Eur. J. Biochem. 210, 161- 168 (1992) 0FEBS 1992

Biosynthesis and metabolism of dipeptidylpeptidase IV in primary cultured rat hepatocytes and Morris hepatoma 7777 cells Nikolaus LOCH Rudolf TAUBER', Andreas BECKER ', Sabine HARTEL-SCHENK' and Werner REUTTER' Institut fur Molekularbiologie und Biochemie, Freie Universitat Berlin, Federal Republic of Germany Institut fur Klinische Chemie und Biochemie, Universitatsklinikum Rudolf-Virchow, Freie Universitat Berlin, Federal Republic of Germany

'

(Received July 6/August 31, 1992) - EJB 92 0958

N-Glycosylation, biosynthesis and degradation of dipeptidylpeptidase IV (EC 3.4.14.5) (DPP IV) were comparatively studied in primary cultured rat hepatocytes and Morris hepatoma 7777 cells (MH 7777 cells). DPP IV had a molecular mass of 105 kDa in rat hepatocytes and of 103 kDa in MH 7777 cells as assessed by SDSjPAGE under reducing conditions. This difference in molecular mass was caused by differences in covalently attached N-glycans. DPP IV from hepatoma cells contained a higher proportion of N-glycans of the oligomannosidic or hybrid type and therefore migrated at a slightly lower molecular mass. In both cell types DPP IV was initially synthesized as a 97-kDa precursor which was completely susceptible to digestion with endo-p-N-acetylglucosaminidaseH converting the molecular mass to 84 kDa. The precursor was processed to the mature forms of DPP IV, glycosylated with N-glycans mainly of the complex type with a half-life of 20 - 25 min. The transit of newly synthesized DPP IV to the cell surface displayed identical or very similar kinetics in both cell types with the major portion of DPP IV appearing at the cell surface after 60min. DPP IV molecules were very slowly degraded in hepatocytes as well as in hepatoma cells with half-lives of approximately 45 h. Inhibition of oligosaccharide processing with I-deoxymannojirimycin led to the formation of DPP IV molecules containing N-glycans of the oligomannosidic type. This glycosylation variant was degraded with the same half-life as complex-type glycosylated DPP IV. By contrast, inhibition of N-glycosylation with tunicamycin resulted into rapid degradation of non-N-glycosylated DPP IV molecules in both cell types. Non-N-glycosylated DPP IV could not be detected at the cell surface indicating an intracellular proteolytic process soon after biosynthesis.

Dipeptidylpeptidase IV (DPP IV) is a serine-type exopeptidase cleaving N-terminal dipeptides from polypeptides with proline or alanine as the penultimate amino acid [l, 21. Though present in various other tissues and cell types, DPP IV is particularly enriched in the brush border membranes of small intestine, kidney proximal tubules and bile canaliculi [ 3 , 41. In the kidney proximal tubules the enzyme is thought to be involved in the metabolism of peptides and confers uptake of proline-containing dipeptides [5]. Moreover, DPP IV has been postulated to play a role in the metabolism of peptide hormones and neuropeptides like substance P, interleukin 2 and casomorphin [6-91. Recently DPP IV was shown to be involved in cell -matrix adhesion and spreading events mediating cellular binding to collagen IV and fibronectin

[lo - 141. In special subsets of T-helper (Tp) cells DPP IV was identified as the CD 26 leukocyte differentiation antigen [I5 171. CD 26 not only represents a cell surface marker enzyme but also plays an important role as a T-cell activation antigen in the human as well as in the mouse system [18,19]. DPP IV is an integral membrane glycoprotein occurring as a homodimer with subunits of 100- 130 kDa depending on the tissue and the species of origin [3,20,21]. As shown by cDNA sequencing and studies on the membrane orientation, DPP IV in the bile canalicular membrane is constituted of a small amino-terminal cytoplasmic domain, a large extracellular domain and one membrane-spanning domain formed by the signal peptide that is not cleaved off during biosynthesis [22 - 271. As deduced from the cDNA sequence, DPP IV has eight potential N-glycosylation sites distributed all over the Correspondence to N. Loch, Institut fur Molekularbiologie und Biochemie, Freie Universitat Berlin, Arnimallee 22, W-1000 Berlin extracellular domain of the molecule [22, 241. Analysis by lectin blotting has suggested that in DPP IV from rat liver 33, Federal Republic of Germany Abbreviations. dMM, 1-deoxymannojirimycin; DPP IV, dipep- the N-glycans are predominantly of the complex type [28], tidylpeptidase IV; DMEM, Dulbecco's modified essential medium; supporting conclusions based on metabolic labeling or carboendo H, endo-P-N-acetylglucosaminidaseH; MH 7777, Morris hepa- hydrate constituent analysis [21, 29, 301. Structural analysis toma 7777; N HS-SS-biotin, sulfosuccinimidyl 2-(biotinamido)ethyl- of the sugar chains of DPP IV from rat kidney brush border 1,3-dithiopropionate;NP-40, Nonidet P-40; PNGase F, peptide N4- membrane has revealed the presence of di-, tri- and ( N-acetyl-P-glucosaminy1)asparagineamidase F. tetraantennary complex-type structures containing a bisecting Enzymes. Endo-P-N-acetylglucosaminidaseH (EC 3.2.1.96); peptide N4-(N-acetyl-P-glucosaminyl)asparagine amidase F (EC N-acetylglucosamine residue [31]. The biological role of the N-glycans of DPP IV is unknown. Studies on the surface 3.2.2.18); sialidase (EC 3.2.1.18); papain (EC 3.4.22.2); dipeptidylpeptidase IV (EC 3.4.14.5). transport of newly synthesized DPP IV in the human colonic

