Vol.
177,
June
No.
BIOCHEMICAL
2, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS Pages
14, 1991
764-770
Biosynthesis of the precursor of a soluble human insulin receptor ectodomain in insect Sf9 cells infected with a recombinant Baculovirus Janice F. Sissom’ and Leland Ellis172s3 ’ Howard
Received
Hughes Medical Institute and 2Department of Biochemistry, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75235-9050
April
15,
1991
Summary: In contrast to transfected mammalian cells, insect St9 cells infected with a recombinant Baculovirus inefficiently process and secrete a soluble derivative of the extracellular domain of the human insulin receptor. The high-mannose form of the receptor precursor that accumulates intracellularly is not grossly aberrant or malfolded, as its interaction with a diverse panel of monoclonal antibodies are comparable to secreted precursor and proteolytically processed receptor, both of which bear partially trimmed oligosaccharide chains. Thus the inefficient step in the biosynthesis of this protein in Sf9 cells is either at, or just preceding, the trimming of its high-mannose oligosaccharide chains. w1991Academic Press,1°C.
The human insulin receptor is synthesized as a single polypeptide chain precursor (o$) that is proteolytically cleaved during its biosynthesis into a- (-135~kDa) and b-subunits (-95-kDa). As the P-subunit contains the single deduced transmembrane domain of the proreceptor sequence, the transmembrane topology of mature heterotetrameric (i.e., [aP12) glycoprotein is rather simple, with a large extracellular domain that binds the hormone and a cytoplasmic protein-tyrosine kinase (l-2). Thus, truncation of the receptor prior to this deduced transmembrane domain leads to the secretion of a soluble extracellular domain (i.e., [c&]~, where PO is the extracellular portion of the P-subunit, -40~kDa) that binds insulin with high affinity, as assessed in both transfected mammalian cells (3-5), and in insect Sf9 cells infected with a recombi3 Corresponding CC@6-291X/91 Copyright All rights
0
of
author.
$1.50 I991 by Academic Press, Inc. reproduction in any form reserved.
764
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
nant Baculovirus (6-7). While proteolytic processing of the soluble receptor precursor and secretion of the mature glycosylated protein are efficient in mammalian cells, they are strikingly less so in Sf9 cells. In the latter case, a significant proportion of the protein produced accumulates intracellularly as a high-mannose (i.e., endoglycosidaseH-sensitive) precursor (6-7), while both the precursor and proteolytically processed forms of the soluble receptor that are secreted bear partially trimmed (i.e., partially endoglycosidaseH-resistant) oligosaccharide chains (6). The observed accumulation of precursor intracellularly could conceivably derive from the accumulation of a malfolded protein which is not further processed and secreted, a more subtle deficit at a requisite step proceeding the trimming of high-mannose oligosaccharide chains of the precursor, or a deficit at the level of trimming per se. As the trimming of high-mannose oligosaccharide chains per se has been reported not to be required for the secretion of other glycoproteins from these cells (8-g), we have in the present study addressed the former possibility by assessing, during the time course of its biosynthesis, the interaction of the intracellular soluble receptor with a diverse panel of receptor-specific monoclonal antibodies (10-12). Methods The recombinant Baculovirus used for the expression of a soluble secreted derivative of the extracellular domain of the human insulin receptor in insect Sf9 cells, and the methods utilized for insulin binding, metabolic labelling, immunoprecipitations and covalent cross-linking with 1251-insulin have been described in detail elsewhere (6, 12,13). Relevant experimental details are given in the Legends to each Figure. Results
