- Email: [email protected]
Coordinate synthesis and degradation of the α-, β- and γ-subunits of the receptor for immunoglobulin E
Mo/ecular Immunology,Vol. 22, No. 9, pp. 1045-1051, 1985 Printedin Great Britain
0161-%X90/85 $3.00 + 0.00 PergamonPress Ltd
COORDINATE SYNTHESIS AND DEGRADATION THE a-, ,9- AND y-SUBUNITS OF THE RECEPTOR FOR IMMUNOGLOBULIN E
OF
RODOLFO QUARTO, JEAN-PIERRE KINET and HENRY METZGER Section on Chemical Immunology, Arthritis and Rheumatism Branch, National Institute of Arthritis, Diabetes, and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD 20205, U.S.A.
(First received 6 December 1984; accepted in revised form 11 February 1985) Abstract-The surface receptor for immunoglobulin E (IgE) on rat basophilic leukemia cells and their normal counterparts has been postulated to consist of four polypeptide chains: a 45-kDa a-chain which binds IgE, a 33-kDa p-component and two disulfide-linked, 9-lo-kDa y-polypeptides. The instability of this complex in mild detergents makes it possible that, in vivo also, the structure may not be stable and that there is an independent assembly or exchange of the chains. We studied this question using surface-labeling and biosynthetic labeling techniques and found that the chains turn over coordinately and do not independently exchange. The results provide further support for the proposal that the aby, complex is the unit receptor for IgE.
INTRODUCTION
Mast cells and analogues of such cells contain a surface glycoprotein that tightly binds monomeric immunoglobulin E (IgE). This glycoprotein has a mol. wt of approx. 45,000 (Conrad and Froese, 1976; Kulczycki et al., 1976; Kumar and Metzger, 1982+30x of which is accounted for by carbohydrate (Kanellopoulos et al., 1980). When cells bearing this protein are solubilized with mild detergents and the IgE-binding component is isolated by affinity chromatography in solvents containing micellar detergent and analyzed by electrophoresis on polyacrylamide gels, only the 45kDa protein is observed. This holds whether the cells have been labeled biosynthetically or by external iodination. Small amounts of label can be irregularly observed in other components (Kulczycki and Parker, 1979; Froese, 1980; Helm and Froese, 1981), but it is not possible to know whether these are contaminants or components that relate to the IgE-binding protein. The use of cross-linking reagents first gave direct evidence that at least one other component was associated with the 45kDa protein in significant amounts (Holowka et al., 1980; Holowka and Metzger, 1982). Subsequently, it was found that, even in the absence of cross-linking reagents, this second, 33-kDa, component co-purified with the 45kDa glycoprotein providing the solutions contained a detergent:lipid ratio similar to that in the original extract of the cells (Rivnay et al., 1982). Ultimately, a third component was observed under these conditions (Perez-Montfort et al., 1983), leading to the proposal that the receptor for IgE consisted of four polypeptides: the 45kDa IgE-binding glycoprotein (GL), a 33-kDa b-chain and two disulfide-linked, 9-lo-kDa, y-chains (Perez-Montfort et al., 1983; Metzger et al., 1984).
Although the stoichiometric data strongly suggest that all of these chains are integral parts of the receptor, the unusual instability of the complex makes it possible that in situ the component peptides can turn over independently. Indeed, in studies in which cells were incubated with 32Pi, we observed labeling of the /?-chains under conditions in which we were able to selectively examine receptors already incorporated into the plasma membrane (Fewtrell et al., 1982). The alternative explanations of those findings were that either chains already incorporated into the membrane could be modified, or that the /Iand y-chains could turn over independently of the a-chains. The present study was undertaken to explore this matter further. MATERIALS AND METHODS
IgE and cells
Rat IgE [IR 162 (Bazin et al., 1974)] and mouse anti-dinitrophenyl IgE [Hi-DNP-e-26.82 (Liu et al., purified as described previously 1980)] were (Kulczycki and Metzger, 1974; Holowka and Metzger, 1982). Iodinations were performed with chloramine-T (McConahey and Dixon, 1966) or with Iodo-beads (Markwell, 1982) obtained from Pierce (Rockford, IL). Arsonylated IgE was prepared as described by Kanellopoulos et al. (1979). Rat basophilic leukemia cells were maintained as described by Barsumian et al. (1981). Binding studies
Cells were saturated with iodinated IgE and analyzed for cell-bound IgE as described previously (Kulczycki and Metzger, 1974). The cells had an average of 3 x lo5 receptors in all the studies described here. 1045
1046
ROD~LFO QUARTO et al.
Labeling of cells Surface iodination. The labeling by lactoperoxidase-catalyzed iodination of cells from stationary cultures has been described previously (Marchalonis, 1969; Kanellopoulos et al., 1979). Alternatively we used Iodo-gen (Markwell and Fox, 1978). Aliquots of 5 x 10’ cells in phosphate-buffered saline were added to a glass vial containing 3OOpg of Iodo-gen and then 500 PCi of carrier-free Na”‘I (Amersham Corp., Arlington, IL) was added. After a 15-min incubation on ice with occasional shaking, the cells were transferred to a tube containing 5 ml complete medium and then processed as described previously (Kanellopoulos et al., 1979). Intrinsic labeling. The method described by PerezMontfort et al. (1983) was used for incorporating L-[3H]leucine (Amersham Corp., code TRK 5 10). For the chase experiments, the washed cells were incubated with IgE where desired and then maintained in complete medium in spinner flasks. PuriJication of receptors
In all cases the receptors were isolated on the basis of the IgE bound to them. Cells were solubilized at 5 x 10’ cells/ml with the detergent 3-[(3-cholamidopropyl)dimethylammonio]-l-propane sulfonate (10 mM) (Pierce) (Hjelmeland, 1980) and protease inhibitors as previously described (Rivnay et al., 1982). The same detergent was used throughout this study. Surface-labeled receptors. These were purified simply by immunoprecipitation using purified rabbit antibodies to IgE, as described previously (PerezMontfort et al., 1983). Biosyntheticaiiy labeled receptors. These were purified essentially as described previously (PerezMontfort et al., 1983). The IgE-receptor complexes were adsorbed either to trinitrophenyl-lysine on Biogel A5m beads (Holowka and Metzger, 1982) or to anti-benzene arsonate-Sepharose 4B (Kanellopoulos et aI., 1979) and were extensively washed with 2 mM detergent in borate-buffered saline, pH 8.0, at 4°C (Alcaraz et al., 1984; Kinet et al., in press). They were then eluted with the appropriate hapten {lOmM dinitrophenyl-e-NH, caproate or 1 mM [(p-arsonicnitrophenyl)-azoltyrosine} in the same buffer. In all subsequent steps the same buffer was used. The complexes were then immunoprecipitated and analyzed by electrophoresis in polyacrylamide gels in sodium dodecyl sulfate as described by PerezMontfort et al. (1983). The ratio z:/I :y was determined by densitometry of autoradiographs of the gels or by slicing the dried slabs and counting in a scintillation counter. RESULTS
Our basic strategy was to develop an estimate of the rates of synthesis and degradation of the IgEbinding a-chain of the receptor and then to compare
these rates with the rates of turnover of the fl- and y-chains. Turnover of the u-chains
The plasma membrane proteins of the cells were surface-labeled with 12’I.Under these conditions only the a-chains of the receptor are labeled even though it can be shown that the /I- and y-chains are present (Perez-Montfort et al., 1983). The cells were then treated in three different ways. One aliquot was reacted with “‘I-labeled mouse IgE and thoroughly washed to remove the unbound IgE. The cells were then incubated at 37°C and sampled periodically. The concns of the reactants were such that saturation of the receptors would be expected (Kulczycki and Metzger, 1974). A second aliquot was similarly reacted with 13’1-IgE but not washed. These cells were then incubated at 37°C in the continuous presence of the labeled IgE and sampled at intervals. The IgE in the medium was sufficient to maintain saturation of the receptors given the K, of IgE for the receptor. A third aliquot was not reacted with IgE before or during the incubation. Samples of these cells were reacted with 13’I-IgE just prior to analysis. The conditions used (legend of Fig. 1) were adequate to ensure saturation of the receptors. The cells in all samples were washed, solubilized with detergent, centrifuged and the supernatant solutions subjected to a “clearing” immunoprecipitation with human IgE and anti human IgE. The supernatants were reacted with anti-mouse IgE and the precipitates were then analyzed by electrophoresis in polyacrylamide gels in sodium dodecyl sulfate. The gels were sliced and the counts in ‘3’I-IgE and ‘251-cr-chains determined. The data from one such experiment are shown in Fig. lA, and illustrate the change with time of the ‘25I-c(:“‘I-IgE ratio in the samples from the three incubation mixtures. The cells that had been loaded with labeled IgE prior to the incubation and then incubated in the absence of IgE show an invariant ratio of ‘25I-~: “‘I-IgE (triangles). These specimens serve as the control or baseline. Because here only those receptors that had bound IgE prior to the incubation were examined, there is no apriori reason why the ratio should have changed. On the other hand, the cells incubated in the continuous presence of labeled IgE show a moderate decline in the ‘25I-c(: 13’I-IgE ratio over time (circles). This would be expected if during the course of the incubation, new and, therefore, unlabeled receptors had appeared on the surface of the cells where they could bind the 13’1-IgE present in the medium. The data in Fig. 1B show that these cells in fact showed an increase in total receptors and this increase in total receptors accounts quantitatively rather well for the decrease in the ‘25I-a:13’1-IgE ratio. Thus, at 24 hr, the 35% decrease in the “‘1-a:IgE ratio (Fig. 1A) is accounted for by the 35% increase in receptors (Fig. 1B).