162 adenocarcinomd cell line Caco-2 showed that inhibition of processing glucosidase 1 by N-methyl-1-deoxynojirimycin inhibited transport to the microvillar membrane by more than So%, presumably due to conformational alterations of the protein backbone induced by the presence of glucosylated oligosaccharides [32]. During internalization and surface recycling, oligosaccharide chains of DPP IV have been shown to undergo continuous loss and re-transfer of outer L-fucose and sialic acid residues thought to represent trimming by cellular glycosidases and a possible repair mechanism [29, 30, 331. Furthermore, as evidenced by use of a monoclonal antibody against a carbohydrate epitope, glycosylation of DPP IV undergoes distinct alterations during transformation of hepatocytes that parallel changes in the expression in hepatoma cells and hepatoma tissues [34]. In an attempt to further characterize the biological role of the N-linked glycans of DPP IV, we have studied N-glycosylation, oligosaccharide processing and degradation of DPP IV comparing different glycoforms of the enzyme in primary cultures of rat hepatocytes and Morris hepatoma 7777 cells.

(by vol.) horse serum. Cells were allowed to adhere overnight. The monolayers obtained were washed twice with ice-cold NaCl/Pi (150 mM NaC1,3 mM KCI, 8 mM Na2HP04, 1 mM KHZPO4,pH 7.4) and preincubated for 60 rnin with DMEM without L-methionine. For pulse labeling in biosynthetic experiments, cells were grown in the presence of 10 MBq/ml ~-[~~S]methionine for 20 min. Thereafter, the medium was withdrawn and the cells were chased in DMEM with 10% (by vol.) horse serum supplemented with 50 pM unlabeled Lmethionine for the indicated times. For turnover studies, pulse labeling was performed for 4 h with 1.85 MBq/ml of L[35S]methionine. Labeling experiments, intended to characterize the structural properties of DPP IV, were performed for 15 h by addition of 1.85 MBq/ml ~-[~~S]methionine in Lmethionine-free DMEM. For labeling in the presence of 2 mM dMM or 1 pg/ml tunicamycin, cells were pretreated with inhibitor at identical concentrations for 1 h before addition of radioactive amino acid. Inhibitor concentrations were maintained throughout the labeling period. lmmunoprecipitation and SDS/PAGE