and Discussion
Within 24 hr of infection of insect St9 cells with a recombinant Baculovirus encoding the extracellular domain of the human insulin receptor, insulin binding activity is detectable in the culture medium. This accumulation of activity peaks at -48 to -72 hr, and begins to decline at -96 hr post-infection, coincident with cell lysis (cf. Fig. 1 of ref. 6). Following immunoabsorption of receptors to a solid-phase support, insulin binding is detectable in non-ionic detergent (TX-100) extracts of infected Sfs cells at -14% to -38% of the maximum binding detected in the medium (cf. Fig. 1 of ref. 6; see also ref. 7). The apparent dissociation constants observed for both the intracellular and secreted forms of insulin binding are comparable to those observed for wild-type or secreted human insu765
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
lin receptors expressed in mammalian cells (i.e., -1 nM; 3,6, 13), and no detectable differences in affinity are observed over the time course of an infection with recombinant Baculovirus (from 21 to 93 hr post-infection; data not shown). The biochemical nature of receptors in both intracellular and extracellular compartments of infected cells is visualized in pulse-chase experiments, in which Sf9 cells (48 hr post-infection with virus) are metabolically-labelled with [35S]-cysteine and [35Sl-methionine for 15 min and then chased for 24 hr. Following immunoprecipitation with any of a panel of receptor-specific monoclonal antibodies, four species of receptor are observed (Fig. 1). In detergent extracts (Fig. 1, upper panel), a prominent -13%kDa non-cleaved precursor accumulates intracellularly, and a -106-kDa proteolytically cleaved a-subunit is faintly visible. Two secreted species are observed in the medium, a -120-kDa non-cleaved precursor and a processed receptor which has been proteolytically cleaved into -106-kDa a- and truncated P-subunits (Fig. 1, lower panel: the truncated -29-kDa P-subunit is not visible on these gel panels; cf. Fig. 3 of ref. 6). The oligosaccharide chains of the intracellular precursor are sensitive to digestion with endoglycosidaseH (and are thus of a highmannose type), while the carbohydrate of both of the secreted species are now partially endoglycosidaseH-resistant (thus, the mannose trees are partially trimmed; see Fig. 4 of ref. 6). Given the significant accumulation of the high-mannose precursor intracellularly (i.e., this species does not chase into the medium over a 24 hr period), does this species represent a malfolded biosynthetic intermediate (note that the pulse chase experiment indicates that this species of receptor is not rapidly degraded)? As illustrated in Fig. 1, there is no discernible difference among the eight monoclonal antibodies with respect to their recognition of either the early (intracellular, high-mannose precursor) versus late (secreted and partially endoglycosidaseH-resistant precursor or cleaved receptor) biosynthetic forms of the receptor, even with those reagents which require the native conformation of the receptor for recognition (i.e., antibodies 83.7, 18.146 and 47.9). Furthermore, in pulse-chase experiments with a 15 minute pulse followed by short periods of chase (0, 10 or 60 min: the upper, middle or lower panels of Fig. 2, respectively), or in experiments with short periods of continuous labelling without chase (1, 2.5, 5, 10 or 15 min; data not shown), no detectable difference in the recognition by these antibodies of the intracellular forms of the receptor is 766
Vol.
177,
No.
BIOCHEMICAL
2, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
. .
Omin
10 min
extract ..
0
1
6Omin
I* s.
18-34
i8-4j
1%43i8-14525-49
Antibody
medium 47-9
836
8314
.
