Extracellular Release as the Major Degradative Pathway of the Insulin-like Growth Factor 11/ Mannose 6-Phosphate Receptor*

The presence of a soluble, truncated form of the IGF- II/Man-6-P receptor in serum has suggested that cleavage from the cell surface may be an initial step in the degradation of this protein (MacDonald, R. G., Tepper, M. A., Clairmont, K. B., Perregaux, S. B., and Czech, M. P. (1989) J. Biol. Chem. 264,3256-3261). In order to test this hypothesis, we pulse-labeled cultured BRL- 3A rat liver cells with [35S]methionine and [35S]cys- teine and measured the fate of labeled receptor at various times after incubation with unlabeled amino acids. It was found that the appearance of labeled IGF- II/Man-6-P receptor in the medium accounts quantitatively for the loss of labeled receptor from the BRL- 3A cells. In similar experiments with Chinese hamster ovary cells, L6 rat myoblasts, and chick embryo fibro- blasts, labeled receptor from the cell membranes de-creases with a time course corresponding to the ap- pearance of soluble receptor in the medium. The release of labeled receptor into the medium can be blocked by the addition of the protease inhibitors aprotinin, chymostatin, or phenylmethylsulfonyl fluoride, but not antipain, leupeptin, and benzamidine. The results are consistent with the hypothesis that the deg- radation and loss of cellular IGF-II/Man-6-P receptors

life cycle. It recycles between plasma membrane and intracellular compartments, while its ligands, IGF-I1 and Man-6-Plinked proteins, are rapidly delivered to lysosomes (Geuze et al., 1988). A number of published reports describe experiments using pulse-chase protocols to address the synthesis, processing, and degradation of the IGF-II/Man-6-P receptor (Goldberg et al., 1983;Creek and Sly, 1983;Sahagian and Neufeld, 1983;MacDonald and Czech, 1985). These reports have demonstrated that the receptor requires a significant length of time to reach maturity (4-8 h), which requires its conversion by glycosylation from the initially detected 245-kDa form to the mature 260-kDa form. During this maturation process the receptor also gains the ability to bind to its ligands. The receptor has a long half-life (24-48 h), which is not altered by agents which disrupt lysosomes. These reports suggest that the degradation of the receptor is not lysosomal (Creek and Sly, 1983;. The mechanism of receptor degradation has remained an important open question. A possible mechanism for cellular degradation of the IGF-II/Man-6-P receptor was suggested by the initial report of a form of this receptor in serum (Kiess et al., 1987). Subsequent work demonstrated that this serum receptor could bind IGF-I1 and Man-6-P containing ligands simultaneously, that it was cytoplasmically altered or truncated as compared with the cellular receptor (MacDonald et al., 1989), and that the circulating receptor was proteolyzed into smaller fragments (MacDonald et al., 1989;Causin et al., 1988). It has also been shown that the IGF-II/Man-6-P receptor can be found in urine (Causin et al., 1988). These data suggested to us that the serum IGF-II/Man-6-P receptor might be a major intermediate form in the degradation pathway of this receptor protein. In the present experiments, this hypothesis was tested by following the fate of IGF-II/Man-6-P receptor in cultured cells pulse-labeled with ["'S]methionine and ['"S] cysteine. We demonstrate here that the loss of labeled IGF-II/Man-6-P receptor in BRL-SA cells upon incubation of cells with unlabeled amino acids can quantitatively account for the appearance of labeled receptor in the medium. This process is blocked by inhibitors of serine proteases including chymostatin. The data suggest that cellular receptor is degraded by proteolysis at the cell surface followed by release into the medium.

Materials-IGF-II/Man-6-P receptors for antibody generation
were purified from rat placental plasma membranes by IGF-II-Sepharose chromatography as previously described (Oppenheimer and Czech, 1983). The anti-IGF-II/Man-B-P receptor antisera used were those previously described (MacDonald et al., 1989). Cell lines were purchased from the American Type Culture Collection. Cell cultured reagents are from GIBCO. Sodium iodoacetate (Fisher Biotech) was recrystallized twice from methanol. All other chemicals were at least reagent grade.
