Soluble LDL-R are formed by cell surface cleavage in response to phorbol esters Michael J. Begg 1, *, Edward D. Sturrock 1 and Deneys R. van der Westhuyzen 2 1 Division of Medical Biochemistry, University of Cape Town, South Africa; 2 Department of Internal Medicine, University of Kentucky Medical Center, and Department of Veteran Affairs Medical Center, Lexington, Kentucky, USA A 140-kDa soluble form of the low density lipoprotein (LDL) receptor has been isolated from the culture medium of HepG2 cells and a number of other cell types. It is produced from the 160-kDa mature LDL receptor by a proteolytic cleavage, which is stimulated in the presence of 4b-phorbol 12-myristate 13-acetate (PMA), leading to the release of a soluble fragment that constitutes the bulk of the extracellular domain of the LDL receptor. By labeling HepG2 cells with [ 35 S]methionine and chasing in the pres- ence of PMA, we demonstrated that up to 20% of LDL- receptors were released into the medium in a 2-h period. Simultaneously, the level of labeled cellular receptors was reduced by 30% in those cells treated with PMA compared to untreated cells, as was the total number of cell surface LDL-receptors assayed by the binding of 125 I-labeled anti- body to whole cells. To determine if endocytosis was required for cleavage, internalization-defective LDL-recep- tors were created by mutagenesis or deletion of the NPXY internalization signal, transfected into Chinese hamster ovary cells, and assayed for cleavage in the presence and absence of PMA. Cleavage was significantly greater in the case of the mutant receptors than for wild-type receptors, both in the absence and presence of PMA. Similar results were seen in human skin fibroblasts homozygous for each of the internalization-defective LDL receptor phenotypes. LDL receptor cleavage was inhibited by the hydoxamate- based inhibitor TAPI, indicating the resemblance of the LDL receptor cleavage mechanism to that of other surface released membrane proteins. Keywords: internalization signal; LDL-R; low density lipo- protein; ectodomain shedding. The low density lipoprotein (LDL) receptor (LDL-R) is a cell surface protein that mediates the uptake and clearance of the cholesterol-rich lipoprotein LDL from the plasma [1]. The LDL-R plays an important role in regulating plasma LDL and cellular cholesterol levels and the activity of this receptor has a direct bearing on plasma cholesterol levels. The LDL receptor is regulated at the transcriptional level in response to intracellular sterol levels by means of sterol sensitive transcription factors SREBP1 and SREBP2 [2,3], however, it can also be upregulated by numerous cytokines [tumor necrosis factor (TNF)-a, interleukin (IL)-1, trans- forming growth factor (TGF)-b, oncostatin M, platelet- derived growth factor and basic fibroblast growth factor (bFGF) [4–8], hormones (insulin and estradiol) [9,10] and second messenger systems [11,12]. Although much is known about the transcriptional mechanisms that control LDL-R expression, little is known about the mechanism(s) that control the turnover and degradation of this important protein. Earlier studies established that the LDL-R degra- dation mechanism(s) is a nonlysosomal process that is dependant on short lived mediator protein(s) and unaffected by the presence of ligand and/or sterol [13,14]. Some evidence suggests that degradation may also be able to modulate LDL-R number. Ness et al.[15]showedthat degradation of LDL-R in hepatocytes is increased in the presence of cholesterol synthesis inhibitors. Kraemer et al. [16] reported that the half-life of LDL-R in rat adipocytes is decreased by 40% in the presence of insulin. The proteolytic mechanism responsible for such regulated LDL-R degra- dation is not known. A regulatory mechanism common to many cell surface proteins is a proteolytic cleavage of the membrane anchor that releases a soluble form into the extracellular medium [17]. Examples include TGF-a,TNF-a, TNF-receptor, angiotensin converting enzyme (ACE), amyloid precursor protein, L -selectin, IL-6 receptor and also members of the LDL receptor gene family, LRP and VLDL-R [17–22]. In addition, this process has been observed to be highly regulated by second messenger systems [23]. A common feature is the release of the extracellular domain via a single proteolytic cleavage at a site just extracellular to the membrane spanning domain. In some cases further proteo- lytic processing of the extracellular domain occurs. The function of these soluble forms varies. In the cases of ACE, TGFa and other mitogens the soluble protein has the Correspondence to E. D. Sturrock, Division of Medical Biochemistry, University of Cape Town, Observatory 7925, South Africa. Fax: + 27 21 406 6470, Tel.: + 27 21 406 6312, E-mail: sturrock@curie.uct.ac.za Abbreviations: LDL, low density lipoprotein; LDL-R, low density lipoprotein receptor; sLDL-R, soluble low density lipoprotein receptor;PMA,4b-phorbol 12-myristate 13-acetate; LPDS, lipopro- tein deficient serum; CHO, Chinese hamster ovary; HSF, human skin fibroblasts; TAPI, TNF-a protease inhibitor; TNF, tumor necrosis factor; IL, interleukin; TGF, transforming growth factor; ACE, angiotensin converting enzyme; PKC, protein kinase C. *Present address: Ribotech Pty Ltd, Biopolymer Unit, 15 A Mail Street, Western Province Park, Good wood 7460, South Africa. (Received 3 September 2003, revised 27 November 2003, accepted 2 December 2003) Eur. J. Biochem. 271, 524–533 (2004) Ó FEBS 2004 doi:10.1046/j.1432-1033.2003.03953.x potential of carrying out its function at sites remote from the original cell surface [19]. Another function may be the rapid downregulation of the cell associated protein. In other cases the released soluble protein has been associated with pathological conditions such as Alzheimer’s disease where aberrant cleavage of the amyloid precursor protein forms the soluble b-amyloid peptide which deposits in plaques resulting in neuronal degeneration. In almost all cases the formation of these soluble proteins is greatly enhanced by phorbol esters and inhibited by hydroxamate-based inhi- bitors such as TAPI (TNF-a protease inhibitor) and batimastat. In this report, we have investigated the degradation of LDL-R in HepG2 cells and show that LDL-Rs are degraded in part by a proteolytic mechanism which cleaves the receptor close to the transmembrane domain, resulting in the release of the extracellular domain as a soluble LDL-R (sLDL-R). Like the release of many other transmembrane proteins, the cleavage of LDL-R is markedly activated by phorbol esters. Investigations using mutant LDL-R and protease inhibitors suggest that cleavage takes place at the cell surface and that the mechanism is closely related to that which generates soluble derivatives of other transmembrane