Endocytosis, Toxins, Immunotoxins and Viruses Society/Host Colloquium Organized by J. M. Lord (University of Warwick) and E. Wawrzynczak (Institute of Cancer Research, Sutton). 643rd Meeting held at the University of Warwick, 22-23 July 1992.
Endocytic and transcytic pathways in Caco-2 cells J. Paul Luzio, Mark R. Jackman and Juliet A. Ellis Department of Clinical Biochemistry, University of CambridgdAddenbrooke'sHospital, Hills Road, Cambridge CB2 2QR, U.K.
cells [3, 51. Membrane proteins transcytosing in the opposite direction have been more difficult to identify, owing to the lack of suitable markers [7]. Apical to basolateral transcytosis of a ligand has been observed in the case of cobalamin, though this involves transfer from intrinsic factor to transcobalamin I1 [8]. The Caco-2 cell line has considerable potential as a model system to study transepithelial transport [9, 101. However, the exploitation of mechanisms controlling transcytosis of membrane proteins between the apical and basolateral domains for targeted drug delivery requires both a greater knowledge of endocytic pathways from each domain and the identification of proteins undergoing apical to basolateral transcytosis. In MDCK cells, a general method of identifying transcytosed membrane glycoproteins has been developed as a result of the isolation and characterization of an MDCK clone resistant to the toxic lectin of Ricinus communk (ricin). Resistance was due to a mutation that disrupted galactosylation of membrane glycoproteins, preventing ricin binding to the cell surface [ 11- 131. This ricin resistant MDCK cell line (MDCKII-RCA') was enriched in cell surface glycoconjugates bearing terminal N-acetylglucosamine residues, enabling domain specific labelling of glycoconjugates by the addition of exogenous UDP-[ 'H]galactose and galactosyltransferase. The transcytosis of membrane proteins labelled in a domain-specific manner was detected by surface biotinylation of the opposite membrane domain, detergent solubilization of these glycoproteins, adsorption onto streptavidin-agarose, and finally, separation by SDS-PAGE. This approach overcomes the need for apical or basolateral membrane marker proteins of high enough abundance to study transcytosis. However, a prerequisite for the appli-
The human colonic adenocarcinoma derived cell line, Caco-2 [l], may be grown as a polarized epithelium on permeable filter supports and develops a number of characteristic enterocyte functions when grown to confluency. The differentiated functions include the development of apical brush border microvilli, the presence of tight junctions, the unique electrical properties of transporting epithelia and polarized distribution of cell surface enzymes and receptors. Caco-2 cells are most like foetal intestinal cells in terms of expressed hydrolases, glycogen content and secretion of apolipoproteins and alphafoetoprotein [2]. When grown on filter supports the cells show polarized secretion of lipoproteins from the basolateral side [2]. They also show the presence of two membrane traffic pathways for delivery of newly synthesized membrane proteins to the apical surface [3]; one involving sorting in the trans-Golgi network analogous to the route found in Madin-Darby canine kidney (MDCK) cells [4, 51, the other involving initial transport from the trans-Golgi network to the basolateral membrane followed by transcytosis to the apical surface, similar to the route found in hepatocytes [6]. Endocytosis has proven easiest to study from the basolateral surface where many common receptors are found. One attempt to study uptake from the apical side showed mixing in a common endosoma1 compartment of apically endocytosed concanavalin A-gold with transferrin endocytosed from the basolateral side [7]. The route of delivery of newly synthesized membrane proteins such as aminopeptidase N demonstrates the existence of the basolateral to apical transcytic pathway in these Abbreviations used: MDCK, Madin-Darby kidney; KCA, Ricinus communis.