All steps were carried out at 4°C. Labeled cells were washed twice with NaCl/Pi, scraped from the dishes with a Materials rubber policeman and collected by centrifugation (10 min, Constituents of tissue culture media were obtained from 5000 x g ) . For detergent extraction, cells were resuspended in Biochrom (Berlin, FRG); other materials for tissue culture 5 ml lysis buffer A (150 mM NaCl, 10 mM Tris/HCl pH 8.0, were purchased from Falcon (Heidelberg, FRG) or Nunc 1 mM EDTA, 2 mM phenylmethylsulfonyl fluoride, 1YOby (Wiesbaden, FRG). ~ - [ ~ ~ S ] M e t h i o nwas i n e from Amersham- vol. NP-40), homogenized by 10 gentle strokes in a Dounce Buchler (Braunschweig, FRG). Protein-A - Sepharose was loose-fitting homogenizer and further kept on ice for 2 h. obtained from Pharmacia (Freiburg, FRG). 1-Deoxy- Detergent-insoluble material was removed by centrifugation mannojirimycin (dMM) was a kind gift from Dr. Schuler (100000 xg, 30 min). The supernatants were precleared by (Bayer AG, Wuppertal, FRG) ; tunicamycin was obtained incubating with 100 mg Sepharose-4B for at least 2 h. from Calbiochem (Frankfurt, FRG); all glycosidases were Sepharose4B was pelleted by centrifugation and discharged. purchased from Boehringer (Mannheim, FRG). To the supernatants 20 pg monoclonal antibody 13.4 coupled Sulfosuccinimidyl 2-(biotinamido)ethyl-1,3-dithiopropionate to 8 mg protein-A-Sepharose was added as suspension in (NHS-SS-biotin) and avidin-agarose were obtained from lysis buffer. The suspension was shaken end-over-end at 4°C Pierce (Oud Beijerland, The Netherlands). Glycyl-prolyl-p- for at least 4 h. Immunocomplexes bound to protein-Anitroanilide tosylate was purchased from Serva (Heidelberg, Sepharose were pelleted by centrifugation, supernatants were FRG). Unless otherwise stated, all other chemicals and re- removed and the pellet was washed five times with washing agents were obtained either from Sigma (Deisenhofen, FRG) buffer B (500 mM NaC1, 10 mM Tris/HCl pH 8.0, 1 mM or from Serva (Heidelberg, FRG). The monoclonal antibody EDTA, 1% by vol. NP-40). In a final washing step NaCI/P, 13.4. used in these experiments has been described previously was used. Immunocomplex coupled to protein-A - Sepharose was eluted by heating (95°C) for 3 min with 50 pl 4% SDS, WI. 28.6% (by vol.) glycerol, 10 mM mercaptoethanol, 50 mM Tris/HCl pH 6.8 (SDS electrophoresis sample buffer). If not Cell culture otherwise stated, electrophoresis was performed in 7.5% Primary hepatocytes were isolated according to the pro- polyacrylamide gels in the presence of 0.1 YOSDS as described cedure of Seglen [36] at a minimum viability of 80-90% as by Laemmli [37]. shown by the trypan-blue exclusion test. Cells were seeded on collagen-I-coated plastic dishes. The hepatoma cell line, Cell surface biotinylation derived from Morris hepatoma 7777, was a kind gift from Prof. Dr. K. von Figura (Gottingen, FRG). Hepatocytes and Cell surface proteins were labeled with biotin essentially hepatoma cells were maintained in Dulbecco’s minimal essen- as described [38]. Metabolically labeled cells were washed tial medium (DMEM) supplemented with penicillin (50 U / three times with ice-cold NaCl/P, plus 0.1 mM CaClz and ml), streptomycin (50 pg/ml), insulin (1 nM) dexamethasone incubated with a freshly prepared solution (2 mg/ml) of NHS(10 nM) and 10% (by vol.) complement-inactivated horse SS-biotin in NaCl/Pi plus 0.1 mM CaClz at 4°C for 10 min. serum in a humidified atmosphere with 5% COz at 37°C. In order to block unreacted NHS-SS-biotin, cells were washed twice with NaC1/Pi plus 0.1 mM CaClz containing 0.1% (massjvol.) bovine serum albumin. Cells were extracted as Labeling of cells detailed in the preceding section with a modified lysis buffer Freshly isolated hepatocytes and confluent hepatoma cells A, which additionally contained 10 mM L-lysine (buffer C). prepared by trypsinization were seeded on collagen-I-coated DPP IV was immunoprecipitated from detergent lysate as dishes (35 mm diameter) at a density of 5 x lo5 cells/well and detailed above and eluted twice with 10 mM glycine/HCl, lo6 cells/well respectively, in DMEM supplemented with 10% pH 3.0. Eluates were immediately neutralized with the same MATERlALS AND METHODS

volume of 500 mM sodium phosphate pH 7.5 and diluted tenfold in buffer A. Then 100 pl of avidin-agarose was added and allowed to react with biotinylated DPP IV for at least 2 h. After five washing steps in buffer B, the avidin-bound biotinylated protein was eluted by boiling for 3 min in 80 p1 4% SDS, 28.6% (per vol.) glycerol, 10 mM mercaptoethanol, 50 mM TrisjHCl pH 6.8.

163

a

A

~ ~ ~ 1 0 - 3

b

I

B

B

I

1

2

I

A

3'

1

2

3

205 116-

Treatment with glycosidases

97 -

66 Immunoadsorbed DPP IV was eluted from protein-A Sepharose with 20 pl 4 % SDS, 10 mM mercaptoethanol, 45 50 mM Tris/HCl pH 6.8 and split into two portions. One of these was incubated with the respective glycosidase at 37 "C for 20 h, whereas the other was mock-incubated under identical 29 conditions without enzyme. Before addition of the giycosidases, all samples were diluted tenfold in the buffer recommended by the manufacturer and a mixture of protease inhibitors (leupeptin, chymostatin, antipain, pepstatin (10 bg Papain: + + - + + each) was added. Treatment with sialidase from Arthrobacter Fig.1. SDS/PAGE of DPP IV isolated from primary cultured rat ureafaciens ( 5 m u ) was performed in 50 mM sodium acetate hepatocytes and MH 7777 cells. Primary cultured rat hepatocytes (A) pH 5.5. For treatment with 5 mU of endo-P-N- and MH 7777 cells (B) were labeled with 1.85 MBq ~ - [ ~ ~ S ] m e t h i o n i n e acetylglucosaminidase H (endo H) from Streptomycesplicatus for 15 h. (a) Cells were detergent-extracted and DPP IV was samples were incubated in 50 mM sodium phosphate pH 6.0. immunoprecipitated. (b) A crude membrane fraction was obtained Digestion with 10 mU peptide-N4-(N-acetyl-~-glucosaminyl)- from the labeled cells and was split into two aliquots, one of which asparagine amidase F (PNGase F) from Flavobacterium was immediately detergent-extracted whereas the other one was meningosepticum was performed in 500 mM sodium phos- treated with papain followed by detergent extraction. DPP IV was immunoprecipitated from (1) hepatocytes (A) and MH 7777 cells (B) phate pH 8.6, containing 1YO(by vol.) Mega-10. after detergent extraction without prior treatment with papain, (2)

