Eur. J. Biochem. 189,609-616 (1990) @) FEBS 1990

Regulation of mannose 6-phosphate/insulin-like growth factor I1 receptor distribution by activators and inhibitors of protein kinase C Thomas BRAULKE, Stefanic TIPPMER, Huey-Jiun CHAO and Kurt von FIGURA Georg-August-Universitit Giittingen, Biochemie 11, Federal Republic of Germany (Received August 7/Novembcr 21, 1989)

-

BJB 89 0990

The tumor-promotor phorbol dibutyrate (PDBt) increases the binding of a neoglycoprotein containing mannose 6-phosphate (Man6P) and of insulin-like growth factor I1 (IGF-11) to the Man6PIIGF-I1 receptor at the cell surface. This effect is dependent on time and concentration and is also seen with synthetic I-oleoyl-2-acetyl-snglycerol, but not with 4a-phorbol, an inactive tumor-promoter. The increase is due to a 3 - 4-fold increase in the number of cell-surface, receptors, accompanied by a 1.6-fold increase in ligand-binding affinity. The internalization rate of the Man6PIIGF-I1 receptor is not affected by PDBt, suggesting that the redistribution of these receptors to the cell surface is due to an accelerated externalization rate. The redistribution of Man6PIIGF-I1 receptors did not impair the sorting of newly synthesized Man6P-containing ligands while uptake of these ligands is 2 - 4-fold increased. Inactivation or down regulation of protein kinase C decreased the binding of the Man6P-containing neoglycoprotein to 65% of controls. Incubation of cells with Man6P, IGF-I, IGF-I1 or epidermal growth factor induces a rapid redistribution of Man6P/IGF-II receptors to the plasma membrane [Braulke, T., Tippmer, S., Neher, E. & von Figura, K. (1989) EMBO J. 8,681 -6861. Incubation with PDBt prevented the effect of growth factors but not that of Man6P on receptor redistribution. Inactivation of protein kinase C did not affect the Man6PIIGF-I1 receptor redistribution induced by Man6P and growth factors. These data suggest that Man6P, growth factors and activation of protein kinase C by phorbol esters and diacylglycerols modulate Man6PIIGF-I1 receptor cell-surface binding by at least two independent mechanisms, receptor redistribution as well as an increase of binding affinity, which might be involved in regulation of endocytosis of ligands.

of the Man6PIIGF-I1 receptor have not been identified yet. However, Man6P and IGF-I1 induce receptor redistribution via different transducing pathways, one of which is controlled by GTP-binding proteins [7]. In rat adipocytes insulin is known to induce redistribution of Man6PIIGF-I1 receptor to the plasma membrane [8, 91. This redistribution is accompanied by a decrease in the phosphoserine and phosphothreonine content of the receptors in clathrin-coated pits [lo]. A potential role of protein kinase C in the regulation of cell-surface receptors has been postulated on the basis of experiments with phorbol esters, which activate protein kinase C. Among the short-term effects of phorbol esters, phosphorylation of several receptors, decrease in affinity and altered distribution were observed (for review see [ll]). For example, phorbol ester treatment causes an increase in phosphorylation of EGF receptor with a concomitant decrease in both highaffinity binding and receptor tyrosine kinase activity [12, 131. Correspondence to T. Braulke, Georg-August-Universitat, Abtei- In addition to these effects, phorbol ester may induce a rapid lung Biochemie 11, Gosslerstrasse 12d, D-3400 Gottingen, Federal redistribution of receptors within the cell. While phorbol ester Republic of Germany treatment results mostly in redistribution of receptor to interAhhreviulions. Man6P, mannose 6-phosphate; IGF, insulin-like nal membranes as described for receptors of insulin [14], tumor growth factor; PDBt, phorbol 12J3-dibutyrate; PMA, phorbol 12- necrosis factor [I 51, asialoglycoprotein and transferrin [16], myristate 13-acetate; OleAcGro, 1-oleoyl-2-acetyl-sn-glycerol; H7, recently the increase of adenosine A2 receptor density [I71 in 1-(5-isoquinolinylsulfonyl-2-methylpiperazine) ; (Man6P-Man4-BSA, cell membranes in response to phorbol ester was reported. pentamannosyl-6-0-phosphate-substitutedbovine serum albumin; In the present study, we report on the effect of activators EGF, epidermal growth factor; MEM, minimal essential medium; and inhibitors of protein kinase C on the distribution and BHK cells, baby hamster kidney cells. Enzymes. Arylsulfatase A (EC 3.1.6.1); protein kinase C (EC function of Man6PIIGF-I1 receptors in human skin fibroblasts. Tumor-promoting phorbol esters and diacylglycerol 2.7.1.37).

The cation-independent receptor for mannose 6-phosphate (Man6P) is identical to the receptor for insulin-like growth factor I1 (IGF-11) [I -41. The Man6PIIGF-I1 receptor is required for targeting newly synthesized lysosomal enzymes to lysosomes and the endocytosis of proteins bearing Man6P residues (for review see [5]). The physiological role of IGF-I1 binding to this receptor is unknown, since many of the biological effects of IGF-I1 seem to be mediated through the IGF-I receptor (for review see [6]). In human fibroblasts Man6P and IGF-11, the two classes of ligands for the Man6PIIGF-I1 receptor, as well as other growth factors like IGF-I or epidermal growth factor (EGF), induce a rapid redistribution of Man6PIIGF-I1 receptors from intracellular membranes to the plasma membrane [7]. The signal transduction mechanisms, through which ManbP, IGF-I1 and other growth factors induce the translocation