proteins. Experimental procedures Materials All tissue culture media was from Highveld Biological, Kelvin, South Africa. Fetal bovine serum was from delta bioproducts, Kempton Park, South Africa. Human LDL (density, 1.019–1.063 gÆmL )1 ) and lipoprotein deficient serum (LPDS) (density > 1.25 gÆmL )1 ) were prepared from whole, male blood, and iodinated by the iodine monochloride method as described previously [24]. IgG-C7 was prepared as described [24] from hybridoma cells obtained from the American Type Culture Collection (CRL 1691). Goat anti-mouse (IgG subfraction) was from Cappel Research Products (Durham, NC, USA.) [ 35 S]Methionine as Tran 35 Slabel TM , and methionine/cys- teine free media were from ICN Radiochemicals (Irvine, CA, USA). Protease inhibitors and 4b-phorbol 12-myristate 13-acetate (PMA) were from either Sigma Chemical Co. or Boehringer Mannheim. TAPI was from Immunex. Mutant fibroblasts GM2408 (HSF-JD) and FH683 (HSF-792stop) and the monoclonal antibodies 4A4 (raised against LDL-R cytoplasmic domain) and HL-1 (raised against LDL-R ligand binding domain) were a kindly supplied by M. S. Brown and J. L. Goldstein (Dallas, TX, USA). Generation of the cDNA constructs Construction and expression of LDL-R mutants. To investigate whether endocytosis plays a role in the ecto- domain shedding of the LDL receptor, three cytoplasmic domain mutants were made (Fig. 1). The mutant 792-LDL- R has a truncated cytoplasmic domain of only two amino acids due to a single base substitution (TGG fi TGA), which converts Trp792 to a stop codon. The mutant receptor, JD-LDL-R (Tyr807Cys), has a single amino acid substitution that disrupts the NPVY internalization signal. To assess whether other signals in the LDL-R cytoplasmic tail influence cleavage, we constructed a third mutant (812- LDL-R), which contains a functional NPVY internalization sequence but lacks the last 27 amino acids of the cytoplasmic domain. Mutagenesis of the LDL-R cytoplasmic tail was carried out according to the method described [25]. Briefly, a 2.1-kb EcoR1/Sma1 fragment was subcloned from the LDL-R expression plasmid pLDL-R2 into the bacteriophage vector, M13mp18. Mutagenic oligonucleotides and primer exten- sion were used to generate the mutated double-stranded vector according to the method of Kunkel et al.[26].After sequencing, to confirm the mutations, a 1.1-kb BglII–SmaI fragment containing the cytoplasmic, transmembrane and O-linked sugar domains was subcloned into pLDL-R2. Fig. 1. LDL-R constructs. LDL-R mutations JD-LDL-R, 812-LDL-R and 792-LDL-R are depicted in relation to the wild-type (WT) receptor. Each of the mutants is shown as an expanded view of the juxtamembrane, transmembrane and cytoplasmic domains in which all these mutations occur. Above the linear bar diagram of the WT-LDL-R is the amino acid sequence with a cysteine which has been substituted for Tyr807 in JD-LDL-R indicated by an asterisk. Ó FEBS 2004 Cell surface cleavage of LDL-R (Eur. J. Biochem. 271) 525 Transfection of the mutant plasmids into LDL-R-negative CHO cells (CHO-A7) and selection of positive clones were carried out according to the procedures described by Davis et al.[27]. Metabolic labeling with [ 35 S]methionine Semi-confluent HepG2 cells were seeded into 35-mm dishes at a split ratio of 1 : 4 and cultured at 37 °CinMEM containing 10% (v/v) fetal bovine serum. After 36 h, LDL- R activity was upregulated by replacing medium with MEM containing LPDS (2.5 mgÆmL )1 ) and culturing for a further 12 h at 37 °C [24]. Cells were then metabolically labeled with [ 35 S]methionine by incubating in methionine/ cysteine-free EMEM/LPDS for 30 min and then for 2 h in methionine/cysteine-free EMEM/LPDS containing 50 lCiÆmL )1 [ 35 S]methionine. The medium was changed to complete MEM/LPDS containing 200 l M unlabeled methionine and incubated at 37 °C for various chase times in the presence or absence of PMA. After the indicated chase periods, the medium was removed from the cells and spun at 15 000 g for 10 min before adding one-tenth volume of 100 m M Hepes pH 7.4, 500 m M NaCl, 20 m M MgCl 2 ,0.5m M leupeptin, 10 m M phenylmethanesulfonyl fluoride and 10% (v/v) Triton X-100. The cells were then washed in buffer A (10 m M Hepes pH 7.4, 150 m M NaCl, 2m M CaCl 2 ), and cell associated LDL-R solubilized in buffer B (10 m M Hepes pH 7.4, 200 m M NaCl, 2 m M CaCl 2 ,2.5m M MgCl 2 ,1m M phenylmethylsulphonyl flou- ride, 0.02 m M Leupeptin, 1% (v/v) Triton X-100). Both medium and cells were immunoprecipitated using pre- formed immune complexes of the monoclonal antibody IgG-C7 as described [28]. The immunoprecipitates were separated by SDS/PAGE, enhanced with salicylate and visualized by fluorography, or direct detection on a Packard Instant Imager 2024 (Packard Instrument Company). Surface LDL-R binding of 125 I-labeled LDL or 125 I-labeled IgG-C7 Human LDL (density, 1.019–1.063 gÆmL )1 )andthe monoclonal antibody IgG-C7 were prepared and labeled with 125 I as described [24]. HepG2 cells were seeded into 35-mm dishes at a split ratio of 1 : 4 and cultured in MEM containing 10% (v/v) fetal bovine serum. After 36 h, LDL-R activity was upregulated by incubating cells in MEM/LPDS (2.5 mgÆmL )1 ) for a further 12 h. Surface LDL-R activity was measured by incubating cells with 1 mL ice cold MEM/LPDS buffered with 20 m M Hepes pH 7.4, containing either 125 I-labeled LDL (10 lgÆmL )1 ) or 125 I-labeled IgG-IgG-C7 (1 lgÆmL )1 )for2hat4°Cas described [24]. This medium was removed and cells were washed four times with 2 mL NaCl/P i containing 0.2% BSA, followed by three times with 2 mL NaCl/P i .The specifically bound fraction of 125 I-labeled LDL that remained cell associated after washing was removed by incubating cells with 0.4% heparin for 1 h at 4 °Cand counted. Bound 125 I-labeled IgGC7 was determined by solubilizing cells in 1 M NaOH and measuring the associated counts. Nonspecific counts, determined by adding excess unlabeled ligand, was subtracted from total counts to give specific binding. Results Phorbol ester enhances the release of sLDL-R The LDL receptor typically has a half-life of  10–12 h in fibroblasts and Chinese hamster ovary (CHO) cells [13,27]. To determine if LDL-R is cleaved in a similar fashion to certain other cell surface proteins, HepG2 cells were pulse- labeled with [ 35 S]methionine and chased in the absence or presence of PMA. The cells (C) and medium (M) were then immunopreciptated with LDL-R-specific antibody IgGC7 and subjected to SDS/PAGE and autoradiography (Fig. 2). Immunoprecipitates of labeled HepG2 cells with the IgG- C7 antibody showed the characteristic 120-kDa precursor LDL receptor present at the start of the chase period but not at later time points when only the 160-kDa mature receptor is seen. The addition of PMA to labeled cells resulted in a