Turnover 1.0
of IgE receptor
1047
Table 1. Ratio of counts in ‘2SI-labeled or-chains to counts in “‘I-labeled IgE on cells incubated in the presence or absence of “‘I-1gE in the medium
d tw
Time of incubation (hr)
-
0 6 12 24
-
1
A
4-l
-0 I
TIME (hrl Fig. 1. Analysis of surface-labelled cells. Cells were surface iodinated with lzsI and then treated in 1 of 3 ways. One aliquot was reacted with “‘I-IgE (3 x 10’ cells/ml, 10 pg/ml IgE) for 30-40 min and maintained in 13’I-IgE (1 pg/ml) throughout the 24-hr incubation (0). A second aliquot was similarly reacted with 13’I-IgE and then washed to remove all unbound IgE prior to the incubation (A). The third aliquot was reacted for 60 min with 10 pg/ml 13’1-IgE only after samples from the incubation were removed just prior to analysis (0). All incubations contained -4 x lo6 cells/ml. The cells in each sample were washed, solubilized with detergent and the receptors immunoprecipitated before being analyzed on gels. A portion of each incubation mixture was also analyzed for cell number and bound “‘I-IgE. (A) Ratio of counts in ‘251-labeleda-chains to the counts in the “‘I-IgE. For each incubation the data were normalized by dividing the ratio determined at each time point by the ratio determined at time zero. The time in hours is the time elapsed after the surface labeling was terminated. mixture calcu(B) Ratio of bound i3’I-IgE/ml incubation lated from the counts of cells/ml and “‘I-IgE bound/cell. The ratios were normalized for each incubation mixture as noted in A. All counts were performed in triplicate. The actual cell counts/ml at zero time and 24 hr were 3.8 x lo6 and 3.4 x lo6 (O), 3.9 x lo6 and 4.5 x lo6 (A), and 4.5 x lo6 and 4.8 x lo6 (Cl).
Normalized ratio” + IgE
I 0.948 (0.145) 0.845 (0.058) 0.465 (0.158)
- IgE 1 0.791 (0.139) 0.507 (0.101) 0.257 (0.1 30)h
“Except as noted the values shown are the averages from four separate experiments. The values in the parentheses are the SDS. “Average from three experiments only. The counts in the a-chains in the fourth experiment were too low to measure accurately.
such a comparison was made are collected in Table 1. Clearly, the difference between the two types of incubations is reproducible. In several of these experiments the total cells/ml and the IgE bound/cell was assessed throughout the incubation and from these data the cell-bound IgE per sample could be calculated such as illustrated in Fig. 1B. Although there was some variability in the absolute values, the receptors/ml in the incubations without IgE never substantially increased and usually declined slightly over the period studied. In four successive experiments the ratio of receptors/ml at 24 hr to that at zero time was 0.89, 1.4, 1.07 and 0.83 respectively. Therefore, the decrease in the ‘*‘I-cr: ‘-“I-IgE ratio in these samples cannot simply be due to the acquisition of newly-synthesized and therefore unlabeled a-chains. There must also have been a corresponding degradation of the chains. Contrariwise, the incubations containing free IgE showed about a 25% increase in the IgE-binding capacity (Fig. 1B). The cells preloaded with IgE but not grown in the presence of IgE were only analyzed for residual bound IgE and not for their total IgE-binding capacity. As shown in Fig. 1B (triangles) there was a moderate decline in the cell-bound IgE, due either to degradation of the bound IgE or dissociation of the IgE, or both. Together, these results are most simply explained by a model in which new cz-chains appear on the surface of the cells in the presence or absence of IgE in the medium, but that those cc-chains to which IgE is bound are less labile than those that are empty. In any case, these observations indicate that, over the period studied, a substantial fraction of those cc-chains that are not bound with IgE are degraded and new cc-chains are taking their place. Turnover of the fl- and y-chains
A decline in the ‘*%Y: 13’1-IgEratio is also expected for the cells grown in the absence of IgE, but what is striking is the more rapid decrease in the ratio for these specimens (squares) compared to those in which the labeled IgE was present continuously (circles). The combined results from four experiments in which
in order to examine the turnover of the p- and labeling with y-chains we used biosynthetic [3H]leucine. Incorporation experiments. Approximately 30% of the receptors on the cells were occupied with arsonylated ‘3’I-labeled rat IgE. The cells were thoroughly washed to remove unbound IgE and then
RODOLFO QUARTO et al.
1048
BIOSYNTHETIC LABELING OF LIGANDED AND UNLIGANDED RECEPTORS ON THE SAME CELLS
pamally
saturate
wtth
1251 (Ars)
receptors rat IgE
I wash
cells
I suspend
cells I” lewoe-free
medum
+ sH-leucine
I sample
at 24, 48 hr +12WArs)
no addition A solubillze
mouse
IgE
cells
I
I purify
IgE receptors
on anti-Ars-Sepharose
+ anti-mouse
+ anti rat IgE
IgE
supernatant
analyze (A)
ppt
1
1 + anti-rat
analyze
ppts
IBI
IgE
(Cl
Fig. 2. Flow diagram of the experimental procedure used to determine the biosynthetic incorporation of [3H]leucine into receptors liganded or not with IgE prior to the incubation with the radiolabeled leucine. Prior to solubilization the cells were washed in order to remove unincorporated [‘Hlleucine and unbound IgE.
incubated for 48 hr with [‘Hlleucine in a leucine-free medium. The receptors on some of the cells were analyzed at 24 hr following the protocol shown in Fig. 2. The remaining cells were similarly analyzed at 48 hr. The samples in “A”-precipitates at 24 and 48 hr showed no incorporation of tritium into the a-, b- or y-components. The specimens in “B”-precipitates showed substantial incorporation into the x-, /I- and y-components at 24 and 48 hr. Although the ratio of counts [(/I + y):cw] was lower than that which is regularly observed-l 4-1.6 vs the usual 2.5-3.0 (Perez-Montfbrt et al., 1983; Fig. 4Bbimportantly this ratio did not change between 24 and 48 hr. The “C-precipitates showed very few counts and, when analyzed on gels, the ratio (/l + y ): a was the same as that seen in the “B”-precipitates. Again, there was no change in the ratio between 24 and 48 hr. It seems likely that these counts represent small amounts of mouse IgE-anti-mouse IgE complexes that had not been recovered during the formation of the “B”-precipitates. These experiments demonstrate that, at least over a period of 48 hr, the a/?y, complex behaves as a single biosynthetic unit: those a-chains already incorporated in the plasma membrane do not acquire labeled /?- and y-chains and (as shown previously) the rates of incorporation of [3H]leucine into the three types of chains of newly synthesized receptors are similar.