610 induce a rapid redistribution of Man6PIIGF-I I receptors from intracellular pools to the cell surface and stimulate the endocytosis of Man6P-containing ligands. Down regulation or inhibition of protein kinase C is associated with a decreased expression of Man6PIIGF-I1 receptors at the cell surface. Furthermore, phorbol ester treatment prevents an additional receptor redistribution by IGF-I, IGF-11, and EGF, while the Man6P-induced redistribution of the receptor is not affected. This lends further support to the idea that Man6P and IGFI1 induce the redistribution of the receptor via different signal transduction pathways. The Man6P-triggered pathway is sensitive to cholera toxin and pertussis toxin [7] and independent of phorbol esters, while the IGF-11-triggered pathway merges with that stimulated by tumor-promoting phorbol esters. EXPERIMENTAL PROCEDURES Materials

The phorbol esters, phorbol 12,13-dibutyrate (PDBt) and phorbol 12-myristate 13-acetate (PMA) and 4a-phorbol (Sigma) were dissolved in ethanol at 2 mg/ml. Man6P-Man4 conjugated to bovine serum albumin [Man6P-Man4-BSA] was prepared and iodinated by means of iodogen (Pierce) as described [18] to a specific activity of 200-280 pCi/pg. The phosphomannan from Hansenula holstii used for preparation of the Man6P-Man4 was kindly provided by Dr M. Slodki (United States Department of Agriculture, Northern Regional Research Center, Peoria IL). [35S]Methionine(1100 Ci/mmol) and '"I were from Amersham and l4C-1abelled molecular mass standards were from New England Nuclear. 1-0leoyl2-acetyl-sn-glycerol (OleAcGro; dissolved in phosphatebuffered salt solution pH 7.4 containing 0.1 'YObovine serum albumin by sonication), Man6P (sodium salt) and 1-(5-isoquinolinylsulfonyl-2-methylpiperazine)(H7; dissolved in water) were obtained from Sigma; recombinant insulin-like growth factors I and I1 were a kind gift by Dr Marki (Ciba-Geigy, Basel). IGF-I1 was iodinated by the chloramine T procedure [19] to a specific activity of 45 -85 pCi/pg. Epidermal growth factor was from Boehringer Mannheim. Arylsulfatase A antiserum was raised by immunizing a goat with purified human placental arylsulfatase A [20]. Formaldehyde-fixed Staphylococcus aureu.y cell wall preparation (Immunoprecipitin) was from Bethesda Research Laboratories. Cell cultures

Human skin fibroblasts were maintained on 35-mm plates in minimal essential medium (MEM) containing 7.5% fetal calf serum plus penicillin and streptomycin.

'' I-Man6P-Man4-BSA binding and internalization studies Binding of '251-Man6P-Man4-BSA to human fibroblasts was determined as described previously [7]. Briefly, cells were washed and incubated for 60 rnin in MEM/albumin (minimal essential medium containing 20 mM Hepes pH 7.2 and 0.1 % bovine serum albumin). Unless otherwise mentioned, cells were incubated with phorbol esters for 30 min, OleAcGro for 15 min, H7 for 60 rnin and/or growth factors and Man6P for 10 rnin at the indicated concentrations at 37°C. After cooling to 4"C, the cells were then incubated twice for 10 rnin with icecold Hank's buffered salt solution containing 2 mM Man6P to release endogenous Man6 P-containing ligands. After washing with Hank's solution, the cells were incubated for 4 h at 4°C

with MEM containing 7.5% fetal calf serum, 20 mM Hepes pH 7.2 and 50 pM '251-Man6P-Man,-BSA is a high-affinity ligand for Man6P-specific receptors. In human skin fibroblasts, which express the Man6PIIGF-I1 receptor and the 46kDa Man6P receptor, all Man6P-binding sites at the cell surface are represented by the Man6PIIGF-I1 receptor [21]. Unbound ligand was removed by washing five times with Hank's solution. Cell-surface-bound 'I-Man6P-Man4-BSA was displaced by two incubations for 10 rnin each with H a n k s solution supplemented with 2 mM Man6P. The cells were solubilized in 1 M NaOH. Radioactivity was determined in the Man6P washes and related to cell protein [22]. The values were, with the exception of the evaluation of number and affinity of Man6PIIGF-I1 receptors, not corrected for unspecific binding. Unspecific binding is defined as binding in the presence of 5 mM Man6P, which accounts for 1015% of total binding in controls. For internalization studies the cells were incubated for 30 rnin in MEMIalbumin and for a further 30 rnin in the absence or presence of PDBt. Thereafter, the cells were incubated in the same medium containing 50 pM '"I-Man6PMan,-BSA (0.8 - 1.0 x lo6 cpm) in the absence and presence of growth factors (10 nM) or Man6P (5 mM) for 30 min. After removal of the medium, the cells were washed five times with ice-cold Hank's buffered salt solution and incubated twice for 10 min with ice-cold Hank's solution containing 2 mM Man6P to release the cell-surface-bound ligands. The cells were solubilized in 1 M NaOH. Aliquots of the medium were removed, proteins precipitated with trichloroacetic acid (0.1 g/ml) and the soluble radioactivity determined in the supernatant. Radioactivities released from the cell surface, in the solubilized cells and in the acid-soluble fraction of the medium were related to cell protein. Uptake is defined as the sum of radioactivity associated with solubilized cells and of the 251-labelleddegradation products released as acid-soluble material into the culture medium.