Radioactive Labeling of Cells-Prior to labeling, cells were grown to 90% confluence. Medium was then removed and replaced by serumfree minimum essential medium buffered with 15 mM Hepes and lacking the labeled amino acid. To this was added 0.15-1.0 mCi of a mixture of [:"S]cysteine and [~"'S]methioninc, either Tran"'S-label (ICN, Costa Mesa, CA) or EXPRE'"S%3 Label (Du Pont-New England Nuclear). Cells were incubated for 30-60 min a t 37 "C, then the medium was removed and replaced by serum-free Dulbccco's modified Eagle's medium (DME). Incubation was continued for the indicated 12131 times in serum-free DME or DME supplemented as indicated prior to homogenization.
Preparation of Membranes from Cultured Cell Lines-Cells were scraped into a buffer consisting of 20 mM Hepes. 0.25 M sucrose, and 1 mM EDTA, pH 7.4, plus protease inhibitors at 0 "C. The protease inhibitors were leupeptin, antipain, and benzamidine at concentrations of 10 mg/ml each, 20 mg/ml aprotinin, 12.5 mg/ml chymostatin, and 1 mM phenylmethylsulfonyl fluoride (PMSF). Following homogenization, total membranes were obtained by a single centrifugation a t 200,000 X g.
To obtain a membrane extract the membranes obtained as described above were solubilized as follows. A volume of the membrane suspension obtained above equivalent to 0.2 mg of protein or the total membrane protein from one well was pelleted by centrifugation for 5 min at 15,000 X g in a microcentrifuge. The supernatant was removed and the pellet resuspended in 50 pl of 15 mM Tris, 0.15 M NaCI, pH 7.4, containing 1% Triton X-100, 1% deoxycholate, and 0.1% sodium dodecyl sulfate with the protease inhibitors described above. Following a 1-h incubation at 4 "C on an end-over-end mixer, the mixture was centrifuged for 10 min at 15,000 X g in a microcentrifuge, and the supernatant fraction was used as the extract. The cellular form of the ICF-II/Man-6-P receptor was isolated from this preparation.
Preparation of Medium for Immunoadsorption-Prior to homogenization of the cells, medium was removed into centrifuge tubes and protease inhibitors added as described above. The medium was then centrifuged at 35,000 X g or 200,000 X g to remove cellular or membrane contaminants. Both techniques produced identical results (data not shown). Medium was frozen at -80 "C until needed. Prior to immunoadsorption, a fraction of the medium corresponding to 0.2mg membranes or the total medium from one well was concentrated using a Centricon 30 (Amicon, Danvers, MA) or a Centrifugal UltraFree with a 30,000 nominal molecular weight limit (Millipore, Bedford, MA) to a volume of 0.1-0.5 ml.
Immunoadsorption of the IGF-IllMan-6-P Receptor--IGF-II/ Man-6-P receptors were immunoadsorbed essentially as described (MacDonald et al., 1989). For immunoadsorption from a membrane extract the 50 pl of extract was diluted to 0.9 ml with 50 mM Hepes, pH 7.4, and to a final concentration of 0.15 M NaCl and 5 mM Man-6-P. For medium, 50 pl of extraction buffer was added to the medium, and it was diluted in the same manner as the membrane extracts. Finally, to each was added 0.1 ml of a 50% slurry of anti-Man-6-P receptor antibody Affi-Gel (Bio-Rad) in 50 mM Hepes, pH 7.4, containing 0.1% Triton X-100. These mixtures were incubated overnight at 4 "C on an end-over-end mixer. Unbound material was removed by withdrawing the supernatant following a 1-min centrifugation in a microcentrifuge at 15,000 X g. This material was then washed in this manner 4 times in 50 mM Hepes, 0.5 M NaCI, pH 7.4, containing 0.1% Triton X-100, then once in each 50 mM Hepes, 0.15 M NaCI, pH 7.4, with 0.05% Triton X-100 and 50 mM Hepes, pH 7.4.
Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis-Samples labeled with were reduced by incubation in electrophoresis sample buffer containing 100 mM dithiothreitol. Electrophoresis was performed as described (Laemmli, 1970) on 6% polyacrylamide gels.
Autoradiogaphy of :''!+Labeled Gels and Quantitation of Results-Gels containing :'nS-labeled material were stained with Coomassie Brilliant Blue and destained. They were then treated with EN"HANCE (Du Pont-New England Nuclear) according to the manufacturer's instructions. The gels were then dried and subjected to autoradiography. Quantitation was performed using an LKB Ultroscan XL densitometer (LKB, Rockville, MD) to scan the autoradiographs and/or by cutting out the bands and counting them in a scintillation counter. Quantitated results shown are means k S.D. for three repeats of each experiment.

RESULTS AND DISCUSSION
In order to determine the relationship between cellular and soluble forms of the IGF-II/Man-6-P receptor, BRL-3A cells were labeled for 1 h with ["SSJmethionine and [:"S]cysteine, then incubated with unlabeled medium for the times indicated (Fig. 1). Immediately following the pulse, there is no labeled receptor found in the medium, and only the 245-kDa precursor is seen in cells. The observed increase in cellular receptor during the early points of the chase period is accounted for by the specificity of the antibody used; the primary translation product (232 kDa) is not recognized by the antibody until it Time@) 0 1 ' 2 3 6 9 12 22 30 "-

Time (hours)
FIG. 1. Quantitation of labeled IGF-II/Man-6-P receptore in medium and cellular membranes of BRL-3A cells at various times following a 1-h labeling period with [""Slmethionine and [3nS]cysteine. BRL-3A rat liver cells were grown to 90% confluence in 35-mm plates, then labeled with 0.4 mCi of [."S]methionine and [:l"S]cysteine for 1 h. The labeling medium was then replaced with serum-free DME for the indicated times. Cells and medium were then separated and receptors purified as described under 'Experimental Procedures." A, total membrane samples from 0.2 mg of membrane protein were electrophoresed on SDS-PACE, stained, destained, and treated with EN'HANCE. A representative autoradiogram is shown. E , medium receptor from one plate was electrophoresed on SDS-PACE, stained, destained, and treated with EN"HANCE. A representative autoradiogram is shown. C, quantitation of the results depicted in panels A and R by cutting of receptor regions from gels and measuring radioactivity in a scintillation counter. Results shown are means * S.D. for three samples for each time point.
has been glycosylated to a 245-kDa form (MacDonald and Czech, 1985). Over the next several hours of the chase period the 245-kDa precursor is converted into the 250-kDa mature receptor, and a truncated receptor is first seen in the medium. At later times, cellular receptor is gradually converted into the truncated form of the receptor found in the medium. Thus, the initial processing of cellular receptor to a form recognized by the antibody prevents quantitation of receptor degradation from the cells over the first few hours (<6 h), and degradation of the medium form of receptor at the later time points prevents quantitation of medium receptor for these times (>22 h). However, quantitation of the results from 6 to 22 h demonstrates that 109 f 39% of the receptor lost from the cells appears in the medium. These results demonstrate that, for this cell line, release of labeled receptor from the cell can account for the loss of labeled receptor from the cells.
In order to determine if release of soluble IGF-II/Man-6-P receptor is cell type-specific or a more general mechanism, a number of cell lines were studied L6 (rat myoblast), CHO-K1 (Chinese hamster ovary), and SL-29 (chick embryo fibroblast) cells. Each cell type was labeled with [%3]methionine and ["S]cysteine for 1 h, then incubated in unlabeled medium for the times indicated (Fig. 2). In each case, labeled IGF-11/ Man-6-P receptor accumulated in the cells a t 0 and about 3.5 h following addition of unlabeled amino acids, then decreased to lower levels a t about 20 h. Labeled serum receptor was absent from the medium at the start of the chase, then accumulated slightly by about 3.5 h, with similar levels seen at about 20 h. These results are similar to the results discussed above for the BRL-3A cells, in that post-translational modification presumably accounts for the additional receptor appearing between 0 and about 3.5 h following the pulse in the cells. During this time, receptor appears in the medium and, as was seen with the BRL-SA cells, serum receptor in the medium is gradually lost, possibly by further degradation and/ or reuptake. Taking these factors into account, it appears that a significant portion of the receptor lost from these cell lines appears in the medium.