Papain treatment

After washing with NaCl/P,, cells metabolically labeled with ~ - [ ~ ~ S ] m e t h i o nwere i n e scraped from the dishes and homogenized in a hypotonic medium (1 mM NaHC03, 0.5 mM CaClz pH 7.4) with 20 strokes in a loose-fitting Dounce homogenizer. Crude membrane pellets were obtained by centrifugation at I00000 x g for 30 min, redispersed in 100 mM TrisjHCl pH 8.8, containing 1 mM EDTA and divided into two portions. For control, one portion was directly detergent-extracted in lysis buffer A as described in the section on immunoprecipitation, and DPP IV was subsequently immunoadsorbed from the NP-40 extract. The other portion was treated with papain (5 pg/lOOpg membrane) for 2 h at 37 "C. Papain was inactivated by the addition of iodoacetic acid at a final concentration of 1 mM. Membranes and watersoluble supernatants were obtained by centrifugation at I00000 x g for 30 min and DPP 1V was immunoadsorbed from both the supernatant and the detergent-extracted membranes. RESULTS Determination of molecular mass

DPP IV immunoprecipitated from primary cultured rat hepatocytes and Morris hepatoma 7777 cells (MH 7777 cells) migrated as a single polypeptide with a molecular mass of 105 kDa (hepatocytes) or 103 kDa (MH 7777 cells) when analyzed by SDSjPAGE under reducing conditions (Fig. 1a). Treatment of membranes from both primary cultured rat hepatocytes and hepatoma cells with papain resulted in the complete release of DPP IV in a water-soluble form (Fig. lb). The molecular mass of the proteolytically released enzyme was slightly reduced. As papain specifically cleaves at a site near the N-terminal membrane-spanning domain of DPP IV

from water-soluble supernatants obtained after digestion with papain, (3) detergent extracts from remaining membrane pellets after treatment with papain. Immunoprecipitates were heated at 95 "C for 5 min in SDS electrophoresis buffer. Eluted enzyme was separated by SDS/ PAGE on a 7.5% gel and analyzed by fluorography.

[3, 291, this indicates an identical, or at least very similar, size of the N-terminal sequences of DPP IV from both cell types. N-Glycosylation

Digestion with endo H cleaving N-glycans of the hybrid and the oligomannosidic type [39] slightly reduced the molecular mass of DPP IV from hepatoma cells (Fig. 2, lanes 7, 8), but not that of DPP IV from hepatocytes (Fig. 2, lanes 5, 6). This indicates that the N-glycans of mature DPP 1V in hepatocytes are almost completely processed to complex-type structures, whereas DPP IV from MH 7777 cells partly retains one or two N-glycan chains of the oligomannosidic or hybrid type. When N-glycan processing of DPP IV was inhibited by the a-mannosidase I inhibtor dMM [41] the unprocessed forms of DPP IV had a molecular mass of 97 kDa in both cell types (Fig. 2, lanes 9,11). These unprocessed forms of DPP IV were completely sensitive to digestion with endo H (Fig. 2, lanes 9-12) resulting in the formation of a polypeptide with a molecular mass of 84 kDa. The different sensitivity to endo H of the processed and the unprocessed forms of DPP IV could also be shown when mixtures of unprocessed and mature DPP IV were treated with this endoglycosidase (Fig. 2, lanes 13 - 16). Treatment with PNGase F cleaving all types of N-glycans [40] converted mature DPP IV from both cell types to polypeptides with an identical molecular mass of 84 kDa (Fig. 2, lanes 18, 20). DPP IV from hepatocytes as well as from hepatoma cells was sensitive to sialidase reducing the molecular mass by approximately 2 kDa (Fig. 2, lanes 1-4).

164

-

11697 66

-

45

-

Neuraminidare

1

2

3

4

-

+

-

+

5

-

6

7

----(----(-+I-

EndoH L

I

I

I

PNGase F dMM

8

9

+ -( I

1

-

-

-

-

-

-

-

-

-

-

-

-

-

-

- -

18

-

-

-

-

-

-

-

-

-

-

+

-

+/-+/-+I-+/-

-

-

-

11

+

_ - - _ - - + - + -

-

-

+ +

12

17

10

_ _ - _

-

I

13 14 15 16

I

-

-

+

+

I

-

-

+ I

-

A

19 20

+

I

-

M, x 10-3

- 116 - 97

Chase [rnin] Endo H

20

10

0

+

-

+ -

-

40

30 f

+

-

-

60

+

-

240

+

-

+

B

M, x 10-3

-

Endo H

10

0

Chase[min] -

+

20

-

Time-dependent sequential deglycosylation of unprocessed DPP IV from both cell types with endo H yielded six distinct intermediates indicating that DPP IV contains at least six N-glycans (not shown). Kinetics of N-glycan processing

The biosynthesis of DPP IV was comparatively studied in hepatocytes -and MH 7777 cells in pulse-chase experiments