'

Cross-linking of 1251-IGF-IIto fibroblasts

Cross-linking of 1251-IGF-IIto cell-surface Man6PIIGFI1 receptors of human fibroblasts was carried out as described [23]. After incubation for 30 rnin with PDBt, the cells were chilled to 4°C and washed three times with a buffer containing 0.1 M Hepes, 0.12 M NaCI, 5 mM KCI, 1.2 mM MgSO,, 8 mM glucose, pH 7.6, and 0.1% bovine serum albumin (buffer A). The cells were then incubated for 16 h at 4°C with buffer A containing 5 ng '251-IGF-II. Subsequently the cells were washed three times with buffer A without albumin and then incubated for 15 rnin at 4°C with the same buffer containing 0.5 mM disuccinimidylsuberate (Pierce Chem. Co.). The cross-linking reaction was quenched by addition of an equal volume of 0.15 M Tris/HCI pH 7.4 containing 1.5 mM EDTA for 15 min. The '251-labelled proteins were analyzed by SDS/ polyacrylamide gel electrophoresis under reducing conditions and visualized by autoradiography. Internalization of 35S-labelledarylsuvatase A

Baby hamster kidney cells transfected with the cDNA of human arylsulfatase A (cell line BHK-HTCP 3) were provided by Dr C. Stein of this institute [24]. After incubation of the cells with [35S]methioninefor 16 h in the presence of 10 mM NH,Cl, the medium was precipitated by ammonium sulfate (0.5 g/ml) and dialyzed against MEM supplemented with 20 mM Hepes pH 7.2. About 0.3% of the total radioactivity

61 1 in the secretions is represented by arylsulfatase A and 3% by ligands for Man6P receptors. Fibroblasts were incubated for 30 rnin in MEM/albumin and for a further 30 rnin in the absence or presence of PDBt prior to addition of the radiolabelled secretions (1.5 -3.5 x lo5 cpm). After incubation for 30min, the cells were washed twice and arylsulfatase A was immunoprecipitated from cell lysates as described [25] and visualized after SDS-gel electrophoresis by fluorography. Man6PIIGF-II receptor affinity chromatography Fibroblasts were labelled with [35S]methionine (0.9 mCi/ ml) for 10 min at 37°C and then chased for 4 h in the absence or presence of PDBt (1 pg/ml). The chase medium also contained 5 mM Man6P to prevent re-uptake of Man6Pcontaining ligands. After the chase the secretory products were precipitated with ammonium sulfate (0.5 g/ml), centrifuged and dialyzed against 50 mM imidazole/HCl pH 7.0, containing 150 mM NaCI, 5 mM sodium glycerol 2-phosphate, 2 mM EDTA, 0.05% Triton X-100 and 0.02% sodium azide (buffer B). The cells were washed five times with icecold Hank's salt solution and solubilized in buffer B containing 1 mM phenylmethylsulfonyl fluoride, 5 mM iodoacetic acid, 0.05 TIU/ml aprotinin, 0.2 pM pepstatin, 0.3 pM leupeptin and 0.14 pg/ml chymostatin. After sonication, the cell extract was centrifuged for 60 rnin at 50000 xg. The supernatant and the dialyzed media were subjected to chromatography on a Man6PIIGF-I1 receptor column. To this end purified Man6PIIGF-I1 receptor from human liver [26] was coupled to Affigel-10 (Bio-Rad) at a concentration of 0.2 mg receptor/ml gel. The samples were loaded onto the receptor column (0.4 ml) equilibrated in buffer B, washed with 2.5 ml buffer B followed by 1.5 ml containing 5 mM glucose 6-phosphate and finally with 2 ml containing 5 mM Man6P. The fractions eluted with Man6P were lyophilized, dissolved in water and precipitated with acetone (final concentration 80%) for 16 h at -20°C. The precipitate was washed in 80% acetone and solubilized in the presence of 5 mM dithiothreitol and 1YOSDS, analyzed by SDS/polyacrylamide gel electrophoresis (IOYOpolyacrylamide), visualized by fluorography and quantified by densitometry. Protein kinase C assay

Concentration

( pglrnl)

Time ( m i n )

Fig. 1. Concentration and time dependence of PDBt on '251-Man6PMan4-BSA binding. (A) Fibroblasts were incubated for 30 rnin in a serum-free medium prior to the addition of various concentrations of PDBt for another 30min. After cooling the cells and removal of endogenous cell-surface ligands by a Man6P wash, the binding of lZ5IMan6P-Man4-BSA, measured as cell-surface "'I, was determined at 4°C. (B) Fibroblasts were incubated for 30 rnin in a serum-free medium and thereafter for different time periods with PDBt (1 pg/ml) prior to the Man6P-Man,-BSA binding assay Table 1 . Effect of activation (phorhol esters und diucylglycerol) and inhibition ( H 7 and long-term PDBt treatment) of protein kinase C on binding of '251-Mnn6P-Man4-BSA Fibroblasts were preincubated with phorbol esters for 30 rnin and with OleAcGro for 15 rnin at 37°C (A) or with H7 for 60 rnin and with PDBt for 16 h (B). Results are given as a percentage of controls f SD; controls bound 0.104 f 0.08 ng/mg cell protein (n = 15) where n = number of experiments ~~