The hypothesis that proteolytic release of IGF-II/Man-6-P receptors from the cell surface leads to the appearance of truncated receptor in the medium was tested. The sensitivity of this process to a spectrum of protease inhibitors was evaluated in BRL-3A cells. When added to the chase medium following a 1-h pulse period with ["'SS]methionine and [""S) cysteine, the protease inhibitors aprotinin, chymostatin, or PMSF completely inhibit the production of the receptor in medium at concentrations of 20 mg/ml, 12.5 mg/ml, and 1 mM, respectively. Benzamidine at 10 mg/ml inhibits the process to a lesser extent (Fig. 3), while 10 mg/ml leupeptin or antipain has no effect on the production of serum receptor in the medium as compared with cells incubated in the absence of protease inhibitors for the chase period. While proteolysis from the cell surface would seem to be the most straightforward way to produce a serum form of a receptor when a cellular form already exists, this mechanism has not previously been identified as a significant route whereby serum receptors are produced (Herington et 01.. Leung et al., 1987;Cower et al., 1 9 m Gussow and Pleogh, 1987). The results described above strongly suggest that proteolyis of the cell surface receptor is important for synthesis of the serum form of the ICF-II/Man-6-P receptor. Furthermore, the proteolytic release of the IGF-II/Man-6-P receptor from the cell appears to account for the loss of cellular receptor. This suggests that the serum IGF-II/Man-6-P receptor is an important degradative intermediate in the removal pathway of the cellular receptor.
While proteolytic release from the cell surface has not been shown to play a significant role in the production of other serum receptor forms or in the degradation of any other known protein, a similar proteolytic process occurs in the final stage of transcytosis by the polymeric Ig receptor. In this process IgA or IgM is bound to the receptor at the basolateral surface, transported across the cell, and released from the apical surface (Ceuze et Hoppe et al., 1985). In the final stage, release of polymeric Ig from the apical surface, receptor is cleaved to release a portion of the receptor, referred to as secretory component, along with the immunoglobulin (Mostov and Blobel, 1982;Solari and Kraehenhuhl, 1984;Sztul et al., 1985). However, while the process of the

IGF-IIIMan-6-P Receptor Degradation
release of the secretory component from the polymeric Ig receptor is similar to the release of the serum receptor form of the IGF-II/Man-6-P receptor, differences do exist. First, the cleavage of the polymeric Ig receptor serves to allow for the release of the Ig, while cleavage of the Man-6-P receptor is not known to serve such a purpose. Second, the two processes are different in respect to protease sensitivity since leupeptin can inhibit the cleavage of the polymeric Ig receptor (Mostov and Blobel, 1982) but not the IGF-II/Man-6-P receptor (Fig. 3).
The findings reported here provide strong evidence for the concept that IGF-II/Man-6-P receptors are degraded by an extracellular pathway in the cells studied here. It is possible that some other receptor proteins share this mechanism of degradation. The cation-dependent Man-6-P receptor, which is also excluded from lysosomes, has been demonstrated to appear at the cell surface (Stein et al., 1987), although it is unable to bind to ligand in that compartment (Kyle et al., 1988). It is possible that such a localization is necessary for the degradation of this protein. In addition, a number of nutrient receptors, such as the low density lipoprotein and transferrin receptor, cycle in a manner similar to the IGF-11/ Man-6-P receptor. The transferrin receptor has also been demonstrated to have a serum form (Kongo et al., 1986) whose levels are thought to correlate with levels of cellular receptor under various conditions (Flowers et al., 1989;Trowbridge, 1989). Furthermore, the low density lipoprotein receptor has been demonstrated to be degraded independently of the lysosome (Grant et al., 1989). An important question for future studies is whether these and other receptor proteins participate in a life cycle involving cellular extrusion prior to their eventual degradation.