marked increase in the levels of a 140-kDa immunopreci- pitable protein in the extracellular medium. The presence of this protein in the medium of untreated cells was hardly noticable (Fig. 2). Recognition of this protein by immune- complexes of the IgG-C7 antibody suggest that it is a soluble form of LDL-R. Immunecomplexes of a second monoclonal antibody HL-1 demonstrated equivalent results to the IgG-C7 antibody (data not shown). Human skin fibroblasts from a receptor negative familial hypercholesterolemia (FH) patient (NS) were analyzed in the same manner for LDL-R cleavage (Fig. 3). The 140-kDa protein was absent from the culture medium of PMA-treated FH fibroblasts but present in the medium of HepG2 cells and normal fibroblasts. Downregulation of LDL-R in the normal cells by pretreatment with 25-hydroxycholesterol resulted in the absence of any detectable 140-kDa soluble protein (Fig. 3). The normal LDL-R expressed in transfected CHO cells (CHO-715) displayed significant 140-kDa protein in the medium while the parent LDL-R-negative cells (CHO-A7) showed no immunoprecipitable protein in the chase medium. Taken together, these data clearly suggest the identity of the 140-kDa protein as a soluble form of LDL-R. Fig. 2. Characterization of the soluble form of LDL-R induced by PMA. HepG2 cells were metabolically labeled for 2 h with [ 35 S]methionine (lane 1) followed by a 4-h chase period (lanes 2–4) in the presence or absence of PMA (100 ngÆmL )1 ). Both cells (C) and medium (M) were collected and immunoprecipitated with IgG-C7, followed by SDS/PAGE and autoradiography. 526 M. J. Begg et al. (Eur. J. Biochem. 271) Ó FEBS 2004 Based on the mobility of the soluble protein on SDS/ PAGE (140 kDa for sLDL-R vs. 160 kDa for the mature LDL-R), and the fact that a monoclonal antibody (4A4) directed against the cytoplasmic domain of LDL-R does not recognize the 140-kDa protein (data not shown), it is likely that the sLDL-R consists of the bulk of the extracellular domain of the LDL receptor. This protein could be generated either by alternative splicing of LDL-R mRNA, eliminating sequences responsible for anchorage in the membrane, or by proteolytic cleavage of the transmembrane LDL-R. mRNA splicing is less likely as sLDL-R was detected only in the medium and not in the associated cells either during or after the pulse. Furthermore, soluble receptors were only readily detected once PMA was added to the medium and this occurred even when labeling of newly synthesized receptors had ceased, i.e., during the chase period. Incorporation of [ 35 S]methionine into newly synthesized receptors ceased within 15 min of commencing the chase as no more 120-kDa precursor LDL receptor protein could be detected after this time period. Even after a 4-h chase period the addition of PMA stimulated the generation of labeled soluble receptors (data not shown). This would not be the case if mRNA splicing was the source of the truncated sLDL-R. The rate at which sLDL-R was released from PMA- treated and untreated HepG2 cells was assessed as shown in Fig. 4. Cells treated with PMA (closed symbols) released sLDL-R at a rate significantly faster than untreated cells (open symbols). The accelerated release induced by PMA lasted for about 2 h after which the rate of release tended to slow down. The slowdown in LDL-R cleavage was not due to sLDL-R degradation in the medium as no significant loss of sLDL-R was detected during a 20-h incubation of sLDL-R-containing medium at 37 °C (data not shown). By 2 h, 18.7% (± 3.5; n ¼ 5) of the total labeled LDL-R was detected as soluble receptor in PMA-treated cells, compared to 4.8% (± 2.1; n ¼ 5) released from untreated cells in the same time period. The regulation of cleavage was further characterized by determining the effect of PMA on sLDL-R release (Fig. 5). No effect was detected at PMA concentrations <1 ngÆmL )1 and sLDL-R release was increased between 1ngÆmL )1 and 10 ngÆmL )1 PMA. No significant increase in sLDL-R is seen > 10 ngÆmL )1 , although in one experiment maximum release was achieved only at 30 ngÆmL )1 .The response of cells to the different PMA concentrations varied between experiments as indicated by the relatively large error bars. This variation is thought to be systematic as all values were either high or low depending on the experiment. The requirement for an active protein kinase C (PKC) in the PMA response was established by using the PKC inhibitor staurosporine (10 l M ) which almost completely abolished the enhanced release at 100 ngÆmL )1 PMA (closed dia- mond, Fig. 5). The number of surface LDL-R is affected by PMA To assess whether PMA-stimulated release of sLDL-R alters the number of LDL-Rs on the cell surface, HepG2 cells were incubated for various times in the presence of PMA after which the number of cell surface receptors was assessed by binding of an anti-LDL-R monoclonal Ig (IgG- C7)at4°C. A 1-h treatment of HepG2 cells with PMA resulted in a 30% decrease in the number of cell surface LDL-R as measured by labeled antibody (Fig. 6). Incuba- ting the cells with PMA for longer periods at 37 °Cresulted in a reversal of the decreased receptor number seen at 1 h, such that by 4 h the number of cell surface LDL-Rs had doubled. This increase is in all likelihood due to PMA stimulation of the PKC dependent, p42/44 MAPK induction of LDL-R transcription in HepG2 cells [29,30]. The initial Fig. 4. Kinetics of sLDL-R release induced by PMA. HepG2 cells, metabolically labeled for 2 h with [ 35 S]methionine, were chased in unlabeled MEM/LPDS for the indicated periods in the presence (d)or absence (s)ofPMA(100ngÆmL )1 ). After the indicated chase periods, the medium was removed, immunoprecipitated with IgG-C7, subjec- ted to SDS/PAGE and followed by autoradiography and quantifica- tion. These data represent the mean (± SEM) of duplicates from three experiments. Fig. 3. Production of sLDL-R by different cell types. HepG2 cells, normal HSF and FH HSF, were either upregulated in DMEM/LPDS (up) or downregulated in DMEM/fetal bovine serum + 1 lgÆmL )1 25OH-cholesterol (down) for 24 h prior to metabolic labeling with [ 35 S] methionine. CHO cells transfected with human LDL-R (715) and its LDL-R negative parent cell line (A7) were maintained in full medium. All cells were pulse-labeled with [ 35 S]methionine for 2 h and chased in DMEM/LPDS in the presence of PMA (100 ngÆmL )1 )for 4 h. The medium was removed from the dishes and subjected to immunoprecipitation and autoradiography described. Ó FEBS 2004 Cell surface cleavage of LDL-R (Eur. J. Biochem. 