Chase experiments. In these experiments a protocol was used similar to that employed in the surfacelabeling studies. Cells were first allowed to incorporate [3H]leucine in a leucine-deficient medium, washed, and then transferred to complete medium containing non-radioactive leucine. Prior to the latter incubation some cells were reacted with “‘1-1gE and then washed or maintained in ‘251-IgE. Others were not reacted with IgE. The latter cells were reacted with ‘*‘I-IgE only after sampling just prior to analysis. The results from two successive experiments are shown in Fig. 3. Under all conditions of incubation (see legend of Fig. 3) there was an approx. 30% increase in cell number over 24 hr (Fig. 3A), the actual increases in the two successive experiments being 7 and 22% (0), 29 and 34% (A), and 44 and 29% (0). The cells loaded with radioactive IgE only at the start of the incubation with complete medium (triangles) showed a 15% decrease in bound ‘251-IgE/ml culture after 24 hr (Fig. 3C), presumably as a result of the dissociation of the IgE, degradation of “‘1-IgEreceptors, or both. Since the number of cells/ml had increased also (- 32% after 24 hr, Fig. 3A), these two factors account for the - 35% decrease in the bound iZ51-IgE/cell (Fig. 3B). In the previous experiments (Fig. 1) only the cultures grown in the continuous presence of IgE showed a substantial (-35%) increase in receptors/ml. In the biosynthetic labeling experiments the cells that had been grown in the absence of IgE also showed an increase in total IgE binding (cf. squares in Figs 1B and 3C). This increase is due to both an increase in cell number (av. 36%) and increase in receptors/cell (av. 31%) so that the total increase in receptors/ml averaged almost 80%. The corresponding value for the cells grown in the continuous presence of IgE (circles) was 98%. It is possible that these increases are due to the transfer of the cells from the deficient to the complete medium. Figure 4A presents the data on the ratio of labeled a-subunits to ‘251-IgE in these same cultures. The cells loaded with ‘251-IgE only at the beginning of the experiment (triangles) served as a control since one would expect the ratio ‘H-a : 12’I-IgE to remain constant. Although at 12 hr this ratio appears to have increased, the error is such that this is not statistically significant. It is hard to find an explanation other than experimental error for such an increase. The cells grown in the continuous presence of “‘1-1gE (circles) show a moderate increase in the a : IgE ratio at 12 hr but none after 24 hr. This is a notable difference from the results with surface labeling where a substantial decrease in the a :IgE ratio was seen (Fig. 1A and Table 1). In those latter experiments only the receptors already inserted in the membrane could have been labeled. In the biosynthetic-chase studies (Figs 3 and 4) any receptors that had been synthesized during the incubation with [3H]leucine but which had not yet been inserted into the plasma membrane, would also have been labeled-indeed possibly to a higher sp. act. than many of those
1049
Turnover of IgE receptor
I
I
I
12
24
radioactive leucine (Fig. 4A squares) fell considerably. If this fall reflected only the turnover of receptors already inserted in the plasma membrane at the start of the incubation, the half-time would be approx. 16 hr. This is somewhat longer than the average that can be calculated from the surfacelabeling studies (- 13 hr, Table 1). The experimental errors are such that this difference is not statistically of biosignificant. However, the recruitment synthetically labeled but not-yet-inserted receptorswhich was postulated to account for the failure to see the expected fall in the tl :IgE ratio of the cells incubated in the presence of IgE+ould similarly dampen the fall in the ratio on the cells incubated in the absence of IgE. On the other hand, the increase in total receptors is likely to be only in part due to
TIMEthrl
Fig. 3. Analysis of the number of cells and the IgE bound to them in cell cultures incubated in non-radioactive complete medium after having been grown in [3HJieuc~ne in teucine-depleted medium. Celis were first grown in a spinner culture with [‘Hlleucine for 12 hr. They were then washed. Two aliquots of the cells were saturated with ‘251-IgE (lO~g/ml IgE, 3 x IO’ cells/ml, 45 min) (0 and a). One of these (A) was washed free of unbound IgE; the other (0) was maintained in ‘23K-IgE(1 ,ag/ml) during the subsequent incubation. A third aliquot (IJ) was not reacted with ‘251-IgE. Each aliquot was then incubated in complete medium at 3 x lo6 cells/ml and sampled at 12 and 24 hr. “‘1-1gE (lO~g/ml) was added to those samples containing cells not previously treated with IgE (a). Each sample was analyzed for cell number and cell-bound IgE. (A) Change in cells/ml with time. The data have been normalized to the start of the incubation. (B) Change in IgE bound/cell. Note: for the cells grown in the presence of IgE (0) or saturated with IgE after sampling (D) the IgE bound reflects the total receptors/cell. For those cells (a) reacted with IgE only prior to the incubation in complete medium (chase), receptors inserted during the incubation would not have been detected. (C) Changes in ‘*‘I-IgE bound/ml of culture. The values were calculated from the results in A and B. The data are from 2 separate experiments with the mean values given by the symbols and the range indicated by the error bars. The circles have been plotted at the correct times; where appropriate for clarity of presentation, the other symbols have been slightly displaced.
already inserted. Incorporation of such receptors into the plasma membrane during the subsequent incubation without radioactive leucine could account for the failure to see a fall in the 3H-u : “‘1-1gE ratio. The 3H-a : ‘*%IgE ratio on the cells loaded with IgE after incubation for 0, 12 or 24 hr with non-
1
5 I
4 3
&
1
0
2
______-_ ~
I
__--_--4
41 I
1
12
24
1
TIME (hrl
Fig. 4. Change in radioactive Ieucine content of the receptors and their subunits during incubation of the cells in nonradioactive leucine in complete medium. After the incubation with [3H]leucine some cells were saturated with ‘251-IgEand maintained with 1 pg ‘%IgE/ml in the subsequent incubation (0). Others were washed after saturation with ‘*%IgE and maintained without IgE during the subsequent incubation (A). A third aliquot was not reacted with ‘ZSI-lgE until completion of the incubation in nonradioactive leucine (5). After removal of a small portion for dete~ination of the number of cells and the IgE bound to them (Fig. 3), the celis in each sample were solubilized and the receptors bound to IgE were purified and analyzed as described in Materials and Methods. In these experiments, all the quantitative data were obtained by direct counting of ge1slices. Autoradiographs of all gels were taken prior to the quantitative analyses and were comparable in quality to fhat illustrated in Fig. 4 of Perez-Montfort et ai. (1983). (A) Ratio of rH]leucine in a-chains relative to the ‘*‘I-IgE to which they are bound. (B) Ratio of radioactivity in the fiand y-chains relative to that in the a-chains. The error bars show the range of results from two successive experiments.