30

40

60

116 97

960

+

-

+

(Fig. 3 ) . Hepatocytes and MH 7777 cells were pulse-labeled for 20 min and DPP IV was immunoprecipitated from detergent extracts obtained from the cells after different chase periods. The immunoprecipitates were divided into two vortions. one of which was mock-incubated whereas the other one was treated with endo H. Immediately after the pulse, newly synthesized DPP IV appeared as a homogenous band with an apparent molecular mass of 97 kDa in both cell types. This precursor was completely endo-H-sensitive being -con-

165 verted to the non-glycosylated 84-kDa form of DPP IV (Fig. 3). As early as 20 rnin after the pulse, an additional band of 105 kDa in hepatocytes and 103 kDa in hepatoma cells appeared that increased in amount during the chase period, whereas the 97-kDa form decreased. The mature 105-kDa polypeptide found in hepatocytes was almost completely resistant to endo H, whereas the mature 103-kDa polypeptide in MH 7777 cells was converted into two isoforms of either 102 kDa or 100 kDa. In hepatocytes the 97-kDa precursor was processed to the mature form of DPP IV with a roughly estimated half-life of 20 - 25 rnin which seemed to be slightly delayed in MH 7777 cells.

A

B

i

I

r

I

M, x 10-3

a

116

-

97

-

I Chase[minl I

0

40

20

I

60 120

0

A

Transport to the cell surface

40

60

120

I

B

I

The kinetics of appearance of newly synthesized DPP IV on the cell surface was studied in pulse-chase experiments in conjunction with surface labeling of proteins with NHS-SSbiotin. Cells were pulse-labeled for 20 rnin with ~ - [ ~ ~ S ] m e t h i o nine followed by a chase with an excess of unlabeled L-methionine. At different times of the chase period, cells were cooled to 4°C and cell surface proteins were covalently labeled with b NHS-SS-biotin. Following detergent lysis, DPP IV was isolated by immunoadsorption and surface-exposed, i.e. biotinylated, DPP IV was subsequently separated from intracellular DPP IV by affinity chromatography on avidin-agarose. Newly synthesized DPP IV appeared on the cell surface of hepatocytes and hepatoma cells after approximately 60 rnin (Fig. 4). No immature 97-kDa species were biotinylated and bound to avidin-agarose, confirming that the applied method of protein biotinylation is selective for cell-surface-exposed proteins in accordance with previous results [38].

20

I

I

1

M, x 10-3 116 97

-

Chase [min]

0

20 40

60 120

0

20 40 60

120

Degradation

Half-lives of degradation of DPP IV were determined in hepatocytes and hepatoma cells in pulse-chase experiments. DPP IV molecules were slowly metabolized in both cell types with half-lives of approximately 45 h as determined by densitometry of the fluorographs (Fig. 5). In order to assess the influence of N-linked glycans on the metabolic stability of DPP IV, the degradation of two glycosylation variants of DPP IV was studied. DPP IV, solely glycosylated with oligomannosidic glycans, was generated by labeling in the presence of dMM. Non-N-glycosylated DPP IV was obtained in the presence of tunicamycin. DPP IV glycosylated with oligomannosidic N-glycans, was metabolized as slowly as DPP IV glycosylated with glycans of the complex-type (Fig. 5). Moreover, the oligomannosidic variant was transported to the cell surface (Fig. 6). By contrast, non-N-glycosylated DPP IV was rapidly degraded within the first 6 h after biosynthetic labeling (Fig. 5) and was not detectable at the cell surface (Fig. 6).

DISCUSSION DPP IV from hepatocytes displayed a molecular mass of 105 kDa, in good accord with estimations performed by other laboratories for DPP IV from rat liver [II, 14,21,24,28]. The molecular mass of DPP IV isolated from MH 7777 cells was slightly lower (103 kDa) when separated by SDS/PAGE. The 105-kDa and the 103-kDa species were expressed at the cell surface and, hence, may be regarded as the mature forms of the enzyme.

As shown by comparison of cDNA and protein sequences, hepatic DPP IV has an uncleaved signal sequence at its NH2 terminus functioning as the membrane-anchoring domain [22, 271. The membrane-spanning domain and the cytoplasmic domain can be removed by digestion with papain which cuts between amino acid residues 34 and 35 [24], thus releasing a water-soluble enzymatically intact form of DPP IV [20]. In hepatocytes, as well as in hepatoma cells, DPP IV was completely released by papain. The reduction in the molecular mass of the proteolytically released enzymes was approximately the same in both cell types, indicating that the sizes of the membrane-spanning and cytoplasmic domains of DPP IV are at least very similar in both cell types. DPP IV from both hepatocytes and hepatoma cells contains at least six N-linked oligosaccharides as determined by stepwise N-deglycosylation. This accords well with the recently published primary sequence of DPP IV containing eight potential Asn-Xaa-Ser/Thr acceptor sites for N-glycosylation [22, 241. As characterized by digestion with sialidase, endo H and PNGase F, the N-glycans of the mature enzyme are mainly of the sialylated complex type. The slight difference in the apparent molecular mass of DPP IV from hepatocytes and hepatoma cells is caused by differences in N-glycosylation. This is concluded from the findings that digestion of the mature and unprocessed forms of DPP IV from hepatocytes and hepatoma cells with PNGase F and endo H, respectively,