271) 527 decrease in cell surface receptor number following PMA treatment was also seen in CHO cells transfected with human LDL-R (CHO-715); however, in these cells the number of surface receptors remained below the control for an extended period and no reversal of this effect was detected (data not shown). This is probably because the transfected gene is not under the control of its native promoter. The loss in LDL receptor surface binding was supported by degradation studies, which demonstrate that PMA enhances the loss of [ 35 S]methionine-labeled receptors significantly (Fig. 7), such that by 2 h, PMA-treated cells have 30% less labeled receptors than untreated cells. Given the lack of steady state conditions in this experiment, it was not possible to determine accurately the half-life of LDL-R following PMA treatment. As an estimate, the apparent half-life of LDL-R was  2 h in PMA-treated cells compared to 5–6 h in untreated cells. The discrepancy between soluble receptor detected in the medium (18.7%) in 2 h vs. the increased loss of total receptor following PMA treatment (40% less labeled receptors in treated cells vs. untreated cells after 2 h) is an indication that other proteolytic pathways are also stimulated by PMA. Other degradative pathways may include the generation of the 125-kDa Band X as reported by Lehrman et al.[31]. Internalization deficient LDL-R undergo increased cleavage While sLDL-R of 140 kDa was readily detected in the medium of PMA-treated cells, no protein of this size was detected in cell lysates, suggesting that receptor cleavage takes place at or near the cell surface, possibly in the endosomal compartment. In order to ascertain if endo- cytosis plays a role in cleavage, two mutant LDL-Rs (792-LDL-R and JD-LDL-R) were constructed which are unable to undergo endocytosis via coated pits. The muta- tions were confirmed to be internalization defective as assayed by 125 I-labeled LDL uptake, with the rates of internalization being 10% and 25% of normal for 792-LDL-R and JD-LDL-R, respectively. The transfected CHO cell lines, CHO-792 and CHO-JD, were pulse-labeled with [ 35 S]methionine and chased in the presence or absence of PMA for 2 h (Fig. 8A). Both 792-LDL-R and JD-LDL-R were cleaved to a significantly greater extent than the wild- type LDL-R. After the 2-h chase period in the presence of PMA,  70% of the total population of labeled 792-LDL-R was present as sLDL-R in the medium, compared to 42% of JD-LDL-R and  20% of wild-type receptors (Fig. 8B). In the absence of PMA, the degree of cleavage for 792-LDL-R, JD-LDL-R and wild-type LDL-R was 40%, 11% and 6%, respectively. These results indicated that endocytosis via coated pits is not required for cleavage and in fact it may inhibit cleavage of LDL-R. In addition it suggested that cleavage takes place on the cell surface. The LDL-R mutant 812-LDL-R that is truncated after Thr811 (Fig. 1) was used to investigate whether other Fig. 6. Effect of PMA on surface LDL-R number. HepG2 cells upregulatedfor12hinMEM/LPDSwereincubatedat37°Cinthe presence (d) or absence (s)ofPMA(100ngÆmL )1 ) for the indicated time period. They were then cooled to 4 °C and incubated for 2 h in the presence of 125 I-labeled monoclonal antibody IgGC7. After sub- stantial washing the remaining cell associated label was determined and normalized to cell protein in each dish. The graph represents the level of IgG-C7 binding to the cell surface as a percentage of the zero hour value. Error bars represent the SEM of four values. Fig. 5. Effect of PMA dose on sLDL-R production. HepG2 cells were pulse-labeled with [ 35 S]methionine and chased for 4 h in the presence (d) or absence (s) of indicated doses of PMA, or in the presence of 100 ngÆmL )1 PMA and 10 l M staurosporine (r). The medium was immunoprecipitated and subjected to SDS/PAGE as described. The dried gels were then exposed to electronic autoradiography and quantitation. The error bars represent the range of four data points from two experiments, with the symbol representing the mean. 528 M. J. Begg et al. (Eur. J. Biochem. 271) Ó FEBS 2004 signals in the LDL-R cytoplasmic tail influence cleavage. The last 27 amino acids of the cytoplasmic domain are required for receptor dimerization [32] and also contain a phosphorylation site at Ser833 [33]. These receptors were shown to be internalization competent (data not shown). Deletion of this domain had no significant effect on cleavage (Fig. 8B) showing that neither the phosphorylation site, nor receptor dimerization appear to play any role in the formation of the 140-kDa sLDL-R. Two of the three cytoplasmic-mutants, JD-LDL-R and 792-LDL-R occur in patients with FH. To confirm the cleavage process observed in the transfected cells, assays were carried out using human skin fibroblasts (HSF) from patients homozygous for these mutations. Both HSF-JD (GM2408) and HSF-792 (FH683) behaved in a very similar fashion to their transfected CHO cell counterparts when stimulated with PMA. The percentage of labeled receptors released as sLDL-R in a 2-h chase period following PMA treatment was 85.9% (± 6.0; n ¼ 4) for HSF-792 compared to 70.1% (± 10.9; n ¼ 6) for CHO- 792. Likewise the percentage release of the JD mutant receptors was 42.2% (± 1.2; n ¼ 4) in human skin fibroblasts and 41.6% (± 8.9; n ¼ 8) in CHO cells. Wild- type LDL receptor release was also very similar – 18.7% for HSF and 20.7% for CHO cells. The most notable difference between CHO cells and HSF cells was that the unstimulated release was much lower in HSF cells than in their CHO counterparts. For example, unstimulated wild- type CHO cells released 6.6% of their LDL-R in 2 h compared to 0.9% for wild-type HSF cells. Similarly, CHO-792 released 40.4% compared to 18.5% HSF. The phenotype of the cytoplasmic domain mutations is thus a direct result of the mutation itself and cannot be ascribed as an artifact of transfection and receptor overexpres- sion. Furthermore, the cell surface cleavage mechanism appeared more sensitive to PMA regulation in HSF than in CHO cells. Enzymes responsible for ectodomain shedding of cell surface proteins have remained largely unidentified except for a few proteases such as ADAM 10 and ADAM 17 (TNF-a converting enzyme or TACE) [34]. From protease inhibitor studies these proteinases fall into two main classes: (a) elastase-like serine proteinases as for c-kit receptor ligands KL-1 and KL-2 [23] and (b) metalloproteinases as for ACE and TNF-a [35,36]. To characterize the protease responsible for the generation of sLDL-R, HepG2 cells were pulse-labeled with 35 S-methionine and chased in the pres- ence of various protease inhibitors. Table 1 shows the effects of multiple inhibitors on sLDL-R release. Only the metalloproteinase inhibitors showed any significant inhibi- tion of release. TAPI, a hydroxamate-based metallopro- teinase inhibitor, shown to inhibit the cleavage of TNF-a, IL-6R, ACE and others [37], inhibited sLDL-R production by as much as 90% (Table 1). The other metalloproteinase inhibitors, EDTA and EGTA inhibited the