0
B-34
18-41
2
18-4318-146
25-49
47-9
t33-7
83-14
Antibody
Fiwre 1. Recognition by monoclonal antibodies of metabolically-labelled soluble receptor expressed in SfS cells. Cells were metabolically-labelled 48 hr post-infection with 50 uCi per milliliter each of 35S-methionine and 35S-cysteine for 15 min, followed by a chase with complete medium for 24 hr. Immunoprecipitates with each of a panel of receptor-specific monoclonal antibodies (see Text) were prepared from either non-ionic detergent (Triton X-100) extracts of cells (upper panel) or culture medium (lower panel). Labelled proteins were visualized by fluorography following SDS-PAGE on a 10% gel. The positions at which each species of receptor migrates as analyzed on this gel system is indicated by a dot at the left of each panel. Top panel (intracellular species observed in cell extracts): upper dot, -138~kDa high-mannose (endoglycosidaseH-sensitive; 6) receptor precursor; lower dot, - 106-kDa processed asubunit. Bottom panel (secreted species observed in culture medium): top dot, -120-kDa non-cleaved precursor; bottom dot, -106- kDa processed a-subunit. Both of these secreted species have partially endogly-
cosidaseH-resistant oligosaccharide chains (6). Fiare 2. Analysis of precursor biosynthesis during pulse-chase experiments. Cells were metabolically-labelled and receptors analyzed as described in the legend of Fig. 1, except that cells were chased for 0, 10 or 60 min. The positions of the intracellular precursor and the pro-
teolytically processed a-subunit are indicated at the left of each panel by the top and bottom dots, respectively
observed during the time course of their biosynthesis. This is confirmed by the use of a set of five polyclonal anti-receptor antibodies, which recognize both native and denatured receptor (11; data not shown). Note that while the -13%kDa non-cleaved precursor does not chase significantly from 0 to 60 min, the -106kDa proteolytically cleaved a-subunit does so (and is presumably in transit out of the cell). Thus, as assessed by the use of this diverse panel of immunoreagents, neither slowly folding receptor intermediates nor grossly malfolded species are observed. Covalent cross-linking with 1251-insulin of immune complexes prepared from culture medium of infected St9 cells reveals radiolabelled 767
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
extract medium
Figure 3. Covalently cross-linking of ‘251-insdin to intraceblar and secreted receptors. Receptors from non-ionic detergent extracts (upper panel) or culture medium (lower panel) of St9 cells 48 hr post-infection were immunoprecipitated with monoclonal antibody 18.41. Immune complexes were incubated with 1251-insulin (monoiodinated porcine insulin, 80 to 120 pCi/pg, DuPont-NEN) in the presence of increasing concentrations of unlabelled porcine insulin (Lily; 0,5 x 10-11, 1 x lo-‘, 5 x lo-‘, lx 10m8, 5 x 10T8, 1 x lo-‘, 5 x 10m7 or 1 x 10e6 M in lanes 1 to 9, respectively), cross-linked with disuccinimidyl suberate and analyzed by SDS-PAGE and radioautography (12-13). Top panel (intracellular species observed in cell extracts): upper dot, -138kDa precursor; lower dot, -106-kDa precessed a-subunit. Bottom panel (secreted species observed in culture medium): top dot, -120~kDa non-cleaved precursor; bottom dot, -106~kDa processed a-subunit.
bands which correspond to the cleaved receptor, as well as the noncleaved precursor (the lower panel of Fig. 3). In cell extracts (the upper panel of Fig. 31, a prominent labelling of the precursor is observed, as well as a small amount of the processed ol-subunit4. The specific nature of this covalent cross-linking is demonstrated by the competition of labelling upon incubating such reactions with increasing concentrations of unlabelled insulin (see the Legend to Fig. 3). Comparable apparent dissociation constants t&s) are observed for all four receptor species. However, note that less labelling of the intracellular non-cleaved precursor species is observed, as compared to that of the two forms of the secreted receptor (cf. Fig. 1 and 31, which suggests that only a subset of the total 4 In a