Pretreatment

Concentration

n

Bound '''I

A

PDBt PMA OleAcGro 4cc-phorbol

1 PgW 10 nM 1 mM 1 Pg/ml

12 10 2 2

312+96 211 + 6 4 198 100

B

u7 PDBt

0.3 mM 1PgW

2 4

65 64k 9

Expt

%

shown). Treatment with the phorbol ester PMA (10 nM) increased the binding 2.1-fold (range 1.6 - 3.6-fold). 4a-Phorbol (1 pg/ml), which lacks tumor-promoting activity, had no effect on binding (Table 1A). If PDBt, PMA and 4a-phorbol were present only during incubation with '251-Man6P-Man4-BSA at 4"C, binding was 87%, 95% and 103% of controls, respectively. This indicates that PDBt, PMA, and 4a-phorbol do not affect the binding of Man6P-Man4-BSA to Man6PIIGF-I1 receptor. Phorbol esters and diacylglycerol interact at a comRESULTS mon binding site on protein kinase C and stimulate the activity Eflects of phorbol esters and H7 of the kinase in a similar manner [28]. If the effects of PDBt on l 2 I-Mnn6P-Man4-BSA binding and PMA on the Man6PIIGF-I1 receptor were mediated Incubation of human fibroblasts at 37 "C with the phorbol through a stimulation of protein kinase C, 1,2-diacyl-snester PDBt increased the binding of '251-Man6P-Man4-BSA glycerol should have a similar effect. Incubation of cells with in a concentration- and time-dependent manner (Fig. 1 A and 1 mM OleAcGro for 15 min did indeed enhance the 12'1B). The maximal effect was observed at 1 pg PDBt/ml and Man6P-Man4-BSA binding twofold (Table 1A). Incubation of fibroblasts for 1 h with H7, which inhibits 60-min incubation, respectively. In further experiments fibroblasts were treated with 1 pg/ml PDBt (2 pM) for 30 min. protein kinase C [29], decreased the binding of 1251-Man6PUnder these conditions PDBt increased the 1251-(Man6P)5- Man4-BSA to 65% of controls (Table IB). Incubation of albumin binding 3.1-fold (range 2.0- 5.4-fold). After removal fibroblasts for 16 h with 2 pM PDBt [30] reducing protein of PDBt the increased binding persisted for at least 2 h (not kinase C activity to less than 15% and decreased the binding

Protein kinase C activity was determined either in the 100000 x g soluble fraction or in Triton X-100 cell lysates from two 35-mm plates according to [27] using a 0.2-ml DEAEcellulose column which had been eluted with 0.3 ml column buffer that contained 150 mM KCI.

612 Table 2. Equilibrium binding constant (Kd) and binding sites (Bmax) for Man6P-Man4-BSA in,fibroblasts pretrated with PDRt Fibroblasts were incubated for 30 nlin at 37'.C with or without PDBt (1 pg/ml). Concentration-dependent binding of (Man6P)5-albumin was determined at 1 -200 pM. All values are the mean f SD or five separate determinations PDBt

-

+

Kd

Bm,

c-

200 -

PM

pg/mg cell protein

24 f 3 15k4

129 f 64 485 _+ 160

97 69 -

of '251-Man6P-Man,-BSA to 64% of controls (Table 1B). In cells pretreated for 1 h with H7 or 16 h with PDBt, addition of PMA and PDBt, respectively, did not stimulate binding of 1251-Man6P-Man4-BSA,indicating that protein kinase C mediates the increased binding of this substance by these phorbol esters. The increase in 251-Man6P-Man4-BSAbinding in fibroblasts treated with protein kinase C activators may be due to increased affinity and/or an increased number of Man6Pl IGF-11 receptors. The dissociation constant (&) and the number of binding sites for 1251-Man6P-Man4-BSAwere examined in controls and in PDBt-treated cells. Pretreatment of fibrobbdsts with PDBt increased the number of binding sites 3.8-fold and decreased the Kd value for Man6P-Man4-BSA from 24 p M in controls to 15 pM (Table 2). Addition of 0.5 m M cycloheximide which inhibited protein synthesis more than 95%, 30min prior to and during the treatment with PDBt did not affect the PDBt-induced increase in 1251Man6P-Man4-BSA binding. Thus, pretreatment of cells with PDBt increases strongly the number of Man6PIIGF-11 receptors at the cell surface, by redistribution of the pre-existing receptors, and increases moderately their affinity for Man6PMan,-BSA.

46

-

30 -

'

Stimulation of '251-IGF-IIbinding by PDBt Increased expression of the Man6PjIGF-I1 receptor should result in an increased binding of IGF-11. As most of the IGF-I1 binding sites on fibroblasts are represented by IGF-I1 binding protein unrelated to the Man6PIIGF-I1 receptor, we had to use chemical cross-linking of the bound 1251-IGF-II followed by SDS/polyacrylamide gel electrophoresis to identify binding to the receptor. Fig. 2 shows the cross-linking of 12sI-IGF-II to cell-surface Man6PIIGF-I1 receptors in fibroblasts. It may be noted that, under the conditions used, no binding to the 135-kDa subunit of the IGF-I receptor is detectable in fibroblasts. The binding of 1251-IGF-IIto the receptor and to a 38-kDa IGF-I1 binding protein is inhibited by an excess of unlabelled IGF-11. Treatment with PDBt increased significantly the binding of '"IIGF-I1 to the Man6PIIGF-I1 receptors. Densitometry of the autoradiographs of two independent experiments revealed a 5.8-fold higher amount of 1251-IGF-II cross-linked to the Man6PIIGF-I1 receptor in PDBt-treated cells. This indicates that the translocation of the Man6PIIGF-I1 receptor induced by phorbol esters results in increased binding of Man6Pcontaining ligands and IGF-11. Stimulation of endocytosis ofMan6P-containing ligands by PDBt Preincubation of the cells for 30 rnin with PDBt led to a 4-fold increase of Man6P-Man4-BSA uptake during a sub-