release by 50–70%. Fig. 8. Effect of cytoplasmic mutations on LDL-R cleavage. LDL-R negative CHO cells (CHO-A7) were transfected with pLDL-R2 con- taining the requisite mutations, and stable clones were selected and seeded into 35-mm dishes. After 24 h, cells were labeled with [ 35 S]methionine for 2 h and chased in unlabeled medium in the pres- ence or absence of PMA for a further 2 h. Immunoprecipitates of cells and medium were subject to SDS/PAGE and fluorography. (A) Autoradiographs of respective mutants labeled as above; for each cell type: (lane 1) cells after 2-h pulse; (lanes 2 and 3) medium after 2-h pulse and 2-h chase in the absence (lane 2) or presence (lane 3) of PMA. (B) The medium bands from B were quantified and expressed as a percent of total receptor label at time zero. The data represents the mean (± SEM) of duplicates from four experiments. Fig. 7. Effect of PMA on cellular LDL-R turnover. HepG2 cells labeled for 2 h with [ 35 S]methionine, were first chased in unlabeled MEM/ LPDS for 1 h followed by a further chase in the presence (d)or absence (s) of PMA (100 ngÆmL )1 ) for the indicated times, after which the cells were solubilized, immunoprecipitated and subject to electronic autoradiography and quantification. The points reflect the quantity of labeled LDL-R remaining in the cell as a percentage of the 0 time point at the start of the chase period. These data represent the mean (± SEM) of duplicates from four experiments. Ó FEBS 2004 Cell surface cleavage of LDL-R (Eur. J. Biochem. 271) 529 Discussion In this study we have investigated LDL-R degradation in HepG2 cells, and report that a 140-kDa sLDL-R is released into the medium by a proteolytic mechanism sensitive to phorbol-ester induction and inhibited by TAPI, a metallo- protease inhibitor. Such solubilizing proteolysis occurs for a number of transmembrane proteins, including TGF-a, TNF-a,TNF-R,ACE,amyloidprecursorprotein, L-selectin and IL-6 receptor [17–20]. The release of sLDL- R into the medium after phorbol ester induction is accompanied by a decrease in both surface LDL-R number (Fig. 6) and total cellular LDL-R (Fig. 7). A similar loss of LDL-R binding after PMA treatment has been reported in U937 cells but the mechanism responsible for this is unknown [38]. The formation of sLDL-R was found in various cell types when stimulated with phorbol esters (Fig. 3). Soluble LDL- Rs have previously been reported to be in the medium of CHO cells that are defective in O-linked glycosylation [39], and to be produced by cells in response to interferon [40]. In the latter case, the 28-kDa soluble receptor, which consists of the N terminus of the receptor, has marked antiviral activity by interfering with vesicular stomatitis virus assem- bly and budding. This 28-kDa N-terminal domain is contained within the 140-kDa soluble receptor reported in this study. Other members of the LDL receptor family, LRP and VLDL-R also undergo surface proteolysis to generate soluble ectodomains. In the case of VLDL-R, the soluble fragment, as well as the corresponding region of the LDL- R, binds minor group rhinoviruses and inhibits viral infection in HeLa cells [21,22]. The important question is whether the cleavage mechanism responsible for the 140 kDa sLDL-R is operative in vivo,andifso,whatis the function of such soluble receptors? Apart from potential antiviral activity, a possible function of these sLDL-R wouldbetobindligand(LDL)andthusinterferewithits uptake and clearance from the plasma by cell bound LDL-R. This occurs in the case of soluble growth factor receptors and soluble cytokine receptors, where the soluble receptors have been shown to act as antagonists by binding to their respective ligands and thereby reducing their effects [19]. Such soluble complexes of receptor and ligand are reported to stabilize the cytokine or growth factor in the extracellular fluid [41]. On the other hand, some cytokines and their receptors, such as IL-6 and its receptor, can act as potent agonists on cells [42]. Cleaved membrane proteins are also involved in various diseases, such as Alzheimer disease and Heymann nephritis [43]. In both of these the pathology is a result of deposition of solublized membrane proteins in plaques. It is tempting to speculate that sLDL-R could become deposited in atherosclerotic plaques and act as a trap for LDL at these sites. Further work is needed to establish to what extent this pathway is operative in vivo and what the potential in vivo activators of this pathway might be. Transferrin receptors and asialoglycoprotein receptors also display a decrease in cell surface number in response to PMA treatment of HepG2 cells [44]. In the case of the latter two receptors, reduced binding is due to redistribution of receptors to intracellular compartments [44]. This redistri- bution scenario does not hold true for the LDL-R as the loss of surface binding is equivalent to the loss of total receptors; also, the mutant LDL-R that is unable to undergo endocytosis and is thus restricted to the cell surface is more susceptible to surface cleavage than the wild-type LDL-R. These mutant receptors also demonstrate an equivalent loss of surface binding (data not shown), indicating that redistribution from the cell surface is not required for the loss in surface receptor number. Furthermore, the kinetics of LDL-R binding loss do not match the much faster kinetics of transferrin receptor and asialoglycoprotein receptor redistribution [44]. The cytoplasmic domain of LDL-R appears to contain elements that are able to modulate cleavage. Receptors with a deletion of the entire cytoplasmic domain (792-LDL-R) or Table 1. The effect of protease inhibitors on the release of sLDL-R into the medium. HepG2 cells were labeled for 2 h with [ 35 S]methionine and then chased in DMEM/LPDS plus 100 ng mL )1 PMA for a further 2 h in the presence of various protease inhibitors. The medium was immuno- precipitated and subjected to SDS/PAGE and quantitative autoradiography as indicated in the methods section. The degree of release was calculated as a percentage of the zero inhibitor control. NA, not applicable; n, Number of experiments performed; for each experiment duplicate dishes were used for each inhibitor; SEM, Standard error of the mean. Protease inhibitors class Protease Concentration of inhibitor sLDL-R release (% of control) SEM n None – – 100 0 9 ALLN Cysteine 260 l M 116 22 4 Leupeptin cysteine/serine 200 l M 128 27 6 E-64 Cysteine 280 l M 91 9 3 Pepstatin Aspartic 1 l M 96 7 3 PMSF Serine 1 m M 79 7 4 Pefabloc Serine 1 m M 92 NA 1 Aprotinin Serine 1.5 l M 101 NA 2 3,4-dichloroisocoumarin Serine 100 l M 128 NA 2 TLCK Serine 130 l M 98 8 4 EDTA Metallo 10 m M 33 18 3 EGTA Metallo 10 m M 50 29 3 Phosphoramidon Metallo 500 l M 95 NA 2 TAPI Metallo 10 l M 10 NA 2 530 M. J. Begg et al. (Eur. J. Biochem. 