previous study, covalent cross-linking with 1251-insulin of partially purified (by wheat germ lectin afhnity and gel filtration chromatography) secreted receptors in solution resulted in the specific labelling of only the a-subunit of the cleaved receptor (and not the secreted precursor; cf. Fig. 6 of ref. 61, while no cross-linked species was visible following incubation of total non-ionic detergent cell extracts with iodinated hormone. As both we (the present study) and others (7) observe 1251-insulin covalently cross-linked to both intracellular cleaved and non-cleaved soluble receptors when bound to immune complexes, this difference is presumed to follow from to the greater efficiency of cross-linking following immunopurification of receptors via receptor-specific monoclonal antibodies. 768
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
intracellular precursor protein has acquired the final prerequisites for the constitution of a high-affininty insulin binding site. Such a “transition state” has also been observed during the biosynthesis of the wildtype insulin receptor in mammalian cells (14). While the molecular correlate of this “state” is not known, the present results in this cell system suggest that this transition is required for both the full acquisition of insulin binding, as well for further movement of the protein within intracellular compartments and trimming of high-mannose oligosaccharide chains. On the one hand, the more efficient overall biosynthesis of this protein observed in mammalian cells may reflect the fact that the Sf9 cells are sick due to the lytic viral infection. However, the fact that significantly greater levels of a soluble protein-tyrosine kinase domain of the receptor can be recovered from the cytoplasm of Sf9 cells (15-17) suggests that there may be distinct differences in the ability of this cell system to overexpress cytosolic versus secreted or transmembrane proteins that must traverse intracellular membrane compartments. Acknowledgments Many thanks to our colleagues in the Ellis lab (esp. Drs. Erik Schaefer and Jeremy Tavard), and to Dr. Flora Katz (UT Southwestern), for their advice and helpful discussions during the course of this work. This research was supported by the Howard Hughes Medical Institute (LE) and a grant from the Robert A. Welch Foundation (LE). References (1) Ebina, Y., Ellis, L., Jarnigan, K, Edery, M., Graf, L., Clauser, E., Ou, J., Masiarz, F., Kan, YW, Goldfine, I.D., Roth, R.A., Rutter, W. J. (1985) Cell 40,747-758 (2) Ullrich, A., Bell, J.R., Chen, E.Y., Herrera, R., Petruzzelli, L.M., Dull, T. J., Gray, A., Coussens, L., Liao, Y.-C., Tsubokawa, M., Mason, A., Seeburg, PH., Grunfeld, C., Rosen, O.M., and Ramachandran, J. (1985) Nature 313,756-761 (3) Ellis, L., Sissom, J., and Levitan, 1, 25-31
A. (1988) J. Molecular
Recognition
(4) Johnson, J.D., Wong, M.L., and Rutter, W.J. (1988) Proc. Natl. Acad. Sci U.S.A. 85, 7516-7520 (5) Whittaker,
J., and Okamoto, A. (1988) J. Biol. Chem. 263,3063-3066
(6) Sissom, J., and Ellis, L. (1989) Biochem. J. 261, 119-126 769
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
(7) Paul, J.I., Tavare, J., Denton, R.M., and Steiner, D.F. (1990) J. Biol. Chem. 265,13074-13083 (8) Jarvis, D.L., and Summers, M.D. (1989) Mol. Cell Biol. 9, 214-223 (9) Jarvis, D.L., Oker-Blum, them. 42,181-191
C., and Summers,
M.D. (1990) J. Cell. Bio-
(10) Soos, M.A., Siddle, K., Baron, M.D., Heward, J.M., Luzio, J.P., Bellatin, J., and Lennox, E.S. (1986) Biochem. J. 235, 199-203 (11) Prigent, S.A., Stanley, K.K., and Siddle, K. (1990) J. Biol. Chem. 265,9970-9977 (12) Schaefer, E.M., Siddle, K., and Ellis, L. (1990) J. Biol. Chem. 265, 13248-13253 (13) Ellis, L., Clauser, E., Morgan, ter, W.J. (1986) Cell 45,721-732 (14) Olson, T.S., Bamberger, 263,7342-7351 (15) Ellis, L., Levitan, 1634-1639
D.O., Edery, M., Roth, R.A., and Rut-
M. J., and Lane, M.D. (1988) J. Biol. Chem.
A., Cobb, M.H., and Ramos, I? (1988) J. Virol. 62,
(16) Levine, B.A., Clack, B., and Ellis, L. (1991) J. Biol. Chem. 266,35653570 (17) Tavare, 1390-1395
J.M., Clack, B., and Ellis, L. (1991) J. Biol. Chem. 266,
770
177,
June
No.