canine
717
I992
Biochemical Society Transactions
718
cation of this method to Caco-2 cells is the isolation and characterization of a ricin resistant Caco-2 (Caco-2-RCA') clone bearing a similar mutation to that seen in MDCKII-RCA' cells. Unfortunately, whilst it has proven possible to isolate ricin-resistant Caco-2 clones, none obtained to date have an appropriate mutation to allow repetition of the elegant experiments performed with MDCK (M. R. Jackman, J. A. Ellis & J. P. Luzio, unpublished work). Nonetheless, the study of the binding and uptake of ricin itself has provided some interesting new data on endocytic pathways in Caco-2 cells. Ricin was iodinated using Na"'I (50 mCi/ml, Radiochemical Centre, Amersham, U.K.) with the solid phase Iodogen technique (Pierce, Europe B.V.) [15], to a specific activity of 5 x lo6 d.p.m./pg of protein. T o assess ricin binding, cells growing in 24-well disposable trays at a density of 4 X lo6 cells/well were washed once with phosphatebuffered saline containing 0.5 mM-Ca2+, 0.5 mMg2+ (PBS+), and incubated for 1 h at 4°C with 25 ng/ml '251-labelled lectin in Dulbecco's modified Eagle's medium (without glutamine, glucose, pyruvate, phenol red and NaHCO,), 20-m~-Hepes,pH 7.4, containing 0.6% (w/v) bovine serum albumin, supplemented with increasing amounts of unlabelled lectin. Cells were then given three 5 min washes with ice-cold PBS+, and the cell bound radioactivity was measured. The amount of lectin which could not be removed from the cells with two 5 min washes with ice-cold PBS+, was used as a background value for each concentration. The amount of lectin binding per cell was calculated and plotted according to Scatchard [ 161. Endocytosis and transcytosis of [1251]ricinbound to the cell surface at 4°C were assessed on cells grown on filter supports [17] after warming cells for defined periods at 37°C and then releasing remaining or transcytosed cell surface bound ricin by the addition of 0.2 M-lactose to the apical or basolateral chamber as appropriate. Scatchard analysis showed that Caco-2 cells have 3.5 X ricin binding sites per cell and the K A for ricin binding is 12.5 x lo6 M - ' (M. R Jackman, J. A. Ellis & J. P. Luzio, unpublished work). Endocytosis of ricin from the apical and basolateral domains altered with the age of the cell layer grown on filter supports. Both apical and basolateral uptake of ricin fell between 7 and 14 days of culture. At 7 days, the percentage of membrane-bound ricin endocytosed from the basolateral surface was more than double that endocytosed from the apical surface after 30 min incubation at 37"C, whereas at 14 days the percentage
Volume 20
endocytosed from each domain was similar and approximately 80% of the apical uptake of the 7 day cells. Using 7 day cultured cell layers it was found that 3.6% of apically bound ricin was transcytosed in 2 h compared with 4.8% of basolaterally bound ricin. This contrasts with the situation in MDCK cells [15] where greater uptake of ricin from both apical and basolateral surfaces is observed and where apical to basolateral transcytosis is approximately 10-fold greater than that in the reverse direction. Endocytosis from the apical and basolateral surfaces of Caco-2 cells differs qualitatively as well as quantitatively. Apical endocytosis of ricin was reversibly inhibited by > 50% by 1 pM-cytochalasin D whereas concentrations as high as 30 p~ had little effect on basolateral endocytosis. In MDCK cells, both apical and basolateral endocytosis of ricin were similarly insensitive to cytochalasin D. Fluorescence microscopy has shown that at concentrations of cytochalasin D that affect apical endocytosis in Caco-2 cells, stress fibres are disrupted but not microvillar actin. Preliminary evidence measuring folate uptake into Caco-2 cells is consistent with some or all of the cytochalasin D effect being due to inhibition of uptake by non-coated pit mechanisms. In other cells, folate uptake has been shown to occur on a glycosyl phosphatidyl inositolanchored receptor that is internalized via caveolae [181. Whilst the investigation of ricin uptake in Caco-2 cells has yielded both qualitative and quantitative information about endocytosis from the apical and basolateral domains, it has not, to date, resulted in the identification of any individual membrane protein that is transcytosed. An alternative approach has been to search for proteins that are delivered from synthesis to both apical and basolateral surfaces. Such proteins are candidates for transcytosis if subsequently endocytosed, since they do not contain signals for domain-specific targeting. Using a polyclonal antiserum raised against a detergent extract of an apical plasma membrane fraction, five integral membrane proteins common to both apical and basolateral domains were identified. In pulse chase experiments, these five proteins were shown to be delivered to each cell surface domain with the same time course [ 171. The antiserum has been used to screen a human colon carcinoma library in the bacteriophage expression vector l g t l 1. Candidate clones encoding three of the five integral membrane proteins have been identified. It is hoped that a strategy similar to that used in the identification of Golgi-specific membrane proteins
Endocytosis, Toxins, Immunotoxins and Viruses
[I91 will prove successful in characterizing proteins Dresent on both aDical and basolateral surfaces of Caco-2 cells. Their endocytosis and transcytosis will then be studied. We thank the Cancer Research Campaign for financial support. M.R.J. is supported as a research student by Ciba-Geigy Pharmaceuticals.