271) Ó FEBS 2004 an amino acid subsitution in the NPVY internalization signal (JD-LDL-R) become hypersensitive to solubilizing cleavage (Fig. 8), presumably due to prolonged residence times on the cell surface and thus more frequent exposure to a cell surface cleaving protease. This hypothesis is supported by an inverse correlation between the ability to internalize via coated pits (internalization index) and the sensitivity of LDL-R to cleavage. 792-LDL-R which has the lowest internalization index (10% of normal) is cleaved at the highest rate, whereas JD-LDL-R is internalized more efficiently than 792-LDL-R (25% of normal) and is less sensitive to ectodomain cleavage than 792-LDL-R, but still cleaved to a greater extent than wild-type receptors. We cannot, however, rule out that changes in the tertiary structure of the cytoplasmic domain are responsible for the enhanced release. In any case, the modulatory element(s) must reside in the N-terminal 22 amino acids of the cytoplasmic domain (Lys790–Glu812) containing the NPVY internalization signal, as a deletion of the last 27 amino acids of the cytoplasmic tail (812-LDL-R) has no effect on cleavage. The modulatory element may in fact be the NPXY motif itself, as cytosolic adaptor proteins containing phospho-tyrosine binding domains are known to interact with the NPXY motif. ARH and Disabled 1 are two cytosolic proteins that have been shown to bind to the LDL-R cytoplasmic domain [45,46]. Recent studies have demonstrated the interaction of numerous signal transduc- tion proteins with the cytoplasmic domains of members of the LDL-R gene family [47]. In contrast to the large release measured for 792-LDL-R in both CHO and HSF cells, Lehrman et al.[31]didnot detect any soluble receptor in the medium of HSF-792. This may be due to the immunoprecipitation protocol used, as we have found that immunoprecipitation with IgG-C7 and Protein A sepharose does not precipitate any product from the extracellular medium whereas precipitation with immunecomplexes of IgG-C7 and goat antimouse IgG, or IgG-C7 linked to sepharose beads precipitates significant amounts of sLDL-R. Of the proteases responsible for the solubilizing of surface proteins, the best characterized is TACE [48–50]. In the main, the solubilizing enzymes responsible for the large array of ectodomain cleavages have not been identified. Two potential classes of solubilizing proteases appear to exist, a serine class with elastase-like sequence specificity [23], and a metalloproteinase class, of which TACE is an example. In our study, only metalloprotein- ase inhibitors inhibited cleavage, with TAPI inhibiting cleavage by up to 90%. TAPI and other closely related hydroxamate-based inhibitors have proved useful in inhibiting the cleavage of proteins such as TNF-a, L -selectin, p60 TNF receptor and IL-6 receptor as well as a number of other cell surface proteins [36,51–54]. The specificity with which TAPI inhibits the cleavage of LDL-R suggests a close relationship between the LDL-R cleaving protease and the family of protease’s of which TACE is a part. This family belongs to a subgroup of adamalysin-like metalloproteinases known as ADAMs, which contain both a disintegrin and a metalloproteinase domain [55,56]. Our data are supported by Guo et al.[57]whodemonstrate the accumulation of amongst others, sLDL-R in the medium of TACE–/– DRM cells transfected with TACE. No sLDL-R is detected in the medium of the untransfected TACE–/– cells [57]. Recent evidence also suggests that reactive oxygen species, including nitric oxide (NO) are responsible for mediating the PMA induced activation of enzymes such as TACE, thus providing a potential natural mechanism for inducing cleavage of LDL-R in vivo [58,59]. Turnover of LDL-R in HepG2 cells is rapid with a half- life in the order of 4 h (Fig. 7) compared to turnover in CHO cells and fibroblasts (t ½ , 10–12 h) [13]. This difference may be significant in the overall regulation of cholesterol homeostasis within the liver, as the faster the rate of turnover of a protein, the more rapid is the response to transcriptional downregulation. In contrast to our findings, Tam et al. [60] measured the half-life of LDL-R in HepG2 cells to be 9–10 h. We are unable to explain the differences in half-life measured in the same cell type; however, evidence from circadian rhythms in rat liver LDL-R expression suggests a LDL-R half-life of  6 h in the liver [61]. The enhanced LDL-R degradation measured in Fig. 7 in the presence of phorbol ester, could not be completely accoun- ted for by the increased release of receptors into the medium, as the increase in [ 35 S]methionine labeled sLDL-R in the medium was approximately 2.5-fold less than the decrease of [ 35 S]methionine labeled cellular LDL-R over the same time period. In addition there was no significant degradation of sLDL-R in the medium. We therefore conclude that PMA induces more than one degradation mechanism for LDL-R, of which the formation of sLDL-R contributes about 50%. Another degradation pathway may involve the production of a 125 kDa LDL-R degradation intermediate referred to a band X, reported by Lehrman et al. [31], which is not recognized by the anti- N-terminal Ig (IgG-C7) used in this study, and is presum- ably due to cleavage of the N terminus from the remainder of the receptor. This potential pathway requires further investigation. In conclusion, LDL receptor number, at least in HepG2 cells, can be regulated not only by transcriptional control but also by enhanced turnover in response to unknown stimuli that signal via PKC. This enhanced turnover involves at least two proteolytic systems, one of which involves cell surface cleavage and the release of soluble receptors into the extracellular medium. Cell surface clea- vage of the LDL-R represents the first described mechanism that can regulate LDL-R expression post-translationally. Acknowledgements We thank S. Jones and L. Gatterdam for excellent technical assistance. This work was supported by American Heart Association grant 9750284 N (D. R. v.d.W.), and by the South African Medical Research Council (E. D. S and D. R. v.d.W). References 1. Brown, M.S. & Goldstein, J.L. (1986) A receptor-mediated pathway for cholesterol homeostasis. Science 232, 34–47. 2. Yokoyama, C.A.U., Wang, X., Briggs, M.R., Admon, A., Wu, J., Hua, X., Goldstein, J.L. & Brown, M.S. (1993) SREBP-1, a basic- helix-loop-helix-leucine zipper protein that controls transcription of the low density lipoprotein receptor gene. Cell 75, 187–197. Ó FEBS 2004 Cell surface cleavage of LDL-R (Eur. J. Biochem. 271) 531 3. Hua, X., Yokoyama, C., Wu, J., Briggs, M.R., Brown, M.S., Goldstein, J.L. & Wang, X. (1993) SREBP-2, a second basic-helix- loop-helix-leucine zipper protein that stimulates transcription by binding to a sterol regulatory element. Proc. Natl Acad. Sci. USA 90, 11603–11607. 