166 A

Non -Glycosylated

High Mannose

Complex

M, x 10-3

116 97

66

-

-

Chaselhl:

0

B

15

35

60

85

0

110

15

35

Complex

60

85

0

110

15

35

85

60

High Mannore

110

Non -Glycosylared

M. x

11697

-

66

-

45

-

ChaseIh]:

0

6

16

30

49

70

92

122

0

6

16

30

49

70

92

122

0

6

16

30

49

70

92

Fig. 5. Degradation of different glycoforms of DPP IV. Hepatocytes (A) and MH 7777 cells (R) were incubated for 1 h in L-methionine-free DMEM in the presence or absence of 2 mM dMM or 1 pgjml tunicamycin, respectively. Cells were pulse-labeled with 1.85 MBq ~ - [ ~ ~ S ] m e t h i o nine for 4 h under maintainance of the indicated inhibitors throughout the labeling period. Subsequently, cells were chased with complete medium containing excess of unlabeld L-methionine. At different chase intervals cells were lysed and detergent-extracted. DPP IV was immunoprecipitated from the extracts and was analyzed by SDSjPAGE and fluorography. DPP IV immunoadsorbed from hepatocytes (A) and MH 7777 cells (B) either glycosylated with N-linked glycans of the complex or oligomannosidic type (high mannose) or without N-linked glycans (Non-glycosylated).

yielded the same unglycosylated polypeptide with an apparent molecular mass of 84 kDa. As may be inferred from the different sensitivity to endo H, cleaving N-glycans of the oligomannosidic and of the hybrid type, N-glycans of DPP IV from hepatocytes are almost completely processed to complextype structures whereas DPP IV from MH 7777 cells retains one or two N-glycans of the oligomannosidic/hybrid type. DPP IV is initially synthesized as a 97-kDa precursor that is processed to the mature 105-kDa species in primary cultured rat hepatocytes and to the 103-kDa species in MH 7777 cells. Conversion from the 97-kDa precursor to the mature glycoprotein begins within 20 min after biosynthesis and is largely completed after 60 - 70 min in both cell types. Maturation from the 97-kDa species to the 105-kDa and the 103-kDa species reflects processing of the N-linked oligosaccharides from structures of the oligomannosidic type to structures of the sialylated complex type. In hepatocytes and hepatoma cells DPP IV acquired endo-H-resistant structures with an estimated half-life of 20 - 30 min. Since GlcNAc transferase I and a-mannosidase I1 rendering N-linked oligosaccharides endo-H-resistant [42] are localized in the medial Golgi subcompartment, the time required for conversion of the glycoprotein from the endo-H-sensitive to the endo-H-resistant type corresponds to the time required for transport from the site of oligosaccharide addition in the endoplasmic reticulum to the central Golgi cisternae. The transport rate measured for DPP IV is comparable to that of other membrane glycoproteins [43-461. Compared to the kinetics of transport from

the endoplasmic reticulum to the medial Golgi apparatus, the time required for newly synthesized DPP IV to reach the cell surface was longer. At the cell surface DPP IV was detectable in larger amounts not earlier than 60 min after biosynthesis. By contrast, 5’-nucleotidase in H4S-hepatoma cells reaches the cell surface as early as 20 min after biosynthesis [44]. Our results indicate that, in addition to the exit from the endoplasmic reticulum, an intra-Golgi event causes a delay in the surface transport of newly synthesized DPP IV. Similar kinetics were reported for DPP IV in Caco 2 cells [43, 471. It has been reported that further conformational maturation accounts for this delay [48]. Only mature species of the enzyme could be detected on the surface, showing that no newly synthesized molecules escape oligosaccharide processing during surface transport. Processing of oligomannosidic precursor oligosaccharides to complex structures is not a prerequisite for surface transport, as the unprocessed variant of DPP IV with oligomannosidic N-glycdns synthesized in the presence of dMM could also be detected on the cell surface. By contrast, non-glycosylated DPP IV synthesized in the presence of tunicamycin failed to reach the cell surface and was rapidly degraded after biosynthesis intracellularly. As has been shown for several glycoproteins, carbohydrate side chains prevent degradation by masking domains which are sensitive to a proteolytic attack (for review see [49]). Moreover, N-glycans may also influence the proper folding of newly synthesized proteins [50, 511. So far, we cannot decide which of these effects accounts for the rapid degradation of the non-

167 M, x 10-3

116

-

This work was supported by the Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg (Re 523/4-1, SFB 312), the MariaSonnenfeld-Stijtung and the Trude Goerke-Stiftung.