1. Fogh, J., Fogh, J. M. & Orfes, T. (1977) J. Natl. Cancer Inst. 59,211-226 2. Traber, M. G., Kayden, H. J. & Rindler, M. J. (1987) J. Lipid Res. 28, 1350-1 362 3. Matter, K., Brauchbar, M., Bucher, K. & Hauri, H.-P. (1990) Cell 60,429-437 4. Lisanti, M. P., Le Bivic, A. Sargiacomo, M. & Rodriguez-Boulan, E. (1989) J. Cell Biol. 109, 21 17-2127 5. Le Bivic, A., Quaroni, A., Nichols, B. & RodriguezBoulan, E. (1990) J. Cell Biol. 111, 1351-1361 6. Bartles, J. R., Feracci, H. M., Stieger, B. & Hubbard. A. L. (1987) J. Cell Biol. 105, 1241-1251 7. Hughson, E. & Hopkins, C. R. (1990) J. Cell Biol. 110,337-348
8. Dix, C. J., Hassan, I. F., Obray, H. Y., Shah, R. & Wilson, G. (1990) Gastroenterology 98, 1272-1279 9. Hidalgo, I. J., h u b , T. J. & Borchardt, R. T. (1989) Gastroenterology 96,736-749 10. Simons, K. & Fuller, S. (1985) Ann. Rev. Cell Biol. 1, 248-288 l l . Meiss, H., creen, R. F. & ~ ~ d ~ iE. ~ ( 1982) Mol. Cell Biol. 2, 1287- 1294 12. Brandli, A. W., Hansen, G. C. & Rodriguez-Boulan. E. (1988) J. Biol. Chem. 263, 16283-16290 13. Brandli, A. ( 1991) Biochem. J. 276, 1 - 12 14. Reference deleted. 15. Fraker, P. J. & Speck, J. C. (1978) Biochem. Biophys. Res. Commun. 80,849-857 16. Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51,660 17. Ellis, J. A., Jackman, M. R. & Luzio, J. P. (1992) Biochem. J. 283,553-560 18. Rothberg, K. G., Ying, Y., Kolhouse, J. F., Kamen, B. A. & Anderson, R. G. W. (1990) J. Cell Biol. 110, 637-649 19. Luzio, J. P., Banting, G., Howell, K., Brake, B., Braghetta, P. & Stanley, K. K. (1990) Biochem. J. 270, 97- 102 Received 10 July 1992
Morphological analysis of the regulation of CD4 endocytosis by ~ 5 6 ' ' ~ Mark Marsh, Pamela Reid, Ian Parsons, Christopher Hermon and Annegret Pelchen-Matthews MRC Laboratory for Molecular Cell Biology, University College London, Gower Street, London WC I E 6BT, U.K.
Introduction Strategies for employing immunotoxins or other immunoglobulin-based cytostatic agents rely on the endocytic properties of cells to deliver the reagent to a specific site in the cell, such as endosomes, lysosomes, or the trans Golgi network. The sites involved in both the processing of immunotoxins and the penetration of their toxic fragments to target sites in the cytosol are becoming increasingly well understood and should enable cell surface molecules to be selected which are best suited for effective delivery. However, the ability to select appropriate delivery molecules will depend on a detailed understanding of their endocytic and intracellular trafficking properties. Although a considerable amount has been learned about endocytosis over the last 20 years, much of this information has been gained from studies of a few well-characterized cell surface proteins. Thus the endocytosis and intracellular sorting of, for example, the receptors for transferrin, low density lipoprotein and ~
_
_
_
_
_
~
~
Abbreviation used HRP, horseradish peroxidase.
epidermal growth factor are understood in some detail [l, 2, 31. However, these well characterized receptors probably account for less than 10% of the total number of proteins expressed at the cell surface [4]. The endocytic properties of the remaining 90% are less well documented, yet these are frequently adopted as potential targets for immunotherapy [see ref. 51. We have been studying the T lymphocyte differentiation antigen CD4. This molecule, a type 1 integral membrane glycoprotein, is expressed on the surface of the T cell subsets which interact with cells expressing major histocompatibility complex class 11, primarily helperhnducer T cells, where it functions as a co-receptor to the T cell antigen receptorlCD3 complex [6]. CD4 is also expressed on some cells of the monocyte lineage, including macrophages, dendritic cells and granulocytes, although its function on these cells is unknown. In addition, CD4 is the primary receptor for the human immunodeficiency viruses, HIV- 1 and HIV2, and functions in the entry of virus into both T lymphocytes and cells of the macrophage/mono-
I992
714 1 1 7
~
~
-
~
~
~
Endocytic and transcytic pathways in Caco-2 cells J. Paul Luzio, Mark R. Jackman and Juliet A. Ellis Department of Clinical Biochemistry, University of CambridgdAddenbrooke'sHospital, Hills Road, Cambridge CB2 2QR, U.K.