'k

IGFll

- +

-

Fig. 2. Autoradiograph from cross-linking rxp.uriment.swith 2'I-IGt;-II to human fibroblasts. Fibroblasts were incubated for 30 min in the absence and presence of PDBt (lpg/ml) before binding and crosslinking of "'I-IGF-I1 in the absence and presence of unlabelled IGF-II (0.5pg). The positions of molecular mass markers (values in kDa), the Man6PIIGF-I1 receptor (+) and the 38-kDa binding protein (*) are indicated Table 3 . Effect of PDBt on endocytosis of ManhP-containing ligands Fibroblasts were treated for 30 min in Hepes-buffered MEM containing 0.1% bovine serum albumin at 37°C and then incubated with either l 2'I-Man6P-Man4-BSAor [3 5S]methionine-labelledsecretions of BHK cells expressing human arylsulfatase A in the absence and presence of PDBt (1 pg/ml) and Man6P (5mM).The cells were solubilized and analyzed for internalized '251-Man6P-Man,-BSAand arylsulfatase A. For the former, results are given i SD ( n = 8). For the latter, fluorograms of two separate experiments (see Fig. 3) were quantified by densitometry.Nonspecific internalization was measured in the presence of 5 mM Man6P. Specific internalization for untreated controls was 0.191 & 0.009 ng 'Z5T-Man6P-Man,-BSA/rngcell protein in 30 rnin (n = 8) Treatment

Internalized ligand '251-Man6P-Man,-BSA

[35S]arylsulfataseA

YOcontrol None

PDBt Man6P

1OOk 14 403 f 139 15k 5

100

237

sequent incubation period of 30 rnin at 37°C in the presence of PDBt (Table 3). The presence of 5 m M Man6P inhibited internalization of 1251-Man6P-Man4-BSAby 85%. N o difference in the concentration required for half-saturation of uptake between control (32 pM) and PDBt-treated cells (25 pM) was observed (not shown). Similar results were obtained with arylsulfatase A. a natural ligand for the Man6PIIGF-I1 receptor. Radiolabelled secretions of BHK cells over-expressing human arylsulfatase A served as a source for arylsulfase A. Fibroblasts were incubated for 30 min at 37°C with the [35S]methionine-labelled

61 3 A

B

C

200 -

80 60

-

97 40 -

69

-

46

-

30

-

20 -

Fig. 3. Eflect ofphorbol ester treatment on endocytosis of 35S-lubelled arylsulfatase A . Fibroblasts were incubated for 30 min in the absence (A, C) or presence (B) of PDBt (1 bg/ml). Then the cells were incubated for 30 min with [35S]methionine-labelledsecretions of cells overexpressing human arylsulfatase A in the absence (A) or presence (B) of PDBt or 5 mM Man6P (C). Arylsulfatase A was immunoprecipitated from cell lysates as described in Materials and Methods. Values on the left are molecular masses (in kDa) of marker proteins 0

secretions in the absence and presence of PDBt. PDBt-treated cells internalized 2.4-fold more labelled arylsulfatase A. The uptake of arylsulfatase A was completely inhibited in the presence of 5 mM Man6P (Fig. 3 and Table 3). ManfiPIIGF-II receptor internalization is not affected by PDBt The similar increase of receptor expression at the cell surface and of ligand internalization suggests that PDBt increases the number of cell-surface receptors by increasing the rate of receptor externalization rather than by decreasing the rate of receptor internalization. Since the rate of receptor externalization cannot be assessed directly, we determined the rate of receptor internalization. Control cells and cells stimulated with PDBt were allowed to bind 1251-Man6P-Man4-BSAat 4°C. After extensive washing, cells were warmed to 37'C in the absence and presence of PDBt. Under both conditions the 1251-Man6P-Man4-BSAdisappeared from the cell surface upon warming with a t I i 2 of less than 1 min (not shown). Control and PDBt-treated cells internalized about 55% and 70% of the prebound '251-Man6P-Man4-BSA, respectively, within 2 min. Between 40-50% and 20-30% of the total radioactivity was released from control and PDBt-treated cells, respectively, into the medium as acid-insoluble material which is due either to dissociation or to rapid recycling of internalized receptor bound '251-Man6P-Man4-BSA. To avoid complications resulting from the rapid kinetics of internalization and to distinguish between dissociation at the cell surface and release after internalization, the cells with prebound '251-Man6P-Man,-BSA were warmed to 2 0 T , at which temperature it disappeared from the cell surface of controls with a t l l z of 13 min (Fig. 4A); in control cells about 55% of the prebound '251-Man6P-Man4-BSA was internalized with a t l j z of 12 min (Fig. 4B) and 30-35% released into the medium (Fig. 4C). The radioactivity released into the medium represented intact l 251-Man6P-Man,-BSA as examined by gel electrophoresis and trichloroacetic acid precipitation. Under the same conditions, in PDBt-treated cells the prebound 1251-(Man6P)5-albumindisappeared from the cell

15

30

45

60

time (rnin)

Fig. 4. Effect of PDBt on the internalization and dissociation at 20'C of 125Z-Man6P-Man4-BSA prebound to the cell surface. Fibroblasts were incubated for 30 min in the absence or presence of PDBt (1 pg/ ml) at 37'-C and then chilled to 4°C. Thc cells were then incubated for 4 h at 4°C with 1ZSI-Man6P-Man4-BSA;after removal of the unbound substance the cells were incubated at 20°C with MEM/ albumin in the absence ( 0 )or presence (0)of PDBt. At various time points the cells were chilled to 4°C and the distribution of 1251Man6P-Man4-BSA, measured as the radioactivity at the cell surface (A), in the cell extract (B) and in the medium (C), was determined. The values are the mean of duplicates of one experiment that is representative for three independent experiments

surface with an identical rate as controls but up to 70% of it was internalized with a t l i z of about 8.5 min. The relative amount of radioactivity released into the medium at 20 "C is smaller compared to that at 37°C but occurred with a similar initial rate in PDBt-treated cells and controls. In addition, the apparent binding constant of '251-Man6P-Man4-BSA at the cell surface at 37°C in controls and PDBt-treated cells was 3-6-fold higher than at 4°C (not shown) [181. We conclude from these results that release of prebound '251-Man6PMan4-BSA is due to dissociation at the cell surface. Since its rate of disappearance from the cell surface and dissociation into the medium is nearly identical in controls and PDBttreated cells, the higher relative amount internalized in the presence of PDBt and the decreased dissociation of ligand into the medium must results from an increased affinity of the Man6PIIGF-I1 receptor for Man6P-Man4-BSA. The sorting of MunhP-containing ligands is not affected by PDBt