1050
RODOLFO QUARTO et al.
recruitment of previously synthesized and therefore labeled receptors. The increase in receptors could also result from insertion of newly synthesized, unlabeled receptors. These would contribute to a fall in the c( : IgE ratio. Thus, in these cells the CI: IgE ratio is likely to reflect a combination of events: (a) decay of labeled membrane receptors, (b) insertion of labeled receptors into the plasma membrane, and (c) insertion of unlabeled receptors into the plasma membrane. Nevertheless, the difference in the CI:IgE ratio for the cells incubated in the absence and presence of IgE is likely to closely reflect turnover of the receptor. As can be seen from Fig. 4A, this was substantial. The latter conclusion makes the results shown in Fig. 4B meaningful. That is, despite considerable turnover of receptors the ratio (B + y2): c( remains constant. The question of what would happen to this ratio was of course the principal matter we were probing in these experiments and the answer is clear. The c(-, p- and y-chains are degraded coordinately. DISCUSSION
In these studies we examined whether the c(-, /Iand y-components that co-purify on the basis of the affinity of c( for IgE behave as a unit during biosynthesis and catabolism. We previously reported that in cells incorporating [3H]leucine, the ratio of counts in c(, B and y remained constant (PerezMontfort et al., 1983). Although those results were consistent with coordinate biosynthesis, they could not provide incontrovertible proof of this. First, because the rates of incorporation into the three polypeptides were not substantially different than the rate of incorporation into the total cellular proteins [see Fig. 5 in Perez-Montfort et al. (1983)] and, second, because such data alone give no information about the fraction of the receptors being analyzed. The data shown in Figs IA and 4A demonstrate that, during the time studied in the present (and the previous) experiments, a considerable proportion of the receptors are turning over. For example, in the experiment with unoccupied receptors (squares in Fig. lA), the total number of receptors did not whereas the specific activity of the change, cc-chain-measured as the ‘25I-c(: ‘3’I-IgE ratio-fell markedly; for the aggregate data (Table 1), 50% in - 13 hr. Therefore, the half-time for turnover of the cc-chain must be about this long. This estimate is somewhat greater than the 8-hr value independently obtained in similar experiments by Furuichi et al. (1985). [These workers have similarly observed an increase in the total IgE-binding capacity of cells incubated in the continuous presence of IgE (cf. circles in Fig. lB)]. Our studies on the incorporation of [3H]leucine over a time sufficiently long that a sizeable fraction of new receptors were synthesized, showed incorporation of label into all three polypeptides in receptors not previously saturated with IgE. On the other hand,
receptors (on the same cells) whose IgE-binding n-chain was already inserted in the plasma membrane, failed to incorporate leucine into either the or-, /I- and y-chains. In complementary “chase” experiments, receptors whose u-, fi- and y-chains had been previously labeled with [3H]leucine showed a parallel rate of decrease in the sp. act. of the three components over a time during which a substantial fraction of the receptors were being degraded and new, unlabeled receptors were being synthesized. As detailed in Results, the combined data also provided evidence for a pool of receptors that had been synthesized during the incubation with [3H]leucine, but had been inserted into the plasma membrane only during the subsequent incubation in non-radioactive leucine. This is in agreement with the studies of Furuichi et al. (1985) who observed that there was an increase in membrane receptors that could not be blocked under conditions where protein synthesis was inhibited. In summary, the current experiments indicate that the B- and y-polypeptides, that in appropriate solvents co-purify at a fixed ratio with the IgE-binding or-chain, also behave as a unit during biosynthesis and degradation. Furthermore, once inserted into the plasma membrane, there does not seem to be any independent exchange of one or more of these components. These results provide further strong support for the tetrameric model of the receptor for IgE.
REFERENCES
Alcaraz G., Kinet J.-P., Kumar N., Wank S. A. and Metzger H. (1984) Phase separation of the receptor for immunoglobulin E and its subunits in Triton X-114. J. biol.
Chem. 259, 14922-14927.
Barsumian E. L., Isersky C., Petrino M. G. and Siraganian R. P. (1981) IgE-induced histamine release from rat basophilic leukemia cell lines: isolation of releasing and nonreleasing clones. Eur. J. Immun. 11, 317-323. Bazin H., Querijean P., Beckers A. and Heremans J. F. (1974) Transplantable immunoglobulin-secreting tumours in rats IV. Sixty-three IgE-secreting immunocytoma tumours. Immunology 26;713-723. Conrad D. H. and Froese A. (1976) Characterization of the target cell receptor for IgE.‘II. Polyacrylamide gel analysis of the surface IgE receptor from normal rat mast cells and from rat basophilic leukemia cells. J. Immun. 116, 319-326. Fewtrell C., Goetze A. and Metzger H. (1982) Phosphorylation of the receptor for immunoglobulin E. Biochemistry 21, 2004-2010. Froese A. (1980) Structure and function of the receptor for IgE. Crit. Rev. Immun. 1, 79-132. Furuichi K., Rivera J. and Isersky C. (1985) The receptor for immunoglobulin E on rat basophilic leukemia cells: effect of ligand binding on receptor expression. Proc. natn. Acad. Sci., U.S.A. 82, 1522-1525. Helm R. M. and Froese A. (1981) The incorporation of tritiated precursors into receptors for IgE of rat basophilic leukemia cells. Immunology 42, 629-636. Hjelmeland L. M. (1980) A nondenaturing zwitterionic
Turnover
of IgE receptor
detergent for membrane biochemistry: design and synthesis. Proc. natn. Acad. Sci. U.S.A. 71, 6368-6370. Holowka D., Hartmann H., Kanellopoulos J. and Metzger H. (1980) Association of the receptor for immunoglobulin E with an endogenous polypeptide on rat basophilic leukemia cells. J. Receptor Res. 1, 41-68. Holowka D. and Metzger H. (1982) Further characterization of the /?-component of the receptor for IgE. Molec. Immun. 19, 219-227. Kanellopoulos J. M., Liu T. Y., Poy G. and Metzger H. (1980) Composition and subunit structure of the cell receptor for immunoglobulin E. J. biof. Chem. 255,
9060-9066. Kanellopoulos, J., Rossi, G. and Metzger H. (1979) Preparative isolation of the cell receptor for immunoglobulin E. J. biol. Chem. 254, 7691-7697. Kinet J.-P.. Alcaraz G., Leonard A.. Wank S. and Metzger H. (in press) Dissociation of the receptor for imn&noglobulin E in mild detergents. Biochemistry. Kulczycki A., Jr and Metzger H. (1974) The interaction of IgE with rat basophilic leukemia cells. II. Quantitative aspects of binding reaction. J. exp. Med. 140, 1676-1695. Kulczycki A., Jr and Parker C. W. (1979) The cell surface receptor for immunoglobulin E. I. The use of repetitive affinity chromatography for the purification of a mammalian receptor. J. biol. Chem. 254, 3187-3193. Kulczycki A., Jr, McNeamey T. A. and Parker C. W. (1976) The rat basophilic leukemia cell receptor for IgE. I. Characterization as a glycoprotein. J. Immun. 117, 661-665. Kumar N. and Metzger H. (1982) Gel filtration in 6 M guanidine hydrochloride of the alpha subunit (and its
1051
fragments) of the receptor for immunoglobuhn E. Molec. Immun. 19, 1561-1567. Liu F.-T., Bohn J. W., Ferry E. L., Yamamoto H., Molinaro C. A., Sherman L.A., Klinman N. R. and Katz D. H. (1980) Monoclonal dinitrophenyl-specific murine IgE antibody: preparation, isolation and characterization. J. Immun. 124, 2128-2736. Marchalonis J. J. (1969) An enzymic method for the trace iodination of immunoglobulins and other proteins. Biothem. J. 113, 299-305. Markwell M. A. K. (1982) A new solid-state reagent to iodinate proteins. I. Conditions for the efficient labeling of antiserum. Analyt. Biochem. 125, 427432. Markwell M. A. K. and Fox C. F. (1978) Surface specific iodination of membrane proteins of viruses and eucaryotic cells using 1,3,4,6-tetrachloro-3cc,6cc-diphenylal~cohnil. Biochemistrv 17. 4807-4817. McConahey P. J. and Dixon F. G. (1966) A method of trace iodination of proteins for immunologic studies. ht. Archs
Allergy appl. Immun. 29, 185-189. Metzger H., Kinet J.-P., Perez-Montfort R., Rivnay B. and W&k S. A. (1984) A tetrameric model for the structure of the mast cell recentor with high affinitv for IgE. In Progress in Immunology V (Edited‘by Yamamura Y. and Tada T.), pp. 493-501. Academic Press, Orlando, FL. Perez-Montfort R., Kinet J.-P. and Metzger H. (1983) A previously unrecognized subunit (y) of the receptor for immunoglobulin E. Biochemistry 22, 5722-5728. Rivnay B., Wank S. A. and Metzger H. (1982) Phospholipids stabilize the interaction between the alpha and beta subunits of the solubilized receptor for immunoglobulin E. Biochemistry 21, 6922-6927.