4. Roth, M., Emmons, L.R., Perruchoud, A. & Block, L.H. (1991) Expressions of the low density lipoprotein receptor and 3-hydroxy- 3-methylglutaryl coenzyme A reductase genes are stimulated by recombinant platelet-derived growth factor isomers. Proc. Natl Acad. Sci. USA 88, 1888–1892. 5. Grove, R.I., Mazzucco, C.E., Radka, S.F., Shoyab, M. & Kiener, P.A. (1991) Oncostatin M up-regulates low density lipoprotein receptors in HepG2 cells by a novel mechanism. J. Biol. Chem. 266, 18194–18199. 6. Moorby, C.D., Gherardi, E., Dovey, L., Godliman, C. & Bowyer, D.E. (1992) Transforming growth factor-beta 1 and interleukin-1 beta stimulate LDL receptor activity in Hep G2 cells. Athero- sclerosis 97, 21–28. 7. Stopeck, A.T., Nicholson, A.C., Mancini, F.P. & Hajjar, D.P. (1993) Cytokine regulation of low density lipoprotein receptor gene transcription in HepG2 cells. J. Biol. Chem. 268, 17489– 17494. 8. Liao, W. & Floren, C.H. (1993) Tumor necrosis factor up-reg- ulates expression of low-density lipoprotein receptors on HepG2 cells. Hepatology 17, 898–907. 9. Wade, D.P., Knight, B.L. & Soutar, A.K. (1989) Regulation of low-density-lipoprotein-receptor mRNA by insulin in human hepatoma HepG2 cells. Eur. J. Biochem. 181, 727–731. 10. Demacker,P.N.,Staels,B.,Stalenhoef,A.F.&Auwerx,J.(1991) Increased removal of beta-very low density lipoproteins after ethinyl estradiol is associated with increased mRNA levels for hepatic lipase, lipoprotein lipase, and the low density lipoprotein receptor in Watanabe heritable hyperlipidemic rabbits. Arterio- scler. Thromb. 11, 1652–1659. 11. Auwerx, J.H., Chait, A. & Deeb, S.S. (1989) Regulation of the low density lipoprotein receptor and hydroxymethylglutaryl coenzyme A reductase genes by protein kinase C and a putative negative regulatory protein. Proc. Natl Acad. Sci. USA 86, 1133– 1137. 12. Auwerx, J.H., Chait, A., Wolfbauer, G. & Deeb, S.S. (1989) Involvement of second messengers in regulation of the low-density lipoprotein receptor gene. Mol. Cell Biol. 9, 2298–2302. 13. Casciola, L.A., van der Westhuyzen, D.R., Gevers, W. & Coetzee, G.A. (1988) Low density lipoprotein receptor degradation is influenced by a mediator protein(s) with a rapid turnover rate, but is unaffected by receptor up- or down-regulation. J. Lipid Res. 29, 1481–1489. 14. Casciola, L.A., Grant, K.I., Gevers, W., Coetzee, G.A. & van der Westhuyzen, D.R. (1989) Low-density-lipoprotein receptors in human fibroblasts are not degraded in lysosomes. Biochem. J. 262, 681–683. 15. Ness, G.C., Zhao, Z. & Lopez, D. (1996) Inhibitors of cholesterol biosynthesis increase hepatic low-density lipoprotein receptor protein degradation. Arch. Biochem. Biophys. 325, 242–248. 16. Kraemer, F.B., Sather, S.A., Park, B., Sztalryd, C., Natu, V., May, K., Nishimura, H., Simpson, I., Cooper, A.D. & Cushman, S.W. (1994) Low density lipoprotein receptors in rat adipose cells: subcellular localization and regulation by insulin. J. Lipid Res. 35, 1760–1772. 17. Ehlers, M.R. & Riordan, J.F. (1991) Membrane proteins with soluble counterparts: role of proteolysis in the release of trans- membrane proteins. Biochemistry 30, 10065–10074. 18. Kahn, J., Ingraham, R.H., Shirley, F., Migaki, G.I. & Kishimoto, T.K. (1994) Membrane proximal cleavage of 1-selectin: identifi- cation of the cleavage site and a 6-kD transmembrane peptide fragment of 1-selectin. J. Cell Biol. 125, 461–470. 19. Massague, J. & Pandiella, A. (1993) Membrane-anchored growth factors. Annu. Rev. Biochem. 62, 515–541. 20. Mullberg, J., Schooltink, H., Stoyan, T., Gunther, M., Graeve, L., Buse, G., Mackiewicz, A., Heinrich, P.C. & Rose-John, S. (1993) The soluble interleukin-6 receptor is generated by shedding. Eur. J. Immunol. 23, 473–480. 21. Quinn, K.A., Grimsley, P.G., Dai, Y.P., Tapner, M., Chesterman, C.N. & Owensby, D.A. (1997) Soluble low density lipoprotein receptor-related protein (LRP) circulates in human plasma. J. Biol. Chem. 272, 23946–23951. 22. Marlovitis, T.C., Abrahamsberg, C. & Blaas, D. (1998) Very-low- density lipoprotein receptor fragment shed from HeLa cells inhibits human rhinovirus infection. J. Virol. 72, 10246–10250. 23.Pandiella,A.,Bosenberg,M.W.,Huang,E.J.,Besmer,P.& Massague, J. (1992) Cleavage of membrane-anchored growth factors involves distinct protease activities regulated through common mechanisms. J. Biol. Chem. 267, 24028–24033. 24. van der Westhuyzen, D.R., Stein, M.L., Henderson, H.E., Marais, A.D., Fourie, A.M. & Coetzee, G.A. (1991) Deletion of two growth-factor repeats from the low-density-lipoprotein receptor accelerates its degradation. Biochem. J. 277, 677–682. 25. Chen, W.J., Goldstein, J.L. & Brown, M.S. (1990) NPXY, a sequence often found in cytoplasmic tails, is required for coated pit-mediated internalization of the low density lipoprotein receptor. J. Biol. Chem. 265, 3116–3123. 26. Kunkel, T.A., Roberts, J.D. & Zakour, R.A. (1987) Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods Enzymol. 154, 367–382. 27. Davis, C.G., Lehrman, M.A., Russell, D.W., Anderson, R.G., Brown, M.S. & Goldstein, J.L. (1986) The J.D. mutation in familial hypercholesterolemia: amino acid substitution in cyto- plasmic domain impedes internalization of LDL receptors. Cell 45, 15–24. 28. Tolleshaug, H., Goldstein, J.L., Schneider, W.J. & Brown, M.S. (1982) Posttranslational processing of the LDL receptor and its genetic disruption in familial hypercholesterolemia. Cell 30, 715–724. 29. Kamps, J.A. & Van Berkel, T.J. (1993) Regulation of low-density- lipoprotein receptors in the human hepatoma cell line Hep G2. Effect of phorbol 12-myristate 13-acetate and low-density lipo- protein. Eur. J. Biochem. 213, 989–994. 30. Kapoor, G.S., Golden, C., Atkins, B. & Mehta, K.D. (2003) pp90RSK- and protein kinase C-dependent pathway regulates p42/44MAPK-induced LDL receptor transcription in HepG2 cells. J. Lipid Res. 44, 584–593. 31. Lehrman, M.A., Goldstein, J.L., Brown, M.S., Russell, D.W. & Schneider, W.J. (1985) Internalization-defective LDL receptors produced by genes with nonsense and frameshift mutations that truncate the cytoplasmic domain. Cell 41, 735–743. 32. van Driel, I.R., Davis, C.G., Goldstein, J.L. & Brown, M.S. (1987) Self association of the low density lipoprotein receptor mediated by the cytoplasmic domain. J. Biol. Chem. 262, 16127–16134. 33. Kishimoto, A., Brown, M.S., Slaughter, C.A. & Goldstein, J.L. (1987) Phosphorylation of serine 833 in cytoplasmic domain of low density lipoprotein receptor by a high molecular weight enzyme resembling casein kinase II. J. Biol. Chem. 262, 1344–1351. 34. Allinson, T.M., Parkin, E.T., Turner, A.J. & Hooper, N.M. (2003) ADAMsfamilymembersasamyloidprecursorproteinalpha- secretases. J. Neurosci. Res. 74, 342–352. 