, and degrade asialo-orosomucoid.
This type of approach offers the possibility of reconstituting and analyzing complex plasma membrane-mediated functions of mammalian cells. L-Cells (18)-Monolayer cultures of confluent L-929 fibroblasts in 75cm" flasks (Costar, Cambridge, Mass.) were treated with trypsin from Gibco (0.5 mg/ml) for 5 min at 37°C. The suspended cells were washed two times with phosphate-buffered saline. To the packed cell pellet containing as little phosphate-buffered saline as possible was added newly prepared P2 plasma membrane fraction of liver cells, 1 to 2 mg of liver membranes in about 0.1 ml/l X IO' cells/fusion. The Co.) in serum-free medium-was added for 1 min at room temperature.
Then, 15 ml of serum-free medium was added. After 2 min more, the cells were collected by centrifugation and were washed once with serum-free medium.
A second identical cycle of fusion was performed after which the cells were washed twice with serum-free medium, and placed back in monolayer culture at 37°C. The cells were 80 to 90% viable after the second cycle of fusion.
Binding That is, none of these cell lines can distinguish between '""I-orosomucoid and IS5 I-orosomucoid from which the terminal sialic acids have been removed either chemically or by neuraminidase.
In contrast, the membrane vesicles derived from rat liver are very efficient in their ability to bind asialo-orosomucoid; 1 mg of these vesicles will bind about 600 ng of asialo-orosomucoid at saturation (Fig.  1) and then fused with L-cells, the radioactive iodide becomes stably associated with the L-cell (Fig. 2). Indeed, the radioactive iodide of the liver membrane vesicle is as stable as the iodide incorporated into the membrane proteins of the L-cell itself (Fig. 2). The iodinated liver membrane proteins inserted into L-cells or HTC cells by polyethylene glycol-mediated fusion have similar and almost identical turnover characteristics as the iodinated proteins of the plasma membrane of the recipient cell." The experiment presented in Fig. 2 shows that the iodinated membrane proteins of the L-cell as well as those iodinated liver membrane proteins inserted into the L-cell by polyethylene glycol-mediated fusion turn over with relatively long half-lives. Further, the presence of asialo-orosoniucoid in the medium has no effect on the turnover of the iodinated hepatocyte proteins introduced into the L-cell. Elsewhere,3X 4 we will show that the hepatocyte receptor for asialoglycoproteins, as assayed with a specific antibody, also has a half-life similar and probably identical with the other iodinated proteins of the liver mem-' E. Hou, R. Warren, and D. Doyle, manuscript in preparation. 4 R. Warren, and D. Doyle, manuscript in preparation. The Pz fraction of rat liver plasma membrane was iodinated with '*$I via the chloramine-T procedure (specific activity, 6.7 x IO* cpm/mg of membrane protein). Confluent L-929 cells from one 75-cm2 flask were removed with trypsin and then reacted with polyethylene glycol through two cycles of fusion with the liver membranes (150 pg of iodinated membrane mixed with 1 mg of unlabeled liver membrane protein was used per fusion).
The fused cells were washed twice with serum-free medium and then distributed among twenty plates (35 X 10 mm) (0, lower graph).
Eight hours later, asialo-orosomucoid (5 pg/ml of medium) was added to some plates (upper graph). At the times indicated, asialo-orosomucoid was removed and added back to these cells. Trichloroacetic acid-insoluble radioactivity was assayed at the times indicated.
Zero time is immediately after the fused cells were placed on the plates (35 x 10 mm). A confluent culture of L-cells growing as a monolayer in one 75-cm* flask was iodinated via lactoperoxidase-catalyzed iodination, 0.5 mCi of ""I/flask. The cells were then removed from the flask with trypsin and were distributed among ten plates (35 x 10 mm) (0, lower graph).
brane and the host cell membrane after fusion of liver membrane with the recipient cell. Long half-lives or slow turnover times relative to the doubling time of the cell seems to be a characteristic property of plasma membrane proteins of many different types of cells in culture (1,(23)(24)(25).
To demonstrate that the biological response that is specified by the hepatocyte receptor is transferred to the recipient cell intact, we assayed for function by measuring the uptake and degradation of asialo-orosomucoid. As shown in Fig. 3, L-cells containing the fused liver membrane, in contrast to the parent L-cell, can bind, take up, and degrade '251-asialo-orosomucoid. Two cycles of fusion were used to transfer the liver membranes to the L-cell because, under optimal conditions, only about 1 to 5% of the liver membranes will become associated with the recipient cell. Two cycles of fusion allow a sufficient number of receptor proteins to be inserted into the L-cell to permit were fused with the Pz fraction of rat liver as described under "Experimental Procedures" and in the legend to Fig. 2. The cells were distributed among 40 plates (35 x 10 mm) which were incubated at 37°C overnight. The next day, the attached cells were washed three times with phosphate-buffered saline. Medium (1 ml), containing ""I-asialo-orosomucoid (4 x 10" cpm, 10 ng, iodinated via the lactoperoxidase-Sepharose method) and 5 pg of unlabeled asialo-orosomucoid was added to each culture. Cellassociated phosphotungstic acid-insoluble radioactivity was determined at the times indicated.
Cells were washed three times with phosphate-buffered saline and material insoluble in phosphotungstic acid was counted in the Biogamma spectrometer.
After 8 h in the presence of asialo-orosomucoid, the medium was removed from the cells and the cells were washed twice with medium not containing asialo-orosomucoid.