BIOCHEMICAL
2, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS Pages
14, 1991
764-770
Biosynthesis of the precursor of a soluble human insulin receptor ectodomain in insect Sf9 cells infected with a recombinant Baculovirus Janice F. Sissom’ and Leland Ellis172s3 ’ Howard
Received
Hughes Medical Institute and 2Department of Biochemistry, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75235-9050
April
15,
1991
Summary: In contrast to transfected mammalian cells, insect St9 cells infected with a recombinant Baculovirus inefficiently process and secrete a soluble derivative of the extracellular domain of the human insulin receptor. The high-mannose form of the receptor precursor that accumulates intracellularly is not grossly aberrant or malfolded, as its interaction with a diverse panel of monoclonal antibodies are comparable to secreted precursor and proteolytically processed receptor, both of which bear partially trimmed oligosaccharide chains. Thus the inefficient step in the biosynthesis of this protein in Sf9 cells is either at, or just preceding, the trimming of its high-mannose oligosaccharide chains. w1991Academic Press,1°C.
The human insulin receptor is synthesized as a single polypeptide chain precursor (o$) that is proteolytically cleaved during its biosynthesis into a- (-135~kDa) and b-subunits (-95-kDa). As the P-subunit contains the single deduced transmembrane domain of the proreceptor sequence, the transmembrane topology of mature heterotetrameric (i.e., [aP12) glycoprotein is rather simple, with a large extracellular domain that binds the hormone and a cytoplasmic protein-tyrosine kinase (l-2). Thus, truncation of the receptor prior to this deduced transmembrane domain leads to the secretion of a soluble extracellular domain (i.e., [c&]~, where PO is the extracellular portion of the P-subunit, -40~kDa) that binds insulin with high affinity, as assessed in both transfected mammalian cells (3-5), and in insect Sf9 cells infected with a recombi3 Corresponding CC@6-291X/91 Copyright All rights
0
of
author.
$1.50 I991 by Academic Press, Inc. reproduction in any form reserved.
764
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
nant Baculovirus (6-7). While proteolytic processing of the soluble receptor precursor and secretion of the mature glycosylated protein are efficient in mammalian cells, they are strikingly less so in Sf9 cells. In the latter case, a significant proportion of the protein produced accumulates intracellularly as a high-mannose (i.e., endoglycosidaseH-sensitive) precursor (6-7), while both the precursor and proteolytically processed forms of the soluble receptor that are secreted bear partially trimmed (i.e., partially endoglycosidaseH-resistant) oligosaccharide chains (6). The observed accumulation of precursor intracellularly could conceivably derive from the accumulation of a malfolded protein which is not further processed and secreted, a more subtle deficit at a requisite step proceeding the trimming of high-mannose oligosaccharide chains of the precursor, or a deficit at the level of trimming per se. As the trimming of high-mannose oligosaccharide chains per se has been reported not to be required for the secretion of other glycoproteins from these cells (8-g), we have in the present study addressed the former possibility by assessing, during the time course of its biosynthesis, the interaction of the intracellular soluble receptor with a diverse panel of receptor-specific monoclonal antibodies (10-12). Methods The recombinant Baculovirus used for the expression of a soluble secreted derivative of the extracellular domain of the human insulin receptor in insect Sf9 cells, and the methods utilized for insulin binding, metabolic labelling, immunoprecipitations and covalent cross-linking with 1251-insulin have been described in detail elsewhere (6, 12,13). Relevant experimental details are given in the Legends to each Figure. Results