0161-%X90/85 $3.00 + 0.00 PergamonPress Ltd
COORDINATE SYNTHESIS AND DEGRADATION THE a-, ,9- AND y-SUBUNITS OF THE RECEPTOR FOR IMMUNOGLOBULIN E
OF
RODOLFO QUARTO, JEAN-PIERRE KINET and HENRY METZGER Section on Chemical Immunology, Arthritis and Rheumatism Branch, National Institute of Arthritis, Diabetes, and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD 20205, U.S.A.
(First received 6 December 1984; accepted in revised form 11 February 1985) Abstract-The surface receptor for immunoglobulin E (IgE) on rat basophilic leukemia cells and their normal counterparts has been postulated to consist of four polypeptide chains: a 45-kDa a-chain which binds IgE, a 33-kDa p-component and two disulfide-linked, 9-lo-kDa y-polypeptides. The instability of this complex in mild detergents makes it possible that, in vivo also, the structure may not be stable and that there is an independent assembly or exchange of the chains. We studied this question using surface-labeling and biosynthetic labeling techniques and found that the chains turn over coordinately and do not independently exchange. The results provide further support for the proposal that the aby, complex is the unit receptor for IgE.
INTRODUCTION
Mast cells and analogues of such cells contain a surface glycoprotein that tightly binds monomeric immunoglobulin E (IgE). This glycoprotein has a mol. wt of approx. 45,000 (Conrad and Froese, 1976; Kulczycki et al., 1976; Kumar and Metzger, 1982+30x of which is accounted for by carbohydrate (Kanellopoulos et al., 1980). When cells bearing this protein are solubilized with mild detergents and the IgE-binding component is isolated by affinity chromatography in solvents containing micellar detergent and analyzed by electrophoresis on polyacrylamide gels, only the 45kDa protein is observed. This holds whether the cells have been labeled biosynthetically or by external iodination. Small amounts of label can be irregularly observed in other components (Kulczycki and Parker, 1979; Froese, 1980; Helm and Froese, 1981), but it is not possible to know whether these are contaminants or components that relate to the IgE-binding protein. The use of cross-linking reagents first gave direct evidence that at least one other component was associated with the 45kDa protein in significant amounts (Holowka et al., 1980; Holowka and Metzger, 1982). Subsequently, it was found that, even in the absence of cross-linking reagents, this second, 33-kDa, component co-purified with the 45kDa glycoprotein providing the solutions contained a detergent:lipid ratio similar to that in the original extract of the cells (Rivnay et al., 1982). Ultimately, a third component was observed under these conditions (Perez-Montfort et al., 1983), leading to the proposal that the receptor for IgE consisted of four polypeptides: the 45kDa IgE-binding glycoprotein (GL), a 33-kDa b-chain and two disulfide-linked, 9-lo-kDa, y-chains (Perez-Montfort et al., 1983; Metzger et al., 1984).
Although the stoichiometric data strongly suggest that all of these chains are integral parts of the receptor, the unusual instability of the complex makes it possible that in situ the component peptides can turn over independently. Indeed, in studies in which cells were incubated with 32Pi, we observed labeling of the /?-chains under conditions in which we were able to selectively examine receptors already incorporated into the plasma membrane (Fewtrell et al., 1982). The alternative explanations of those findings were that either chains already incorporated into the membrane could be modified, or that the /Iand y-chains could turn over independently of the a-chains. The present study was undertaken to explore this matter further. MATERIALS AND METHODS
IgE and cells
Rat IgE [IR 162 (Bazin et al., 1974)] and mouse anti-dinitrophenyl IgE [Hi-DNP-e-26.82 (Liu et al., purified as described previously 1980)] were (Kulczycki and Metzger, 1974; Holowka and Metzger, 1982). Iodinations were performed with chloramine-T (McConahey and Dixon, 1966) or with Iodo-beads (Markwell, 1982) obtained from Pierce (Rockford, IL). Arsonylated IgE was prepared as described by Kanellopoulos et al. (1979). Rat basophilic leukemia cells were maintained as described by Barsumian et al. (1981). Binding studies
Cells were saturated with iodinated IgE and analyzed for cell-bound IgE as described previously (Kulczycki and Metzger, 1974). The cells had an average of 3 x lo5 receptors in all the studies described here. 1045
1046
ROD~LFO QUARTO et al.
Labeling of cells Surface iodination. The labeling by lactoperoxidase-catalyzed iodination of cells from stationary cultures has been described previously (Marchalonis, 1969; Kanellopoulos et al., 1979). Alternatively we used Iodo-gen (Markwell and Fox, 1978). Aliquots of 5 x 10’ cells in phosphate-buffered saline were added to a glass vial containing 3OOpg of Iodo-gen and then 500 PCi of carrier-free Na”‘I (Amersham Corp., Arlington, IL) was added. After a 15-min incubation on ice with occasional shaking, the cells were transferred to a tube containing 5 ml complete medium and then processed as described previously (Kanellopoulos et al., 1979). Intrinsic labeling. The method described by PerezMontfort et al. (1983) was used for incorporating L-[3H]leucine (Amersham Corp., code TRK 5 10). For the chase experiments, the washed cells were incubated with IgE where desired and then maintained in complete medium in spinner flasks. PuriJication of receptors
In all cases the receptors were isolated on the basis of the IgE bound to them. Cells were solubilized at 5 x 10’ cells/ml with the detergent 3-[(3-cholamidopropyl)dimethylammonio]-l-propane sulfonate (10 mM) (Pierce) (Hjelmeland, 1980) and protease inhibitors as previously described (Rivnay et al., 1982). The same detergent was used throughout this study. Surface-labeled receptors. These were purified simply by immunoprecipitation using purified rabbit antibodies to IgE, as described previously (PerezMontfort et al., 1983). Biosyntheticaiiy labeled receptors. These were purified essentially as described previously (PerezMontfort et al., 1983). The IgE-receptor complexes were adsorbed either to trinitrophenyl-lysine on Biogel A5m beads (Holowka and Metzger, 1982) or to anti-benzene arsonate-Sepharose 4B (Kanellopoulos et aI., 1979) and were extensively washed with 2 mM detergent in borate-buffered saline, pH 8.0, at 4°C (Alcaraz et al., 1984; Kinet et al., in press). They were then eluted with the appropriate hapten {lOmM dinitrophenyl-e-NH, caproate or 1 mM [(p-arsonicnitrophenyl)-azoltyrosine} in the same buffer. In all subsequent steps the same buffer was used. The complexes were then immunoprecipitated and analyzed by electrophoresis in polyacrylamide gels in sodium dodecyl sulfate as described by PerezMontfort et al. (1983). The ratio z:/I :y was determined by densitometry of autoradiographs of the gels or by slicing the dried slabs and counting in a scintillation counter. RESULTS
Our basic strategy was to develop an estimate of the rates of synthesis and degradation of the IgEbinding a-chain of the receptor and then to compare
these rates with the rates of turnover of the fl- and y-chains. Turnover of the u-chains
The plasma membrane proteins of the cells were surface-labeled with 12’I.Under these conditions only the a-chains of the receptor are labeled even though it can be shown that the /I- and y-chains are present (Perez-Montfort et al., 1983). The cells were then treated in three different ways. One aliquot was reacted with “‘I-labeled mouse IgE and thoroughly washed to remove the unbound IgE. The cells were then incubated at 37°C and sampled periodically. The concns of the reactants were such that saturation of the receptors would be expected (Kulczycki and Metzger, 1974). A second aliquot was similarly reacted with 13’1-IgE but not washed. These cells were then incubated at 37°C in the continuous presence of the labeled IgE and sampled at intervals. The IgE in the medium was sufficient to maintain saturation of the receptors given the K, of IgE for the receptor. A third aliquot was not reacted with IgE before or during the incubation. Samples of these cells were reacted with 13’I-IgE just prior to analysis. The conditions used (legend of Fig. 1) were adequate to ensure saturation of the receptors. The cells in all samples were washed, solubilized with detergent, centrifuged and the supernatant solutions subjected to a “clearing” immunoprecipitation with human IgE and anti human IgE. The supernatants were reacted with anti-mouse IgE and the precipitates were then analyzed by electrophoresis in polyacrylamide gels in sodium dodecyl sulfate. The gels were sliced and the counts in ‘3’I-IgE and ‘251-cr-chains determined. The data from one such experiment are shown in Fig. lA, and illustrate the change with time of the ‘25I-c(:“‘I-IgE ratio in the samples from the three incubation mixtures. The cells that had been loaded with labeled IgE prior to the incubation and then incubated in the absence of IgE show an invariant ratio of ‘25I-~: “‘I-IgE (triangles). These specimens serve as the control or baseline. Because here only those receptors that had bound IgE prior to the incubation were examined, there is no apriori reason why the ratio should have changed. On the other hand, the cells incubated in the continuous presence of labeled IgE show a moderate decline in the ‘25I-c(: 13’I-IgE ratio over time (circles). This would be expected if during the course of the incubation, new and, therefore, unlabeled receptors had appeared on the surface of the cells where they could bind the 13’1-IgE present in the medium. The data in Fig. 1B show that these cells in fact showed an increase in total receptors and this increase in total receptors accounts quantitatively rather well for the decrease in the ‘25I-a:13’1-IgE ratio. Thus, at 24 hr, the 35% decrease in the “‘1-a:IgE ratio (Fig. 1A) is accounted for by the 35% increase in receptors (Fig. 1B).