Man6PIIGF-I1 receptors in the Golgiltrans-Golgi network are necessary for binding and segregation of newly synthesized Man6P-containing lysosomal enzyme precursors from the secretory pathway. To determine whether the PDBtinduced Man6PIIGF-I1 receptor redistribution to the plasma

614 Medium

Table 4. Effect of Mun6P und growth factors on '2sI-Man6P-Man4BSA binding tofibroblustspretreuted with PDBtfor 30 rnin or 16 h Pretreatment of PDBt for 30 rnin or 16 h resulted in 258% and 64% 1Z51-Man6P-Man,-BSA binding of untreated controls, respectively. All values are referred to the respective controls, f SD

Cells

Addition during chase

200 97

-

Addition

69 -

none ( n = 12)

e

46

-

e

30 -

Fig. 5. Man6P-containing ligunds in cells and secretions in the presence of PDBt. Fibroblasts were labelled for 10min at 37°C with [35SImethionine and chased for h in the absence and presence of PDBt (1 pg/ml). Additionally, the chase medium contained 5 mM Man6P to prevent reuptake of ManbP-containing ligands. The extracts of the media and of half of the cells were applied to a Man6t'/ IGF-I1 receptor affinity column. The fractions eluted with Man6P were separated by SDS/polyacrylamide gel electrophoresis, visualized by fluorography and quantified by densitometry. The Position of peptides with higher labelling intensity in cell extracts of control cells compared to PDBt-treated cells are indicated by asteriks. The positions of the two polypeptides which are pronounced in PDBttreated cells are indicated by the arrows

membrane had led to a functionally relevant deficit of receptors in the Golgiltrans-Golgi network, the distribution of Man6P-containing ligands between cells and media was studied. Cells were labelled with [3sS]methionine for 10 rnin and then chased for 4 h in the absence and presence of PDBt. Additionally, 5 mM Man6P was present in the medium to prevent a reuptake of Man6P-containing ligands appearing at the cell surface. The presence of Man6P alone during the chase had no effect on secretion of Man6P-containing ligands (not shown). Extracts of cells and media were applied to a Man6PIIGF-11 receptor affinity column and the specifically bound material was eluted with 5 mM Man6P and analyzed by SDS/polyacrylamide gel electrophoresis. In the presence of PDBt 7.8% of total [3sS1methionine-labelledMan6P-containing ligands were secreted during the chase period compared to 6.2% in control cells (Fig. 5). This indicates that intracellular retention of Man6P-containing ligands is not affected by PDBt treatment. The apparent molecular masses of intracellularly retained Man6P-containing ligands differed significantly in treated cells and controls. We ascribe this tentatively to an effect of PDBt on the proteolytic processing of Man6Pcontaining ligands. It is characteristic of many lysosomal proteins that they undergo significant size changes due to limited proteolysis in prelysosomal organelles and in lysosomes [31]. PDBt inhibits the growth~factor-inducedMan6PIIGF-II receptor redistribution Treatment of fibroblasts for 10 rnin at 37°C with Man6P o r growth factors such as IGF-I, IGF-I1 or EGF causes a

rapid and transient redistribution of Man6PIIGF-I1 receptors to the plasma membrane [7]. We examined, therefore, whether

'"1-binding after pretreatment PDBt (30 min) (n = 5)

PDBt (16 h) (n = 3)

100 f 19 131f23 131 & 34 108 f 16 119 f 29

100 f 14 132f 6 314 40 232 f 40 172 t 40

"A control

None Man6P IGF-I IGF-I1 EGF

1OOf 8 137 f 19 295 f 99 175 52 190 & 40

the redistribution of Man6PIIGF-I1 receptors induced by phorbol esters was additive or non-additive to that induced by Man6P and growth factors. Results presented in Table 4 demonstrate that pretreatment Of Cells with PDBt prevented a further increase in binding of lZ51-Man6P-Man4-BSAby growth factors. In contrasts, Man6P induced a significant increase ( P = 0.02, Wilcoxon test) in this binding which was not affected by PDBt. Similar results were obtained in cells pretreated with PMA (not shown). These results indicate that Man6P and phorbol ester induce the receptor redistribution through independent signal transduction pathways, while the signal transduction pathways stimulated by growth factors are either inhibited by phorbol ester or merge - with that stimulated by phorbol esters. To examine whether the growth-factor-induced redistribution of Man6PIIGF-I1 receptors depends on protein kinase C, fibroblasts were incubated for 16 h with PDBt to downregulate protein kinase C. Depletion of protein kinase C activity resulted in a lower basal binding of 1251-Man6P-Man,BSA (see Table 1 B). Stimulation of these cells for 10 rnin with Man6P or growth factors, however, led to the same relative increase in binding as in controls (Table 4). Treatment with the H7- inhibitor also did not affect the increase of "'1Man6P-Man4-BSA binding stimulated by Man6P and growth factor (not shown). These results indicate that protein kinase C is not required for the induction of redistribution of Man6P/ IGF-I1 receptors induced by Man6P and growth factors.