and Discussion
Within 24 hr of infection of insect St9 cells with a recombinant Baculovirus encoding the extracellular domain of the human insulin receptor, insulin binding activity is detectable in the culture medium. This accumulation of activity peaks at -48 to -72 hr, and begins to decline at -96 hr post-infection, coincident with cell lysis (cf. Fig. 1 of ref. 6). Following immunoabsorption of receptors to a solid-phase support, insulin binding is detectable in non-ionic detergent (TX-100) extracts of infected Sfs cells at -14% to -38% of the maximum binding detected in the medium (cf. Fig. 1 of ref. 6; see also ref. 7). The apparent dissociation constants observed for both the intracellular and secreted forms of insulin binding are comparable to those observed for wild-type or secreted human insu765
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
lin receptors expressed in mammalian cells (i.e., -1 nM; 3,6, 13), and no detectable differences in affinity are observed over the time course of an infection with recombinant Baculovirus (from 21 to 93 hr post-infection; data not shown). The biochemical nature of receptors in both intracellular and extracellular compartments of infected cells is visualized in pulse-chase experiments, in which Sf9 cells (48 hr post-infection with virus) are metabolically-labelled with [35S]-cysteine and [35Sl-methionine for 15 min and then chased for 24 hr. Following immunoprecipitation with any of a panel of receptor-specific monoclonal antibodies, four species of receptor are observed (Fig. 1). In detergent extracts (Fig. 1, upper panel), a prominent -13%kDa non-cleaved precursor accumulates intracellularly, and a -106-kDa proteolytically cleaved a-subunit is faintly visible. Two secreted species are observed in the medium, a -120-kDa non-cleaved precursor and a processed receptor which has been proteolytically cleaved into -106-kDa a- and truncated P-subunits (Fig. 1, lower panel: the truncated -29-kDa P-subunit is not visible on these gel panels; cf. Fig. 3 of ref. 6). The oligosaccharide chains of the intracellular precursor are sensitive to digestion with endoglycosidaseH (and are thus of a highmannose type), while the carbohydrate of both of the secreted species are now partially endoglycosidaseH-resistant (thus, the mannose trees are partially trimmed; see Fig. 4 of ref. 6). Given the significant accumulation of the high-mannose precursor intracellularly (i.e., this species does not chase into the medium over a 24 hr period), does this species represent a malfolded biosynthetic intermediate (note that the pulse chase experiment indicates that this species of receptor is not rapidly degraded)? As illustrated in Fig. 1, there is no discernible difference among the eight monoclonal antibodies with respect to their recognition of either the early (intracellular, high-mannose precursor) versus late (secreted and partially endoglycosidaseH-resistant precursor or cleaved receptor) biosynthetic forms of the receptor, even with those reagents which require the native conformation of the receptor for recognition (i.e., antibodies 83.7, 18.146 and 47.9). Furthermore, in pulse-chase experiments with a 15 minute pulse followed by short periods of chase (0, 10 or 60 min: the upper, middle or lower panels of Fig. 2, respectively), or in experiments with short periods of continuous labelling without chase (1, 2.5, 5, 10 or 15 min; data not shown), no detectable difference in the recognition by these antibodies of the intracellular forms of the receptor is 766
Vol.
177,
No.
BIOCHEMICAL
2, 1991
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
. .
Omin
10 min
extract ..
0
1
6Omin
I* s.
18-34
i8-4j
1%43i8-14525-49
Antibody
medium 47-9
836
8314
.