REFERENCES

1. Hopsu-Havu, V. K. & Sarimo, S. R. (1967) Hoppe-Seyler’s Z . Physiol. Chem. 348,1540 - 1550. 97 2. Kenny, A. J., Booth, A. G., George, S. G., Ingram, J., Kershaw, D., Wood, E. J. &Young, A. R. (1976) Biochem. J . 155,169182. 66 3. Gossrau, R. (1979) Histochemistry 60,231 -248. 4. Hartel, S., Gossrau, R., Hanski, C . & Reutter, W. (1988) Histochemistry 89, 151- 161. 5. Tiruppathi, C., Miyamoto, Y., Ganapathy, V., Roesel, R. A,, Whitford, G. M. & Leibach, F. H. (1990) J . Biol. Chem. 265, 1476- 1483. 45 6. Palmieri, F. E. &Ward, P. E. (1982) Biochim. Biophys. Acta 755, 522 - 525. 7. Nausch, I. & Heymann, E. (1985) J . Neurochem. 44,1354- 1357. 8. Mentlein, R. (1988) FEBS Lett. 234, 251 -256. 9. Nausch, I., Mentlein, R. & Heymann, E. (1990) Biol. Chem. Hoppe-Seyler 371, 1113- 1118. 10. Hanski, C., Huhle, T. & Reutter, W. (1985) Biol. Chem. HoppeSeyler 366, 1169- 1176. a b c 11. Walborg Jr., E. F., Tsuchida, S., Weeden, D. S., Thomas, M. W., Fig.6. Cell surface expression of different glycoforms of DPP IV. Barrick, A,, McEntire, K. D., Allison, J. P. & Hixson, D. C. Primary cultured rat hepatocytes were grown for 60 min in l-methio(1985) Exp. Cell Res. 158, 509-5518, nine-free DMEM in the presence of either 2 mM dMM or l ~ g / m l 12. Hanski, C., Huhle, T., Gossrau, R. & Reutter, W. (1988) Exp. tunicamycin or without inhibitor. Cells were then pulse-labeled with Cell Res. 178, 64 - 72. 3.7 MBq/ml ~ - [ ~ ~ S ] m e t h i o nfor i n e2 h. Subsequently cells were sur13. Bauvois, B. (1988) Biochem. J . 252,723-732. face-labeled with NHS-SS-biotin as described in Materials and 14. Piazza, G. A,, Callanan, H. M., Mowery, J. & Hixson, D. C. Methods, lysed and detergent-extracted. DPP IV was immunoad(1989) Biochem. J . 262,327 - 334. sorbed from detergent extracts and eluted from the protein-A15. Mentlein, R., Heymann E., Scholz, W., Feller, A. C. & Flad, H.Sepharose. Biotinylated DPP IV molecules were adsorbed to avidinD. (1984) Cell. Zmmunol. 89, 11 -19. agarose eluted with SDS electrophoresis sample buffer and analyzed 16. Feller, A. C., Parwaresch, M. R. & Lennert, K. (1984) Leukemia by SDSjPAGE and fluorography. Surface-biotinylated DPP IV metaRes. 8, 397 - 406. bolically labeled with ~ - [ ~ ~ S ] m e t h i o n(a) i nin e the absence of inhibitor, 17. Ulmer, A. J., Mattern, T., Feller, A. C., Heymann, E. & Flad, (b) in the presence of dMM, or (c) in the presence of tunicamycin. H.-D. (1990) Scand. J . Immunol. 31,429-435. 18. Hegen, M., Niedobitek, G., Klein, C. E., Stein, H. & Fleischer, B (1990) J. Zmmunol. 144, 2908 - 291 4. 19. Vivier, I., Marguet, D., Naquet, P., Bonicel, J., Black, D., Li, C. X.-Y., Bernard, A.-M., Gorvel, J.-P. & Pierres, M. (1991) J. glycosylated DPP IV molecules. We are currently investigating Immunol. 141,4101 -4109. whether these molecules are subject to degradation in the endoplasmic reticulum. As has been shown for several mem- 20. Macnair, R. D. & Kenny, A. J. (1979) BioLhem. J. 179,379-395. brane and secretory glycoproteins, e.g. subunits of the T cell 21. Elovson, J. (1980) J. Biol. Chem. 255, 5807-5815. 22. Hong, W. & Doyle, D. (1987) Proc. Natl Acad. Sci. USA 84, receptor [52], cystic fibrosis transmembrane conductance 7962 - 7966. regulator 1531 and mutant a,-antitrypsin 1541, incompletely 23. Hong, W. &. Doyle, D. (1987) J . Biol. Chem. 263,16892-16898. assembled or wrongly folded polypeptides are retained within 24. Ogata, S., Misumi, Y . & Y. Ikehara (1989) J. Biol. Chem. 264, this compartment and become rapidly degraded by a proteo3596 - 3601. lytic system that most likely resides within the endoplasmic 25. Hong, W. & Doyle, D. (1990) J . Cell Biol. 111, 323-328. reticulum. Recently, this pathway of degradation was demon- 26. Darmoul, D., Lacasa, M., Chantret, I. Swallow, D. M. & Trugnan, G. (1990) Annu. Hum. Genet. 54, 191 -197. strated for the transferrin receptor synthesized in the presence of tunicamycin 1551 and for a mutated transferrin receptor 27. Marguet, D., Bernard, A.-M., Vivier, I., Darmoul, D., Naquet, P. & Pierres, M. (1992) J . Biol. Chem. 267, 2200-2208. lacking the N-glycosylation site closest to the transmembrane region [56]. By contrast to these glycoproteins, N-linked oligo- 28. Bartles, J. R., Braiterman, L. T. & Hubbard, A. L. (1985) J . Biol. Chem. 260, 12792- 12802. saccharides are not required for the formation of meta- 29. Kreisel, W., Volk, B. A,, Buchsel, R. & Reutter, W. (1980) Proc. bolically stable species nor for surface transport of other memNatl Acad. Sci. USA 77, 1828- 1831. brane glycoproteins including 5’-nucleotidase [44], the low- 30. Volk, B. A,, Kreisel, W., Kottgen, E., Gerok, W. & Reutter, W. density-lipoprotein receptor [57], and the asialoglycoprotein (1983) FEBS Lett. 163, 150-152. receptor [%I. Therefore, the cotranslational addition of N- 31. Yamashita, K., Tachibana, Y . , Matsuda, Y., Katunuma, N., Kochibe, N. & Kobata, A. (1988) Biochemistry 27,5565 - 5573. linked oligosaccharides to the nascent polypeptide distinctively affects the biosynthetic folding of membrane glyco- 32. Matter, K., McDowell, W., Schwartz, R. T. & Hauri, H.-P. (1 989) J . Biol. Chem. 264, 13131 -13139. proteins and their metabolic stability. By contrast to the inhibition of N-glycosylation by tunicamycin, inhibition of N- 33. Kreisel, W., Hanski, C., Tran-Thi, T.-A., Katz, N., Decker, K. & Reutter, W. (1988) J . Biol. Chem. 263, 11736-21742. glycan processing by dMM had no influence on the turnover 34. Hartel-Schenk, S., Loch, N., Zimmermann, M. & Reutter, W. rate of the enzyme, indicating that the transfer of N-glycan (1991) Eur. J . Biochem. 196, 349-355. units, but not their subsequent processing, confers stability to 35. Becker, A,, Neumeier, R., Heidrich, C., Loch, N., Hartel, S. & the glycoprotein. Reutter, W. (1986) Biol. Chem. Hoppe-Seyler 367, 681 -688.