DISCUSSION The Man6PlIGF-I1 receptors at the cell surface are in equilibrium with intracellular receptors residing in the Golgi/ trans-Golgi network, endosomes and other non-identified membranes [32, 331. The exchange between cell-surface and intracellular receptors is rapid and replaces the entire cellsurface pool of receptors every 3 rnin [IS]. The exchange does not depend on the binding Man6P-containing ligands [18] or IGF-I1 1341. The two classes of receDtor ligands. however. induce in fibroblasts a rapid redistribdtion orMan6P/IGF-II receptors to the cell surface. They induce the redistribution through binding to different sites of the same receptor and use different signal transduction pathways. The signal transduction pathway stimulated by Man6P is sensitive to cholera toxin and pertussis toxin. The signal transduction pathway stimulated by IGF-I1 is resistant to both toxins and shares

61 5 some elements with the signal transduction pathways by which IGF-I and EGF induce redistribution of the Man6PIIGF-I1 receptors [7]. In the present study we show that tumor-promoting phorbol esters and diacylglycerol, which activate protein kinase C, induce a similar redistribution of Man6PIIGF-I1 receptors to the surface of fibroblasts as do Man6P, IGF-I, IGF-I1 and EGF. Since inhibition or down-regulation of protein kinase C decreases the receptor expression at the cell surface, protein kinase C may be involved in the mechanisms that regulate the traffic and pool sizes of the receptor. In addition to the redistribution of receptors to the cell surface, which resulted in increased binding of Man6P-containing ligands, IGF-I1 and receptor antibodies (not shown), PDBt increased moderately the apparent affinity for Man6P-Man4BSA, a ligand of the Man6P-binding site. The increase in the apparent affinity may be real or apparent. Improved binding of the polyvalent ligand Man6P-Man4-BSA (which contains more that 30 Man6P groupslpolypeptide molecule) to an increased number of cell-surface receptors could mimick an increase of affinity. Since binding of the presumably monovalent ligand IGF-I1 is about 6-fold increased, while the number of receptors is only about 4-fold increased, phorbol esters may increase the affinity for IGF-11. The presence of IGF-I1 binding proteins prevented the determination of the affinity constant of the Man6PIIGF-11 receptor for IGF-11. The kinetics of receptor traffic are fundamental for the receptor function. They determine the pool sizes of the receptors at the sites of ligand binding and determine the rates of ligand transport. PDBt did not affect the rate of internalization of cell-surface-bound ligands. This implies that the increased expression of Man6P/IGF-I1 receptors at the cell surface results from a stimulation of the externalization rate of receptors. This may be accomplished by an acceleration of externalization rate and/or by feeding receptors that normally cycle between endosomes and the Golgi into the pool that recycles to the cell surface. The mechanism by which phorbol esters induce the externalization rate of the Man6PIIGF-I1 receptor is not clear. Preliminary results show that treatment of 32P-labelled fibroblasts with PDBt increases selectively the phosphorylation of cell-surface Man6PIIGF-I1 receptors (unpublished results). It is therefore conceivable that changes in phosphorylation of the receptor induced by protein kinase C contributes to the receptor redistribution at the cell surface. An important function of the Man6PIIGF-I1 receptor is the binding of newly synthesized lysosomal enzymes in the Golgi/truns-Golgi network and their transport to a prelysosomal compartment. If lysosomal enzymes fail to bind to the receptor, they become secreted [5, 33, 351. The increase of cell-surface Man6PIIGF-I1 receptors from the normal 10 20% of total receptors to up to 50% in phorbol-ester-treated cells may result in a decrease of receptors in the GolgiltransGolgi network below the level that is sufficient for targeting of newly synthesized Man6P-containing ligands. The present data show that the phorbol-ester-induced redistribution of Man6PIIGF-I1 receptors does not impair the intracellular retention of newly synthesized Man6P-containing ligands. The altered distribution pattern of intracellular Man6P-containing polypeptides in the phorbol-ester-treated cells may result from proteolytic processing which is typical of many lysosomal proteins [31]. Phorbol esters and growth factors stimulate the cell-surface expression of Man6PIIGF-I1 receptors in a non-additive manner. An explanation could be that either signal-transduction pathways stimulated by growth factors and phorbol esters

fuse and act through a common mechanism or that phorbol esters inhibit the growth-factor-triggered signal transduction pathway. The ability of growth factors (and Man6P) to stimulate the cell-surface expression of Man6PIIGF-I1 in cells, in which protein kinase C is down-regulated or inhibited, indicates that the signal transduction pathway mediating the growth-factor-induced receptor redistribution does not involve protein kinase C. The additive effects of phorbol esters and Man6P on the receptor redistribution suggest that the signal transduction pathways triggered by phorbol esters and Man6 P are different and independent. The different effects of phorbol esters on the Man6P- and growth-factor-induced receptor redistribution support the earlier hypothesis that the signal transduction pathways initiated by binding of Man6P, on the one hand, and IGF-11, on the other hand, are different ~71. Little is known about signal transduction pathways coupled to Man6PIIGF-I1 receptors. Recently it has been observed that IGF-I1 stimulates the formation of inositol trisphosphate and diacylglycerol in basolateral membranes of renal proximal tubular cells [36] and that Man6P enhances the effect of IGF-I1 [37]. In competent BALB/c3T3 cells primed with EGF, IGF-11 Stimulates the calcium influx via a mechanism sensitive to pertussis toxin [38]. Furthermore, the coupling of Man6P/IGF-I1 receptor to Gi2 proteins in phospholipid vesicles could be demonstrated [39]. In our own experiments we could not detect changes in concentration of intracellular calcium [7]or inositol trisphosphate (unpublished results) after stimulation of human fibroblasts with both Man6P and IGF-11. This indicates that the binding of ligands to the Man6P/IGF-I1 receptor in fibroblasts is not coupled with the formation of inositol trisphosphate and a calcium channel. This study was supported by the Deutsche Forschungsgemeinschuft (SFB 236) and the Fonds der Chemischen Industrie. We thank Prof. Dr H.-D. Soling for suggestions and critical reading of the manuscript.