0
B-34
18-41
2
18-4318-146
25-49
47-9
t33-7
83-14
Antibody
Fiwre 1. Recognition by monoclonal antibodies of metabolically-labelled soluble receptor expressed in SfS cells. Cells were metabolically-labelled 48 hr post-infection with 50 uCi per milliliter each of 35S-methionine and 35S-cysteine for 15 min, followed by a chase with complete medium for 24 hr. Immunoprecipitates with each of a panel of receptor-specific monoclonal antibodies (see Text) were prepared from either non-ionic detergent (Triton X-100) extracts of cells (upper panel) or culture medium (lower panel). Labelled proteins were visualized by fluorography following SDS-PAGE on a 10% gel. The positions at which each species of receptor migrates as analyzed on this gel system is indicated by a dot at the left of each panel. Top panel (intracellular species observed in cell extracts): upper dot, -138~kDa high-mannose (endoglycosidaseH-sensitive; 6) receptor precursor; lower dot, - 106-kDa processed asubunit. Bottom panel (secreted species observed in culture medium): top dot, -120-kDa non-cleaved precursor; bottom dot, -106- kDa processed a-subunit. Both of these secreted species have partially endogly-
cosidaseH-resistant oligosaccharide chains (6). Fiare 2. Analysis of precursor biosynthesis during pulse-chase experiments. Cells were metabolically-labelled and receptors analyzed as described in the legend of Fig. 1, except that cells were chased for 0, 10 or 60 min. The positions of the intracellular precursor and the pro-
teolytically processed a-subunit are indicated at the left of each panel by the top and bottom dots, respectively
observed during the time course of their biosynthesis. This is confirmed by the use of a set of five polyclonal anti-receptor antibodies, which recognize both native and denatured receptor (11; data not shown). Note that while the -13%kDa non-cleaved precursor does not chase significantly from 0 to 60 min, the -106kDa proteolytically cleaved a-subunit does so (and is presumably in transit out of the cell). Thus, as assessed by the use of this diverse panel of immunoreagents, neither slowly folding receptor intermediates nor grossly malfolded species are observed. Covalent cross-linking with 1251-insulin of immune complexes prepared from culture medium of infected St9 cells reveals radiolabelled 767
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
extract medium
Figure 3. Covalently cross-linking of ‘251-insdin to intraceblar and secreted receptors. Receptors from non-ionic detergent extracts (upper panel) or culture medium (lower panel) of St9 cells 48 hr post-infection were immunoprecipitated with monoclonal antibody 18.41. Immune complexes were incubated with 1251-insulin (monoiodinated porcine insulin, 80 to 120 pCi/pg, DuPont-NEN) in the presence of increasing concentrations of unlabelled porcine insulin (Lily; 0,5 x 10-11, 1 x lo-‘, 5 x lo-‘, lx 10m8, 5 x 10T8, 1 x lo-‘, 5 x 10m7 or 1 x 10e6 M in lanes 1 to 9, respectively), cross-linked with disuccinimidyl suberate and analyzed by SDS-PAGE and radioautography (12-13). Top panel (intracellular species observed in cell extracts): upper dot, -138kDa precursor; lower dot, -106-kDa precessed a-subunit. Bottom panel (secreted species observed in culture medium): top dot, -120~kDa non-cleaved precursor; bottom dot, -106~kDa processed a-subunit.
bands which correspond to the cleaved receptor, as well as the noncleaved precursor (the lower panel of Fig. 3). In cell extracts (the upper panel of Fig. 31, a prominent labelling of the precursor is observed, as well as a small amount of the processed ol-subunit4. The specific nature of this covalent cross-linking is demonstrated by the competition of labelling upon incubating such reactions with increasing concentrations of unlabelled insulin (see the Legend to Fig. 3). Comparable apparent dissociation constants t&s) are observed for all four receptor species. However, note that less labelling of the intracellular non-cleaved precursor species is observed, as compared to that of the two forms of the secreted receptor (cf. Fig. 1 and 31, which suggests that only a subset of the total 4 In