35. Ramchandran, R. & Sen, I. (1995) Cleavage processing of angiotensin-converting enzyme by a membrane-associated metalloprotease. Biochemistry 34, 12645–12652. 36. Gearing, A.J., Beckett, P., Christodoulou, M., Churchill, M., Clements, J., Davidson, A.H., Drummond, A.H., Galloway, W.A., Gilbert, R. & Gordon, J.L. (1994) Processing of tumour 532 M. J. Begg et al. (Eur. J. Biochem. 271) Ó FEBS 2004 necrosis factor-alpha precursor by metalloproteinases. Nature 370, 555–557. 37. Hooper, N.M., Karran, E.H. & Turner, A.J. (1997) Membrane protein secretases. Biochem. J. 321, 265–279. 38. Rouis, M., Goldstein, S., Thomopoulos, P., Berthelier, M., Hervy, C. & Testa, U. (1984) Phorbol esters inhibit the binding of Low density lipoproteins (LDL) to U-937 monocytelike cells. J. Cell Physiol. 121, 540–546. 39. Kozarsky, K., Kingsley, D. & Krieger, M. (1988) Use of a mutant cell line to study the kinetics and function of O-linked glycosyla- tion of low density lipoprotein receptors. Proc. Natl Acad. Sci. USA 85, 4335–4339. 40. Fischer,D.G.,Tal,N.,Novick,D.,Barak,S.&Rubinstein,M. (1993) An antiviral soluble form of the LDL receptor induced by interferon. Science 262, 250–253. 41. Klein, B. & Brailly, H. (1995) Cytokine-binding proteins: stimu- lating antagonists. Immunol. Today 16, 216–220. 42. Vollmer,P.,Peters,M.,Ehlers,M.,Yagame,H.,Matsuba,T., Kondo, M., Yasukawa, K., Buschenfelde, K.H. & Rose-John, S. (1996) Yeast expression of the cytokine receptor domain of the soluble interleukin-6 receptor. J. Immunol. Methods 199, 47–54. 43. Orlando, R.A. & Farquhar, M.G. (1993) Identification of a cell line that expresses a cell surface and a soluble form of the gp330/ receptor-associated protein (RAP) Heymann nephritis antigenic complex. Proc. Natl Acad. Sci. USA 90, 4082–4086. 44. Fallon, R.J. & Schwartz, A.L. (1986) Regulation by phorbol Esters of Asialoglycoprotein and transferrin receptor distribution and ligand affinity in a hepatoma cell line. J. Biol. Chem. 261, 15081–15089. 45. He, G., Gupta, S., Yi, M., Michaely, P., Hobbs, H.H. & Cohen, J.C. (2002) ARH is a modular adaptor protein that interacts with the LDL receptor, clathrin, and AP-2. J. Biol. Chem. 277, 44044– 44049. 46.Trommsdorff,M.,Borg,J.P.,Margolis,B.&Herz,J.(1998) Interaction of cytosolic adaptor proteins with neuronal apolipo- protein E receptors and the amyloid precursor protein. J. Biol. Chem. 273, 33556–33560. 47. Gotthardt, M., Trommsdorff, M., Nevitt, M.F., Shelton, J., Richardson, J.A., Stockinger, W., Nimpf, J. & Herz, J. (2000) Interactions of the low density lipoprotein receptor gene family with cytosolic adaptor and scaffold proteins suggest diverse bio- logical functions in cellular communication and signal transduc- tion. J. Biol. Chem. 275, 25616–25624. 48. Moss, M.L., Jin, S.L., Milla, M.E., Burkhart, W., Carter, H.L., Chen, W.J., Clay, W.C., Didsbury, J.R., Hassler, D., Hoffman, C.R.,Kost,T.A.,Lambert,M.H.,Leesnitzer,M.A.,McCauley, P., McGeehan, G., Mitchell, J., Moyer, M., Pahel, G., Rocque, W., Overton, L.K., Schoenen, F., Seaton, T., Su, J.L., Warner, J. & Becherer, J.D. (1997) Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-alpha. Nature 385, 733–736. 49. Peschon, J.J., Slack, J.L., Reddy, P., Stocking, K.L., Sunnarborg, S.W., Lee, D.C., Russell, W.E., Castner, B.J., Johnson, R.S., Fitzner, J.N., Boyce, R.W., Nelson, N., Kozlosky, C.J., Wolfson, M.F., Rauch, C.T., Cerretti, D.P., Paxton, R.J., March, C.J. & Black, R.A. (1998) An essential role for ectodomain shedding in mammalian development. Science 282, 1281–1284. 50. Moss, M.L., Jin, S.L., Becherer, J.D., Bickett, D.M., Burkhart, W., Chen, W.J., Hassler, D., Leesnitzer, M.T., McGeehan, G., Milla, M., Moyer, M., Rocque, W., Seaton, T., Schoenen, F., Warner, J. & Willard, D. (1997) Structural features and bio- chemical properties of TNF-alpha converting enzyme (TACE). J. Neuroimmunol. 72, 127–129. 51. Bennett, T.A., Lynam, E.B., Sklar, L.A. & Rogelj, S. (1996) Hydroxamate-based metalloprotease inhibitor blocks shedding of 1-selectin adhesion molecule from leukocytes: functional con- sequences for neutrophil aggregation. J. Immunol. 156, 3093–3097. 52. Mullberg, J., Durie, F.H., Otten-Evans, C., Alderson, M.R., Rose-John,S.,Cosman,D.,Black,R.A.&Mohler,K.M.(1995) A metalloprotease inhibitor blocks shedding of the IL-6 receptor and the p60 TNF receptor. J. Immunol. 155, 5198–5205. 53. Arribas, J., Coodly, L., Vollmer, P., Kishimoto, T.K., Rose-John, S. & Massague, J. (1996) Diverse cell surface protein ectodomains are shed by a system sensitive to metalloprotease inhibitors. J. Biol. Chem. 271, 11376–11382. 54. Parvathy, S., Oppong, S.Y., Karran, E.H., Buckle, D.R., Turner, A.J. & Hooper, N.M. (1997) Angiotensin-converting enzyme secretase is inhibited by zinc metalloprotease inhibitors and requires its substrate to be inserted in a lipid bilayer. Biochem. J. 327, 37–43. 55. Killar, L., White, J., Black, R. & Peschon, J. (1999) Adamalysins. A family of metzincins including TNF-alpha converting enzyme (TACE). Ann. N. Y. Acad. Sci. 878, 442–452. 56. Moss, M.L. & Lambert, M.H. (2002) Shedding of membrane proteins by ADAM family proteases. Essays Biochem. 38, 141– 153. 57. Guo, L., Eisenman, J.R., Mahimkar, R.M., Peschon, J.J., Paxton, R.J.,Black,R.A.&Johnson,R.S.(2002)Aproteomicapproach for the identification of cell surface proteins shed by metallopro- teases. Mol. Cell. Proteomics 1, 30–36. 58. Zhang, Z., Oliver, P., Lancaster, J.R., Schwarzenberger, P.O., Joshi, M.S., Cork, J. & Kolls, J.K. (2001) Reactive oxygen species mediate tumor necrosis factor alpha-converting, enzyme-depen- dent ectodomain shedding induced by phorbol myristate acetate. FASEB J. 15, 303–305. 59. Zhang, Z., Kolls, J.K., Oliver, P., Good, D., Schwarzenberger, P.O., Joshi, M.S., Ponthier, J.L. & Lancaster, J.R. (2000) Acti- vation of tumor necrosis factor-alpha-converting enzyme-medi- ated ectodomain shedding by nitric oxide. J. Biol. Chem. 275, 15839–15844. 60. Tam,S.P.,Brissette,L.,Ramharack,R.&Deeley,R.G.(1991) Differences between the regulation of 3-hydroxy-3-methylglutaryl- coenzyme A reductase and low density lipoprotein receptor in human hepatoma cells and fibroblasts reside primarily at the translational and post-translational levels. J. Biol. Chem. 266, 16764–16773. 61. Balasubramaniam, S., Szanto, A. & Roach, P.D. (1994) Circadian rhythm in hepatic low-density-lipoprotein (LDL)-receptor expression and plasma LDL levels. Biochem. J. 298, 39–43. Ó FEBS 2004 Cell surface cleavage of LDL-R (Eur. J. Biochem. 271) 533 . compared to untreated cells, as was the total number of cell surface LDL-receptors assayed by the binding of 125 I-labeled anti- body to whole cells. To. adaptor proteins containing phospho-tyrosine binding domains are known to interact with the NPXY motif. ARH and Disabled 1 are two cytosolic proteins that