REFERENCES I . Morgan, D. O., Edmanm, J . C., Standring, D. N., Fried, V. A,, Smith, M. C., Roth, R. A. & Rutter, W. J. (1987) Nuture 329, 301 -307. 2. Roth, R. A,, Stover, C., Hari, J., Morgen, D. O., Smith, M. C., Sara, V. &Fried, V. A. (1987) Biochem. Biophys. Res. Commun. 149, 600-606. 3. Braulke, T., Causin, C., Waheed, PI., Junghans, U., Hasilik, A., Maly, P., Humbel, R. E. & von Figura, K. (1988) Biochem. Biophys. Res. Commun. 150, 1287- 1293. 4. Kiess, W., Blickenstaff, G. D., Sklar, M. M., Thomas, C. L., Nissley, S. P. & Sahagian, G. (1988) J . Biol. Chem. 263, 93399344. 5. von Figura, K. & Hasilik, A. (1986) Annu. Rev. Biochem. 55, 167-193. 6. Roth, R. A. (1988) Science 239, 1269-1271. 7. Braulke, T., Tippmcr, S., Neher, E. & von Figura, K . (1989) E M B O J . 8, 681 -686. 8. Oppenheimer, C. L., Pessin, J. E., Massague, J., Gitomer, W. & Czech, M. P. (1983) J . B i d . Chem. 258, 4824-4830. 9. Wdrdzala, L. J., Simpson, I. A,, Rechler, M. M. & Cushman, S. W. (1984) J . B i d . Chem. 259, 8378-8383. 10. Corvera, S., Folander, K., Clairmont, K. B. &Czech, M. P. (1988) Proc. Nut1 Acad Sci. U S A 85, 7567 - 7571. 11. Nishizuka, Y. (1986) Science 233, 305-312. 12. Whitcty, B. & Glaser, L. (1986) J . Cell Bid. 103, 1355-1362. 13. Davis, R. J. &Czech, M. P. (1986) Biochem. J . 233, 435-441.

616 14. Hachiya, H. L., Takayama, S., White, M. F. &King. G. L. (1987) J . Biol. Chem. 262, 6417 - 6424. IS. Johnson, S. E. & Baglioni, C. (1988) J . B i d . Chem. 263, 56865692. 16. Fallon, R. J . & Schwartz, A . L. (1 986) J . B i d . Chern. 261, 15081 15089. 17. Murayama, I., Nomura, Y. & Ui, M. (1989) J . Biol. Chern. 264, 15 186- 15 191. 18. Braulke, T., Gartung, C., Hasilik, A. & von Figura, K. (1987) J . Cell Biol. 104, 1251- 1256. 19. Zapf, J., Walter, M. & Froesch, E. R . (1981) J . Clin. Invest. 68, 1321- 1330. 20. Waheed, A., Hasilik, A. & von Figura, K. (1982) Hoppe-Seyler’s Z. Physiol. Chem. 363,425-430. 21. Stein, M., Zijderhand-Bleekemolen, J. E., Geuze, H., Hasilik, A . & von Figura, K. (1987) E M B O 1.6,2677-2681. 22. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J . Biol. Chem. 193, 265-275. 23. Ballard, F. J . Ross, M., Upton, F. M. & Francis, G. L. (1988) Biochenz. J . 249, 721 - 726. 24. Stein, C., Gieselmann, V., Kreysing, J., Schmidt, B., Pohlmann, R., Waheed, A., Meyer, H. E., O’Brien, J. S. & von Figura, K . (1989) J . B i d . Chem. 264, 1252-1259. 25. Braulke, T., Hille, A,, Huttncr, W. B., Hasilik, A. & von Figura, K . (1987) Biochem. Biophys. Res. Commun. 143, 178-185. 26. von Figura, K., Gieselmann, V. & Hasilik, A. (1985) Biochem. J . 225, 543 - 547.

27. Niles, R. M. & Locwy, B. P. (1989) Cancer Res. 4Y. 44834487. 28. Castagna, M., Tdkai, Y., Kaibuchi, K., Sano, K., Kikkawa, U. & Nishizuka, Y. (1982) J . B i d . Chem. 257,7847-7851. 29. Kawamolo, S. & Hidaka, H . (1984) Biochem. Biophys. Res. Commun. 125, 258-264. 30. Ballester, R. & Rosen, 0. M. (1985) J . Biol. Chem. 260, 1519415199. 31. Skudlarek, M. D., Novak, E. & Swank, R . T. (1984) in Lysosonws in biology andpathology, vol. 7 (Dingle, J. T., Dean, R . T. & Sly, W. cds) pp. 17 - 34, Elsevier, Amsterdam. 32. von Figura, K., Gieselmann, V. & Hasilik, A. (1984) EMBO J . 3,1281 -1286. 33. Gartung C., Braulke, T., Hasilik, A. & von Figura, K . (1985) E M B O .J. 4, 1725- 1730. 34. Oka. Y. & Czech, M. P. (1986) J . Biol. Chern. 261, 9090-9093. 35. Braulke, T., Geuze, H., Slot, J. W., Hasilik, A. & von Figura. K. (1987) Eur. J . Cell Biol. 43, 316-321. 36. Rogers, S. A. & Hammerman (1988) Proc. Nut1 Acad. Sci. USA 85, 1-5. 37. Rogers. S. A. & Hammerman (1989) J. Biol. Chern. 264, 42734276. 38. Nishimoto, I., Hata, Y., Ogata, E. & Kojima, I. (1987) J . B i d . Chem. 262, 12120-12126. 39. Nishimoto, I., Murayama, Y., Katada, T., Ui, M. & Ogata, E. (2989) .I. Biol. Chem. 264, 14029- 14038.