Turnover 1.0
of IgE receptor
1047
Table 1. Ratio of counts in ‘2SI-labeled or-chains to counts in “‘I-labeled IgE on cells incubated in the presence or absence of “‘I-1gE in the medium
d tw
Time of incubation (hr)
-
0 6 12 24
-
1
A
4-l
-0 I
TIME (hrl Fig. 1. Analysis of surface-labelled cells. Cells were surface iodinated with lzsI and then treated in 1 of 3 ways. One aliquot was reacted with “‘I-IgE (3 x 10’ cells/ml, 10 pg/ml IgE) for 30-40 min and maintained in 13’I-IgE (1 pg/ml) throughout the 24-hr incubation (0). A second aliquot was similarly reacted with 13’I-IgE and then washed to remove all unbound IgE prior to the incubation (A). The third aliquot was reacted for 60 min with 10 pg/ml 13’1-IgE only after samples from the incubation were removed just prior to analysis (0). All incubations contained -4 x lo6 cells/ml. The cells in each sample were washed, solubilized with detergent and the receptors immunoprecipitated before being analyzed on gels. A portion of each incubation mixture was also analyzed for cell number and bound “‘I-IgE. (A) Ratio of counts in ‘251-labeleda-chains to the counts in the “‘I-IgE. For each incubation the data were normalized by dividing the ratio determined at each time point by the ratio determined at time zero. The time in hours is the time elapsed after the surface labeling was terminated. mixture calcu(B) Ratio of bound i3’I-IgE/ml incubation lated from the counts of cells/ml and “‘I-IgE bound/cell. The ratios were normalized for each incubation mixture as noted in A. All counts were performed in triplicate. The actual cell counts/ml at zero time and 24 hr were 3.8 x lo6 and 3.4 x lo6 (O), 3.9 x lo6 and 4.5 x lo6 (A), and 4.5 x lo6 and 4.8 x lo6 (Cl).
Normalized ratio” + IgE
I 0.948 (0.145) 0.845 (0.058) 0.465 (0.158)
- IgE 1 0.791 (0.139) 0.507 (0.101) 0.257 (0.1 30)h
“Except as noted the values shown are the averages from four separate experiments. The values in the parentheses are the SDS. “Average from three experiments only. The counts in the a-chains in the fourth experiment were too low to measure accurately.
such a comparison was made are collected in Table 1. Clearly, the difference between the two types of incubations is reproducible. In several of these experiments the total cells/ml and the IgE bound/cell was assessed throughout the incubation and from these data the cell-bound IgE per sample could be calculated such as illustrated in Fig. 1B. Although there was some variability in the absolute values, the receptors/ml in the incubations without IgE never substantially increased and usually declined slightly over the period studied. In four successive experiments the ratio of receptors/ml at 24 hr to that at zero time was 0.89, 1.4, 1.07 and 0.83 respectively. Therefore, the decrease in the ‘*‘I-cr: ‘-“I-IgE ratio in these samples cannot simply be due to the acquisition of newly-synthesized and therefore unlabeled a-chains. There must also have been a corresponding degradation of the chains. Contrariwise, the incubations containing free IgE showed about a 25% increase in the IgE-binding capacity (Fig. 1B). The cells preloaded with IgE but not grown in the presence of IgE were only analyzed for residual bound IgE and not for their total IgE-binding capacity. As shown in Fig. 1B (triangles) there was a moderate decline in the cell-bound IgE, due either to degradation of the bound IgE or dissociation of the IgE, or both. Together, these results are most simply explained by a model in which new cz-chains appear on the surface of the cells in the presence or absence of IgE in the medium, but that those cc-chains to which IgE is bound are less labile than those that are empty. In any case, these observations indicate that, over the period studied, a substantial fraction of those cc-chains that are not bound with IgE are degraded and new cc-chains are taking their place. Turnover of the fl- and y-chains
A decline in the ‘*%Y: 13’1-IgEratio is also expected for the cells grown in the absence of IgE, but what is striking is the more rapid decrease in the ratio for these specimens (squares) compared to those in which the labeled IgE was present continuously (circles). The combined results from four experiments in which
in order to examine the turnover of the p- and labeling with y-chains we used biosynthetic [3H]leucine. Incorporation experiments. Approximately 30% of the receptors on the cells were occupied with arsonylated ‘3’I-labeled rat IgE. The cells were thoroughly washed to remove unbound IgE and then
RODOLFO QUARTO et al.
1048
BIOSYNTHETIC LABELING OF LIGANDED AND UNLIGANDED RECEPTORS ON THE SAME CELLS
pamally
saturate
wtth
1251 (Ars)
receptors rat IgE
I wash
cells
I suspend
cells I” lewoe-free
medum
+ sH-leucine
I sample
at 24, 48 hr +12WArs)
no addition A solubillze
mouse
IgE
cells
I
I purify
IgE receptors
on anti-Ars-Sepharose
+ anti-mouse
+ anti rat IgE
IgE
supernatant
analyze (A)
ppt
1
1 + anti-rat
analyze
ppts
IBI
IgE
(Cl
Fig. 2. Flow diagram of the experimental procedure used to determine the biosynthetic incorporation of [3H]leucine into receptors liganded or not with IgE prior to the incubation with the radiolabeled leucine. Prior to solubilization the cells were washed in order to remove unincorporated [‘Hlleucine and unbound IgE.
incubated for 48 hr with [‘Hlleucine in a leucine-free medium. The receptors on some of the cells were analyzed at 24 hr following the protocol shown in Fig. 2. The remaining cells were similarly analyzed at 48 hr. The samples in “A”-precipitates at 24 and 48 hr showed no incorporation of tritium into the a-, b- or y-components. The specimens in “B”-precipitates showed substantial incorporation into the x-, /I- and y-components at 24 and 48 hr. Although the ratio of counts [(/I + y):cw] was lower than that which is regularly observed-l 4-1.6 vs the usual 2.5-3.0 (Perez-Montfbrt et al., 1983; Fig. 4Bbimportantly this ratio did not change between 24 and 48 hr. The “C-precipitates showed very few counts and, when analyzed on gels, the ratio (/l + y ): a was the same as that seen in the “B”-precipitates. Again, there was no change in the ratio between 24 and 48 hr. It seems likely that these counts represent small amounts of mouse IgE-anti-mouse IgE complexes that had not been recovered during the formation of the “B”-precipitates. These experiments demonstrate that, at least over a period of 48 hr, the a/?y, complex behaves as a single biosynthetic unit: those a-chains already incorporated in the plasma membrane do not acquire labeled /?- and y-chains and (as shown previously) the rates of incorporation of [3H]leucine into the three types of chains of newly synthesized receptors are similar.