a previous study, covalent cross-linking with 1251-insulin of partially purified (by wheat germ lectin afhnity and gel filtration chromatography) secreted receptors in solution resulted in the specific labelling of only the a-subunit of the cleaved receptor (and not the secreted precursor; cf. Fig. 6 of ref. 61, while no cross-linked species was visible following incubation of total non-ionic detergent cell extracts with iodinated hormone. As both we (the present study) and others (7) observe 1251-insulin covalently cross-linked to both intracellular cleaved and non-cleaved soluble receptors when bound to immune complexes, this difference is presumed to follow from to the greater efficiency of cross-linking following immunopurification of receptors via receptor-specific monoclonal antibodies. 768
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
intracellular precursor protein has acquired the final prerequisites for the constitution of a high-affininty insulin binding site. Such a “transition state” has also been observed during the biosynthesis of the wildtype insulin receptor in mammalian cells (14). While the molecular correlate of this “state” is not known, the present results in this cell system suggest that this transition is required for both the full acquisition of insulin binding, as well for further movement of the protein within intracellular compartments and trimming of high-mannose oligosaccharide chains. On the one hand, the more efficient overall biosynthesis of this protein observed in mammalian cells may reflect the fact that the Sf9 cells are sick due to the lytic viral infection. However, the fact that significantly greater levels of a soluble protein-tyrosine kinase domain of the receptor can be recovered from the cytoplasm of Sf9 cells (15-17) suggests that there may be distinct differences in the ability of this cell system to overexpress cytosolic versus secreted or transmembrane proteins that must traverse intracellular membrane compartments. Acknowledgments Many thanks to our colleagues in the Ellis lab (esp. Drs. Erik Schaefer and Jeremy Tavard), and to Dr. Flora Katz (UT Southwestern), for their advice and helpful discussions during the course of this work. This research was supported by the Howard Hughes Medical Institute (LE) and a grant from the Robert A. Welch Foundation (LE). References (1) Ebina, Y., Ellis, L., Jarnigan, K, Edery, M., Graf, L., Clauser, E., Ou, J., Masiarz, F., Kan, YW, Goldfine, I.D., Roth, R.A., Rutter, W. J. (1985) Cell 40,747-758 (2) Ullrich, A., Bell, J.R., Chen, E.Y., Herrera, R., Petruzzelli, L.M., Dull, T. J., Gray, A., Coussens, L., Liao, Y.-C., Tsubokawa, M., Mason, A., Seeburg, PH., Grunfeld, C., Rosen, O.M., and Ramachandran, J. (1985) Nature 313,756-761 (3) Ellis, L., Sissom, J., and Levitan, 1, 25-31
A. (1988) J. Molecular
Recognition
(4) Johnson, J.D., Wong, M.L., and Rutter, W.J. (1988) Proc. Natl. Acad. Sci U.S.A. 85, 7516-7520 (5) Whittaker,
J., and Okamoto, A. (1988) J. Biol. Chem. 263,3063-3066
(6) Sissom, J., and Ellis, L. (1989) Biochem. J. 261, 119-126 769
Vol.
177,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
(7) Paul, J.I., Tavare, J., Denton, R.M., and Steiner, D.F. (1990) J. Biol. Chem. 265,13074-13083 (8) Jarvis, D.L., and Summers, M.D. (1989) Mol. Cell Biol. 9, 214-223 (9) Jarvis, D.L., Oker-Blum, them. 42,181-191
C., and Summers,
M.D. (1990) J. Cell. Bio-
(10) Soos, M.A., Siddle, K., Baron, M.D., Heward, J.M., Luzio, J.P., Bellatin, J., and Lennox, E.S. (1986) Biochem. J. 235, 199-203 (11) Prigent, S.A., Stanley, K.K., and Siddle, K. (1990) J. Biol. Chem. 265,9970-9977 (12) Schaefer, E.M., Siddle, K., and Ellis, L. (1990) J. Biol. Chem. 265, 13248-13253 (13) Ellis, L., Clauser, E., Morgan, ter, W.J. (1986) Cell 45,721-732 (14) Olson, T.S., Bamberger, 263,7342-7351 (15) Ellis, L., Levitan, 1634-1639
D.O., Edery, M., Roth, R.A., and Rut-
M. J., and Lane, M.D. (1988) J. Biol. Chem.
A., Cobb, M.H., and Ramos, I? (1988) J. Virol. 62,
(16) Levine, B.A., Clack, B., and Ellis, L. (1991) J. Biol. Chem. 266,35653570 (17) Tavare, 1390-1395
J.M., Clack, B., and Ellis, L. (1991) J. Biol. Chem. 266,
770