-

-

-

168 36. Seglen, P. 0. (1967) Methods Cell Biol. 13, 30-83. 37. Laemmli, U. K. (1970) Nature 227, 680-685. 38. Busch, G., Hoder, D., Reutter, W. & Tauber, R. (1989) Eur. J. Cell Biol. 50, 257- 262. 39. Tarentino, A. L. & Maley, F (1974) J . Biol. Chem. 249,811 -817. 40. Plummer, T. H., Jr., Elder, J. H., Alexander, S., Phelan, A. W. & Tarentino, A. L. (1984) J. Biol. Chem. 259, 10700-10704. 41. Fuhrmann, U., Bause, E., Legler, G. & Ploegh, H. L. (1984) Nature 307, 755-758. 42. Hubbard, S. C. & Ivatt, R. J. (1981) Annu. Rev. Biochem. 50, 555- 583. 43. Hauri, H.-P., Sterchi, E., Bienz, D., Fransen, .I. A. M. & Marxer, A. (1985) J. Cell Biol. 101, 838-851. 44. Van den Bosch, R., Geuze, H. J., du Maine, A. P. M. & Strous, G. J. (1986) Eur. J . Biochem. 160,49-54. 45. Wada, I., Himeno, M. & Kato, K. (1986) . I Biol. . Chem. 261, 2222 - 2227. 46. Lorkowski, G., Zijderhand-Bleekemolen, J. E., Erdos, E. G., von Figura, K. & Hasilik, A. (1987) Biochem. J. 248, 345 - 350. 47. Stieger, B., Matter, K., Baur, B., Bucher, K., Hochli, M. & Hauri, H.-P. (1988) J . Cell Biol. 106, 1853-1861.

48. Matter, K . & Hauri, H.-P. (1991) Biochemistry 30, 1916-1923. 49. Olden, K., Parent, J. B. & White, S. L. (1982) Biochim. Biophys. Acta 650, 209 - 232. 50. Machamer, C . E. & Rose, J. K. (1988) J . Biol. Chem. 263,5955 5960. 51. Danielson, E. M. (1992) Biochemistry 31, 2266-2272. 52. Lippincott-Schwartz, J., Bonifacino, J. S., Yuan, L. C. & Klausner, R. D. (1988) Cell 54, 209 - 220. 53. Cheng, S. H., Gregory, R. J., Marshall, J. Paul, S., Souza, D. W., White, G. A., O’Riordan, C. R. & Smith, A. E. (1990) Cell 63, 827 - 834. 54. Le, A,, Graham, K. S. & Sifers, R. N. (1990) J . Biol. Chem. 265, 14001 -14007. 55. Reckshow, C . L. & Enns, C. A (1988) J . Biol. Chem. 263,12977301. 56. Hoe, M. H. &Hunt, R. C. (1992) J. Biol. Chem. 267,4916-4923. 57. Filipovic, I. (1989) J. Biol. Chem. 264, 8815-8820. 58. Breitfeld, P. P., Rup, D. & Schwartz, A. L. (1984) J . Biol. Chem. 259, 10414 - 10421.