Chase experiments. In these experiments a protocol was used similar to that employed in the surfacelabeling studies. Cells were first allowed to incorporate [3H]leucine in a leucine-deficient medium, washed, and then transferred to complete medium containing non-radioactive leucine. Prior to the latter incubation some cells were reacted with “‘1-1gE and then washed or maintained in ‘251-IgE. Others were not reacted with IgE. The latter cells were reacted with ‘*‘I-IgE only after sampling just prior to analysis. The results from two successive experiments are shown in Fig. 3. Under all conditions of incubation (see legend of Fig. 3) there was an approx. 30% increase in cell number over 24 hr (Fig. 3A), the actual increases in the two successive experiments being 7 and 22% (0), 29 and 34% (A), and 44 and 29% (0). The cells loaded with radioactive IgE only at the start of the incubation with complete medium (triangles) showed a 15% decrease in bound ‘251-IgE/ml culture after 24 hr (Fig. 3C), presumably as a result of the dissociation of the IgE, degradation of “‘1-IgEreceptors, or both. Since the number of cells/ml had increased also (- 32% after 24 hr, Fig. 3A), these two factors account for the - 35% decrease in the bound iZ51-IgE/cell (Fig. 3B). In the previous experiments (Fig. 1) only the cultures grown in the continuous presence of IgE showed a substantial (-35%) increase in receptors/ml. In the biosynthetic labeling experiments the cells that had been grown in the absence of IgE also showed an increase in total IgE binding (cf. squares in Figs 1B and 3C). This increase is due to both an increase in cell number (av. 36%) and increase in receptors/cell (av. 31%) so that the total increase in receptors/ml averaged almost 80%. The corresponding value for the cells grown in the continuous presence of IgE (circles) was 98%. It is possible that these increases are due to the transfer of the cells from the deficient to the complete medium. Figure 4A presents the data on the ratio of labeled a-subunits to ‘251-IgE in these same cultures. The cells loaded with ‘251-IgE only at the beginning of the experiment (triangles) served as a control since one would expect the ratio ‘H-a : 12’I-IgE to remain constant. Although at 12 hr this ratio appears to have increased, the error is such that this is not statistically significant. It is hard to find an explanation other than experimental error for such an increase. The cells grown in the continuous presence of “‘1-1gE (circles) show a moderate increase in the a : IgE ratio at 12 hr but none after 24 hr. This is a notable difference from the results with surface labeling where a substantial decrease in the a :IgE ratio was seen (Fig. 1A and Table 1). In those latter experiments only the receptors already inserted in the membrane could have been labeled. In the biosynthetic-chase studies (Figs 3 and 4) any receptors that had been synthesized during the incubation with [3H]leucine but which had not yet been inserted into the plasma membrane, would also have been labeled-indeed possibly to a higher sp. act. than many of those
1049
Turnover of IgE receptor
I
I
I
12
24
radioactive leucine (Fig. 4A squares) fell considerably. If this fall reflected only the turnover of receptors already inserted in the plasma membrane at the start of the incubation, the half-time would be approx. 16 hr. This is somewhat longer than the average that can be calculated from the surfacelabeling studies (- 13 hr, Table 1). The experimental errors are such that this difference is not statistically of biosignificant. However, the recruitment synthetically labeled but not-yet-inserted receptorswhich was postulated to account for the failure to see the expected fall in the tl :IgE ratio of the cells incubated in the presence of IgE+ould similarly dampen the fall in the ratio on the cells incubated in the absence of IgE. On the other hand, the increase in total receptors is likely to be only in part due to
TIMEthrl
Fig. 3. Analysis of the number of cells and the IgE bound to them in cell cultures incubated in non-radioactive complete medium after having been grown in [3HJieuc~ne in teucine-depleted medium. Celis were first grown in a spinner culture with [‘Hlleucine for 12 hr. They were then washed. Two aliquots of the cells were saturated with ‘251-IgE (lO~g/ml IgE, 3 x IO’ cells/ml, 45 min) (0 and a). One of these (A) was washed free of unbound IgE; the other (0) was maintained in ‘23K-IgE(1 ,ag/ml) during the subsequent incubation. A third aliquot (IJ) was not reacted with ‘251-IgE. Each aliquot was then incubated in complete medium at 3 x lo6 cells/ml and sampled at 12 and 24 hr. “‘1-1gE (lO~g/ml) was added to those samples containing cells not previously treated with IgE (a). Each sample was analyzed for cell number and cell-bound IgE. (A) Change in cells/ml with time. The data have been normalized to the start of the incubation. (B) Change in IgE bound/cell. Note: for the cells grown in the presence of IgE (0) or saturated with IgE after sampling (D) the IgE bound reflects the total receptors/cell. For those cells (a) reacted with IgE only prior to the incubation in complete medium (chase), receptors inserted during the incubation would not have been detected. (C) Changes in ‘*‘I-IgE bound/ml of culture. The values were calculated from the results in A and B. The data are from 2 separate experiments with the mean values given by the symbols and the range indicated by the error bars. The circles have been plotted at the correct times; where appropriate for clarity of presentation, the other symbols have been slightly displaced.
already inserted. Incorporation of such receptors into the plasma membrane during the subsequent incubation without radioactive leucine could account for the failure to see a fall in the 3H-u : “‘1-1gE ratio. The 3H-a : ‘*%IgE ratio on the cells loaded with IgE after incubation for 0, 12 or 24 hr with non-
1
5 I
4 3
&
1
0
2
______-_ ~
I
__--_--4
41 I
1
12
24
1
TIME (hrl
Fig. 4. Change in radioactive Ieucine content of the receptors and their subunits during incubation of the cells in nonradioactive leucine in complete medium. After the incubation with [3H]leucine some cells were saturated with ‘251-IgEand maintained with 1 pg ‘%IgE/ml in the subsequent incubation (0). Others were washed after saturation with ‘*%IgE and maintained without IgE during the subsequent incubation (A). A third aliquot was not reacted with ‘ZSI-lgE until completion of the incubation in nonradioactive leucine (5). After removal of a small portion for dete~ination of the number of cells and the IgE bound to them (Fig. 3), the celis in each sample were solubilized and the receptors bound to IgE were purified and analyzed as described in Materials and Methods. In these experiments, all the quantitative data were obtained by direct counting of ge1slices. Autoradiographs of all gels were taken prior to the quantitative analyses and were comparable in quality to fhat illustrated in Fig. 4 of Perez-Montfort et ai. (1983). (A) Ratio of rH]leucine in a-chains relative to the ‘*‘I-IgE to which they are bound. (B) Ratio of radioactivity in the fiand y-chains relative to that in the a-chains. The error bars show the range of results from two successive experiments.