cells [3, 51. Membrane proteins transcytosing in the opposite direction have been more difficult to identify, owing to the lack of suitable markers [7]. Apical to basolateral transcytosis of a ligand has been observed in the case of cobalamin, though this involves transfer from intrinsic factor to transcobalamin I1 [8]. The Caco-2 cell line has considerable potential as a model system to study transepithelial transport [9, 101. However, the exploitation of mechanisms controlling transcytosis of membrane proteins between the apical and basolateral domains for targeted drug delivery requires both a greater knowledge of endocytic pathways from each domain and the identification of proteins undergoing apical to basolateral transcytosis. In MDCK cells, a general method of identifying transcytosed membrane glycoproteins has been developed as a result of the isolation and characterization of an MDCK clone resistant to the toxic lectin of Ricinus communk (ricin). Resistance was due to a mutation that disrupted galactosylation of membrane glycoproteins, preventing ricin binding to the cell surface [ 11- 131. This ricin resistant MDCK cell line (MDCKII-RCA') was enriched in cell surface glycoconjugates bearing terminal N-acetylglucosamine residues, enabling domain specific labelling of glycoconjugates by the addition of exogenous UDP-[ 'H]galactose and galactosyltransferase. The transcytosis of membrane proteins labelled in a domain-specific manner was detected by surface biotinylation of the opposite membrane domain, detergent solubilization of these glycoproteins, adsorption onto streptavidin-agarose, and finally, separation by SDS-PAGE. This approach overcomes the need for apical or basolateral membrane marker proteins of high enough abundance to study transcytosis. However, a prerequisite for the appli-
The human colonic adenocarcinoma derived cell line, Caco-2 [l], may be grown as a polarized epithelium on permeable filter supports and develops a number of characteristic enterocyte functions when grown to confluency. The differentiated functions include the development of apical brush border microvilli, the presence of tight junctions, the unique electrical properties of transporting epithelia and polarized distribution of cell surface enzymes and receptors. Caco-2 cells are most like foetal intestinal cells in terms of expressed hydrolases, glycogen content and secretion of apolipoproteins and alphafoetoprotein [2]. When grown on filter supports the cells show polarized secretion of lipoproteins from the basolateral side [2]. They also show the presence of two membrane traffic pathways for delivery of newly synthesized membrane proteins to the apical surface [3]; one involving sorting in the trans-Golgi network analogous to the route found in Madin-Darby canine kidney (MDCK) cells [4, 51, the other involving initial transport from the trans-Golgi network to the basolateral membrane followed by transcytosis to the apical surface, similar to the route found in hepatocytes [6]. Endocytosis has proven easiest to study from the basolateral surface where many common receptors are found. One attempt to study uptake from the apical side showed mixing in a common endosoma1 compartment of apically endocytosed concanavalin A-gold with transferrin endocytosed from the basolateral side [7]. The route of delivery of newly synthesized membrane proteins such as aminopeptidase N demonstrates the existence of the basolateral to apical transcytic pathway in these Abbreviations used: MDCK, Madin-Darby kidney; KCA, Ricinus communis.