Phosphotungstic acid-soluble radioactivity in the medium and acid-insoluble radioactivity associated with the cells was determined at the times indicated. At the time indicated, '2"I-asialoorosomucoid (4 x 10" cpm) and unlabeled asialo-orosomucoid, 5 pg, were added back to the cell cultures.
Acid-insoluble radioactivity associated with the cells was assayed. Finally, the asialo-orosomucoid in the medium was removed by washing the cells again and cellassociated acid-insoluble radioactivity and acid-soluble radioactivity in the medium were determined. The L-cell before fusion with the liver membranes will bind 500 cpm or less of asialo-orosomucoid when added at the same concentration and under the same conditions as specified above for the fused cells.
assay of function. It should be mentioned that the PZ fraction of rat liver, while heavily enriched in plasma membrane fragments, is not homogeneous.
Since the receptor for asialoorosomucoid is also present in other membrane fractions of rat liver (12), there is some possibility that membranes other than the plasma membrane transfer the receptor to the L-cell. In Fig. 3, the cells after fusion were placed back in culture in the presence of '251-asialo-orosomucoid (5 pg/ml). This concentration of asialo-orosomucoid is very much in excess of the amount that can be processed in the L-cell by the relatively limited number of receptors that have been inserted even by two cycles of fusion. For example, at most, 100 pg of liver membranes is transferred to lo7 cells after two cycles of fusion in which 1 mg of liver membrane is reacted twice with lo7 cells. As shown in Fig. 1, this amount of liver membrane can only bind on the order of 60 ng of the asialo-orosomucoid at saturation at 4°C. Elsewhere,3 we will show that after 6 to 8 h, the amount of asialo-orosomucoid associated with the Lcell reaches a plateau. The time required to reach this plateau represents the time required to achieve a steady state of binding, interiorization, and degradation of the asialo-orosomucoid by the cells relative to the concentration of the asialoorosomucoid in the medium. This time of about 6 to 8 h for the modified L-cells is actually very similar to the time required by primary cultures of isolated rat hepatocytes to do the same series of reactions and attain a steady state between the cell-associated asialo-orosomucoid and that in the me-Insertion and Function of Membrane Receptors in Foreign Cells dium. Hence, we presume that the L-cell is functioning very much like the hepatocyte in this series of reactions. The individual steps in the series can be analyzed separately and we will report on the properties of the binding, interiorization, and degradation steps in more detail in subsequent communications. We point out here that binding of the asialoglycoprotein by the receptor is a necessary but not the only prerequisite for the cell to accomplish interiorization and degradation. The L-cell is supplying some components for these latter steps because we3 can show that HTC cells after polyethylene glycol-mediated fusion with liver membranes can bind asialoglycoproteins, but the modified HTC cells cannot degrade the bound protein. When the asialo-orosomucoid is removed from the medium of the modified L-cell at about the time of the plateau, between 5 and 10 h in Fig. 3, the cell will complete the degradation of the asialo-orosomucoid still associated with the cell to small molecular weight acid-soluble material (Fig.  3).
The hepatic asialoglycoprotein receptor once inserted into the L-cell is, as mentioned above, stable and apparently is not degraded during the steps involved in the degradation of the bound asialoglycoprotein.
These steps, as mentioned, presumably include binding, interiorization of membrane units containing the receptor and bound asialoglycoprotein, fusion with lysosomes, and degradation of the asialoglycoprotein to acidsoluble material (1,23,26). If these steps, indeed, are involved in the mode of this receptor action, they do not affect significantly the receptor itself because the modified L-cell, after interiorizing and degrading the bound asialo-orosomucoid, is still capable of binding and degrading asialo-orosomucoid when the cells are exposed for a second time to high concentrations of the glycoprotein in the medium (Fig. 3). Hence, the hepatic receptors are still present and functional on the cell surface. Also, it is possible to demonstrate by fluorescence microscopy that hepatocyte membranes are localized at the surface of L-cells after fusion with liver vesicles." An explanation for the ability of the hepatic receptor for asialoglycoproteins to function for extended times after its insertion into L-cells while the asialoglycoprotein that it binds is degraded is that the receptor is capable of cycling in and out of the cell surface. Tanabe et al. (27) recently have presented some evidence for this type of receptor recycling in intact rats and we4 also have evidence that recycling is part of the reason for the stability of the asialoglycoprotein receptor in isolated rat hepatocytes. Fig. 3 also shows that there is a slow component in the degradation of the asialo-orosomucoid. This is seen most easily after the asialo-orosomucoid is removed from the medium between 25 and 30 h. Isolated hepatocytes also show this type of behavior4 (28); the biphasic kinetics are due to high affinity receptor-mediated turnover and also a low affnity nonreceptor-mediated turnover of the asialoglycoprotein. The remarkable biological stability of plasma membrane proteins coupled with the ability to insert foreign membrane proteins into a tissue culture cell of choice and have these proteins function should help to resolve some of the steps involved in the mechanism of receptor action. It should be mentioned that it is possible that this method can be used to put any foreign membrane protein into a recipient cell. Indeed, while our experiments were in progress, a report appeared showing that similar techniques could be used to activate a cell's adenyl cyclase by fusion of a foreign membrane-bound glucagon receptor (29). Finally, since several cycles of polyethylene glycol-mediated fusion can be done on one cell, it might be possible to supply more than one component of a complex system to the recipient cell. In this way, it might be possible to "reconstitute" in the recipient cell a complex membrane-mediated response by obtaining the component pieces from other cells lacking one or more of the membrane proteins involved in the response.

functions by mouse L-cells. plasma membrane of a fibroblast. Acquisition of the hepatocyte receptor
Transfer of the hepatocyte receptor for serum asialo-glycoproteins to the