1050
RODOLFO QUARTO et al.
recruitment of previously synthesized and therefore labeled receptors. The increase in receptors could also result from insertion of newly synthesized, unlabeled receptors. These would contribute to a fall in the c( : IgE ratio. Thus, in these cells the CI: IgE ratio is likely to reflect a combination of events: (a) decay of labeled membrane receptors, (b) insertion of labeled receptors into the plasma membrane, and (c) insertion of unlabeled receptors into the plasma membrane. Nevertheless, the difference in the CI:IgE ratio for the cells incubated in the absence and presence of IgE is likely to closely reflect turnover of the receptor. As can be seen from Fig. 4A, this was substantial. The latter conclusion makes the results shown in Fig. 4B meaningful. That is, despite considerable turnover of receptors the ratio (B + y2): c( remains constant. The question of what would happen to this ratio was of course the principal matter we were probing in these experiments and the answer is clear. The c(-, p- and y-chains are degraded coordinately. DISCUSSION
In these studies we examined whether the c(-, /Iand y-components that co-purify on the basis of the affinity of c( for IgE behave as a unit during biosynthesis and catabolism. We previously reported that in cells incorporating [3H]leucine, the ratio of counts in c(, B and y remained constant (PerezMontfort et al., 1983). Although those results were consistent with coordinate biosynthesis, they could not provide incontrovertible proof of this. First, because the rates of incorporation into the three polypeptides were not substantially different than the rate of incorporation into the total cellular proteins [see Fig. 5 in Perez-Montfort et al. (1983)] and, second, because such data alone give no information about the fraction of the receptors being analyzed. The data shown in Figs IA and 4A demonstrate that, during the time studied in the present (and the previous) experiments, a considerable proportion of the receptors are turning over. For example, in the experiment with unoccupied receptors (squares in Fig. lA), the total number of receptors did not whereas the specific activity of the change, cc-chain-measured as the ‘25I-c(: ‘3’I-IgE ratio-fell markedly; for the aggregate data (Table 1), 50% in - 13 hr. Therefore, the half-time for turnover of the cc-chain must be about this long. This estimate is somewhat greater than the 8-hr value independently obtained in similar experiments by Furuichi et al. (1985). [These workers have similarly observed an increase in the total IgE-binding capacity of cells incubated in the continuous presence of IgE (cf. circles in Fig. lB)]. Our studies on the incorporation of [3H]leucine over a time sufficiently long that a sizeable fraction of new receptors were synthesized, showed incorporation of label into all three polypeptides in receptors not previously saturated with IgE. On the other hand,
receptors (on the same cells) whose IgE-binding n-chain was already inserted in the plasma membrane, failed to incorporate leucine into either the or-, /I- and y-chains. In complementary “chase” experiments, receptors whose u-, fi- and y-chains had been previously labeled with [3H]leucine showed a parallel rate of decrease in the sp. act. of the three components over a time during which a substantial fraction of the receptors were being degraded and new, unlabeled receptors were being synthesized. As detailed in Results, the combined data also provided evidence for a pool of receptors that had been synthesized during the incubation with [3H]leucine, but had been inserted into the plasma membrane only during the subsequent incubation in non-radioactive leucine. This is in agreement with the studies of Furuichi et al. (1985) who observed that there was an increase in membrane receptors that could not be blocked under conditions where protein synthesis was inhibited. In summary, the current experiments indicate that the B- and y-polypeptides, that in appropriate solvents co-purify at a fixed ratio with the IgE-binding or-chain, also behave as a unit during biosynthesis and degradation. Furthermore, once inserted into the plasma membrane, there does not seem to be any independent exchange of one or more of these components. These results provide further strong support for the tetrameric model of the receptor for IgE.
REFERENCES
Alcaraz G., Kinet J.-P., Kumar N., Wank S. A. and Metzger H. (1984) Phase separation of the receptor for immunoglobulin E and its subunits in Triton X-114. J. biol.
Chem. 259, 14922-14927.
Barsumian E. L., Isersky C., Petrino M. G. and Siraganian R. P. (1981) IgE-induced histamine release from rat basophilic leukemia cell lines: isolation of releasing and nonreleasing clones. Eur. J. Immun. 11, 317-323. Bazin H., Querijean P., Beckers A. and Heremans J. F. (1974) Transplantable immunoglobulin-secreting tumours in rats IV. Sixty-three IgE-secreting immunocytoma tumours. Immunology 26;713-723. Conrad D. H. and Froese A. (1976) Characterization of the target cell receptor for IgE.‘II. Polyacrylamide gel analysis of the surface IgE receptor from normal rat mast cells and from rat basophilic leukemia cells. J. Immun. 116, 319-326. Fewtrell C., Goetze A. and Metzger H. (1982) Phosphorylation of the receptor for immunoglobulin E. Biochemistry 21, 2004-2010. Froese A. (1980) Structure and function of the receptor for IgE. Crit. Rev. Immun. 1, 79-132. Furuichi K., Rivera J. and Isersky C. (1985) The receptor for immunoglobulin E on rat basophilic leukemia cells: effect of ligand binding on receptor expression. Proc. natn. Acad. Sci., U.S.A. 82, 1522-1525. Helm R. M. and Froese A. (1981) The incorporation of tritiated precursors into receptors for IgE of rat basophilic leukemia cells. Immunology 42, 629-636. Hjelmeland L. M. (1980) A nondenaturing zwitterionic
Turnover
of IgE receptor
detergent for membrane biochemistry: design and synthesis. Proc. natn. Acad. Sci. U.S.A. 71, 6368-6370. Holowka D., Hartmann H., Kanellopoulos J. and Metzger H. (1980) Association of the receptor for immunoglobulin E with an endogenous polypeptide on rat basophilic leukemia cells. J. Receptor Res. 1, 41-68. Holowka D. and Metzger H. (1982) Further characterization of the /?-component of the receptor for IgE. Molec. Immun. 19, 219-227. Kanellopoulos J. M., Liu T. Y., Poy G. and Metzger H. (1980) Composition and subunit structure of the cell receptor for immunoglobulin E. J. biof. Chem. 255,
9060-9066. Kanellopoulos, J., Rossi, G. and Metzger H. (1979) Preparative isolation of the cell receptor for immunoglobulin E. J. biol. Chem. 254, 7691-7697. Kinet J.-P.. Alcaraz G., Leonard A.. Wank S. and Metzger H. (in press) Dissociation of the receptor for imn&noglobulin E in mild detergents. Biochemistry. Kulczycki A., Jr and Metzger H. (1974) The interaction of IgE with rat basophilic leukemia cells. II. Quantitative aspects of binding reaction. J. exp. Med. 140, 1676-1695. Kulczycki A., Jr and Parker C. W. (1979) The cell surface receptor for immunoglobulin E. I. The use of repetitive affinity chromatography for the purification of a mammalian receptor. J. biol. Chem. 254, 3187-3193. Kulczycki A., Jr, McNeamey T. A. and Parker C. W. (1976) The rat basophilic leukemia cell receptor for IgE. I. Characterization as a glycoprotein. J. Immun. 117, 661-665. Kumar N. and Metzger H. (1982) Gel filtration in 6 M guanidine hydrochloride of the alpha subunit (and its
1051
fragments) of the receptor for immunoglobuhn E. Molec. Immun. 19, 1561-1567. Liu F.-T., Bohn J. W., Ferry E. L., Yamamoto H., Molinaro C. A., Sherman L.A., Klinman N. R. and Katz D. H. (1980) Monoclonal dinitrophenyl-specific murine IgE antibody: preparation, isolation and characterization. J. Immun. 124, 2128-2736. Marchalonis J. J. (1969) An enzymic method for the trace iodination of immunoglobulins and other proteins. Biothem. J. 113, 299-305. Markwell M. A. K. (1982) A new solid-state reagent to iodinate proteins. I. Conditions for the efficient labeling of antiserum. Analyt. Biochem. 125, 427432. Markwell M. A. K. and Fox C. F. (1978) Surface specific iodination of membrane proteins of viruses and eucaryotic cells using 1,3,4,6-tetrachloro-3cc,6cc-diphenylal~cohnil. Biochemistrv 17. 4807-4817. McConahey P. J. and Dixon F. G. (1966) A method of trace iodination of proteins for immunologic studies. ht. Archs
Allergy appl. Immun. 29, 185-189. Metzger H., Kinet J.-P., Perez-Montfort R., Rivnay B. and W&k S. A. (1984) A tetrameric model for the structure of the mast cell recentor with high affinitv for IgE. In Progress in Immunology V (Edited‘by Yamamura Y. and Tada T.), pp. 493-501. Academic Press, Orlando, FL. Perez-Montfort R., Kinet J.-P. and Metzger H. (1983) A previously unrecognized subunit (y) of the receptor for immunoglobulin E. Biochemistry 22, 5722-5728. Rivnay B., Wank S. A. and Metzger H. (1982) Phospholipids stabilize the interaction between the alpha and beta subunits of the solubilized receptor for immunoglobulin E. Biochemistry 21, 6922-6927.