canine
717
I992
Biochemical Society Transactions
718
cation of this method to Caco-2 cells is the isolation and characterization of a ricin resistant Caco-2 (Caco-2-RCA') clone bearing a similar mutation to that seen in MDCKII-RCA' cells. Unfortunately, whilst it has proven possible to isolate ricin-resistant Caco-2 clones, none obtained to date have an appropriate mutation to allow repetition of the elegant experiments performed with MDCK (M. R. Jackman, J. A. Ellis & J. P. Luzio, unpublished work). Nonetheless, the study of the binding and uptake of ricin itself has provided some interesting new data on endocytic pathways in Caco-2 cells. Ricin was iodinated using Na"'I (50 mCi/ml, Radiochemical Centre, Amersham, U.K.) with the solid phase Iodogen technique (Pierce, Europe B.V.) [15], to a specific activity of 5 x lo6 d.p.m./pg of protein. T o assess ricin binding, cells growing in 24-well disposable trays at a density of 4 X lo6 cells/well were washed once with phosphatebuffered saline containing 0.5 mM-Ca2+, 0.5 mMg2+ (PBS+), and incubated for 1 h at 4°C with 25 ng/ml '251-labelled lectin in Dulbecco's modified Eagle's medium (without glutamine, glucose, pyruvate, phenol red and NaHCO,), 20-m~-Hepes,pH 7.4, containing 0.6% (w/v) bovine serum albumin, supplemented with increasing amounts of unlabelled lectin. Cells were then given three 5 min washes with ice-cold PBS+, and the cell bound radioactivity was measured. The amount of lectin which could not be removed from the cells with two 5 min washes with ice-cold PBS+, was used as a background value for each concentration. The amount of lectin binding per cell was calculated and plotted according to Scatchard [ 161. Endocytosis and transcytosis of [1251]ricinbound to the cell surface at 4°C were assessed on cells grown on filter supports [17] after warming cells for defined periods at 37°C and then releasing remaining or transcytosed cell surface bound ricin by the addition of 0.2 M-lactose to the apical or basolateral chamber as appropriate. Scatchard analysis showed that Caco-2 cells have 3.5 X ricin binding sites per cell and the K A for ricin binding is 12.5 x lo6 M - ' (M. R Jackman, J. A. Ellis & J. P. Luzio, unpublished work). Endocytosis of ricin from the apical and basolateral domains altered with the age of the cell layer grown on filter supports. Both apical and basolateral uptake of ricin fell between 7 and 14 days of culture. At 7 days, the percentage of membrane-bound ricin endocytosed from the basolateral surface was more than double that endocytosed from the apical surface after 30 min incubation at 37"C, whereas at 14 days the percentage
Volume 20
endocytosed from each domain was similar and approximately 80% of the apical uptake of the 7 day cells. Using 7 day cultured cell layers it was found that 3.6% of apically bound ricin was transcytosed in 2 h compared with 4.8% of basolaterally bound ricin. This contrasts with the situation in MDCK cells [15] where greater uptake of ricin from both apical and basolateral surfaces is observed and where apical to basolateral transcytosis is approximately 10-fold greater than that in the reverse direction. Endocytosis from the apical and basolateral surfaces of Caco-2 cells differs qualitatively as well as quantitatively. Apical endocytosis of ricin was reversibly inhibited by > 50% by 1 pM-cytochalasin D whereas concentrations as high as 30 p~ had little effect on basolateral endocytosis. In MDCK cells, both apical and basolateral endocytosis of ricin were similarly insensitive to cytochalasin D. Fluorescence microscopy has shown that at concentrations of cytochalasin D that affect apical endocytosis in Caco-2 cells, stress fibres are disrupted but not microvillar actin. Preliminary evidence measuring folate uptake into Caco-2 cells is consistent with some or all of the cytochalasin D effect being due to inhibition of uptake by non-coated pit mechanisms. In other cells, folate uptake has been shown to occur on a glycosyl phosphatidyl inositolanchored receptor that is internalized via caveolae [181. Whilst the investigation of ricin uptake in Caco-2 cells has yielded both qualitative and quantitative information about endocytosis from the apical and basolateral domains, it has not, to date, resulted in the identification of any individual membrane protein that is transcytosed. An alternative approach has been to search for proteins that are delivered from synthesis to both apical and basolateral surfaces. Such proteins are candidates for transcytosis if subsequently endocytosed, since they do not contain signals for domain-specific targeting. Using a polyclonal antiserum raised against a detergent extract of an apical plasma membrane fraction, five integral membrane proteins common to both apical and basolateral domains were identified. In pulse chase experiments, these five proteins were shown to be delivered to each cell surface domain with the same time course [ 171. The antiserum has been used to screen a human colon carcinoma library in the bacteriophage expression vector l g t l 1. Candidate clones encoding three of the five integral membrane proteins have been identified. It is hoped that a strategy similar to that used in the identification of Golgi-specific membrane proteins
Endocytosis, Toxins, Immunotoxins and Viruses
[I91 will prove successful in characterizing proteins Dresent on both aDical and basolateral surfaces of Caco-2 cells. Their endocytosis and transcytosis will then be studied. We thank the Cancer Research Campaign for financial support. M.R.J. is supported as a research student by Ciba-Geigy Pharmaceuticals.
1. Fogh, J., Fogh, J. M. & Orfes, T. (1977) J. Natl. Cancer Inst. 59,211-226 2. Traber, M. G., Kayden, H. J. & Rindler, M. J. (1987) J. Lipid Res. 28, 1350-1 362 3. Matter, K., Brauchbar, M., Bucher, K. & Hauri, H.-P. (1990) Cell 60,429-437 4. Lisanti, M. P., Le Bivic, A. Sargiacomo, M. & Rodriguez-Boulan, E. (1989) J. Cell Biol. 109, 21 17-2127 5. Le Bivic, A., Quaroni, A., Nichols, B. & RodriguezBoulan, E. (1990) J. Cell Biol. 111, 1351-1361 6. Bartles, J. R., Feracci, H. M., Stieger, B. & Hubbard. A. L. (1987) J. Cell Biol. 105, 1241-1251 7. Hughson, E. & Hopkins, C. R. (1990) J. Cell Biol. 110,337-348
8. Dix, C. J., Hassan, I. F., Obray, H. Y., Shah, R. & Wilson, G. (1990) Gastroenterology 98, 1272-1279 9. Hidalgo, I. J., h u b , T. J. & Borchardt, R. T. (1989) Gastroenterology 96,736-749 10. Simons, K. & Fuller, S. (1985) Ann. Rev. Cell Biol. 1, 248-288 l l . Meiss, H., creen, R. F. & ~ ~ d ~ iE. ~ ( 1982) Mol. Cell Biol. 2, 1287- 1294 12. Brandli, A. W., Hansen, G. C. & Rodriguez-Boulan. E. (1988) J. Biol. Chem. 263, 16283-16290 13. Brandli, A. ( 1991) Biochem. J. 276, 1 - 12 14. Reference deleted. 15. Fraker, P. J. & Speck, J. C. (1978) Biochem. Biophys. Res. Commun. 80,849-857 16. Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51,660 17. Ellis, J. A., Jackman, M. R. & Luzio, J. P. (1992) Biochem. J. 283,553-560 18. Rothberg, K. G., Ying, Y., Kolhouse, J. F., Kamen, B. A. & Anderson, R. G. W. (1990) J. Cell Biol. 110, 637-649 19. Luzio, J. P., Banting, G., Howell, K., Brake, B., Braghetta, P. & Stanley, K. K. (1990) Biochem. J. 270, 97- 102 Received 10 July 1992
Morphological analysis of the regulation of CD4 endocytosis by ~ 5 6 ' ' ~ Mark Marsh, Pamela Reid, Ian Parsons, Christopher Hermon and Annegret Pelchen-Matthews MRC Laboratory for Molecular Cell Biology, University College London, Gower Street, London WC I E 6BT, U.K.
Introduction Strategies for employing immunotoxins or other immunoglobulin-based cytostatic agents rely on the endocytic properties of cells to deliver the reagent to a specific site in the cell, such as endosomes, lysosomes, or the trans Golgi network. The sites involved in both the processing of immunotoxins and the penetration of their toxic fragments to target sites in the cytosol are becoming increasingly well understood and should enable cell surface molecules to be selected which are best suited for effective delivery. However, the ability to select appropriate delivery molecules will depend on a detailed understanding of their endocytic and intracellular trafficking properties. Although a considerable amount has been learned about endocytosis over the last 20 years, much of this information has been gained from studies of a few well-characterized cell surface proteins. Thus the endocytosis and intracellular sorting of, for example, the receptors for transferrin, low density lipoprotein and ~
_
_
_
_
_
~
~
Abbreviation used HRP, horseradish peroxidase.
epidermal growth factor are understood in some detail [l, 2, 31. However, these well characterized receptors probably account for less than 10% of the total number of proteins expressed at the cell surface [4]. The endocytic properties of the remaining 90% are less well documented, yet these are frequently adopted as potential targets for immunotherapy [see ref. 51. We have been studying the T lymphocyte differentiation antigen CD4. This molecule, a type 1 integral membrane glycoprotein, is expressed on the surface of the T cell subsets which interact with cells expressing major histocompatibility complex class 11, primarily helperhnducer T cells, where it functions as a co-receptor to the T cell antigen receptorlCD3 complex [6]. CD4 is also expressed on some cells of the monocyte lineage, including macrophages, dendritic cells and granulocytes, although its function on these cells is unknown. In addition, CD4 is the primary receptor for the human immunodeficiency viruses, HIV- 1 and HIV2, and functions in the entry of virus into both T lymphocytes and cells of the macrophage/mono-
I992
714 1 1 7
~
~
-
~
~
~
