Asparagine-linked Oligosaccharides on Formyl Peptide Chemotactic Receptors of Human Phagocytic Cells*

affinity-la-beled (where represents and ethylene consist of two isoelectric forms with type differences in both apparent size and charge (neutrophils: kDa, PI 5.8, and kDa, PI 5.6 and 6.0; differentiated HL-60 cells: kDa, PI and

Formyl peptide chemotactic receptors affinity-labeled with N-formyl-Nle-Leu-Phe-Nle-['261]iodo-Tyr-Lys (where Nle represents norleucine) and ethylene glycol bis(succinimidy1 succinate) consist of two isoelectric forms with cell type differences in both apparent size and charge (neutrophils: 55-70 kDa, PI 5.8, and 6.2.; monocytes: 60-75 kDa, PI 5.6 and 6.0; differentiated HL-60 cells: 62-85 kDa, PI 5.6 and 6.0). Endo-/3-N-acetylglucosaminidase F (endo F) cleavage of N-linked oligosaccharides from formyl peptide receptor generates 40-50-and 33-kDa products that can be affinity-labeled. Whereas both PI forms of this receptor from neutrophils are cleaved by endo F to 33-kDa final products, this cleavage does not eliminate PI differences. Tunicamycin decreases expression of formyl peptide receptor on differentiating HL-60 and causes a dose-dependent decrease in size of the major product seen after affinity labeling (0.5 pg/ml: 38-48 kDa; 2 pg/ml: 32 kDa). Thus, the formyl peptide receptor polypeptide backbone from all three cell types contains at least two N-linked oligosaccharide side chains which contribute to the cell type differences in M, and are not required for ligand binding. Papain treatment of intact cells generates a membrane-bound formyl peptide receptor fragment that can be affinity-labeled and is of similar size (29-31 kDa) in all three cell types. Endo F treatment of the affinity-labeled papain fragment of formyl peptide receptor does not alter its size, suggesting that this fragment does not contain the N-linked oligosaccharide cleaved by endo F from intact receptor. The results indicate that at least two Nlinked oligosaccharide chains are located on the distal 1-3-kDa portion of the receptor polypeptide backbone. sulfoxide (Me2SO') (6) or N6,02-dibutyryladenosine 3',5'monophosphate (Bt2cAMP) (7). These differentiated HL-60 cells (d-HL-60), like neutrophils, respond functionally to formyl peptides and express specific, saturable binding of radiolabeled formyl peptide (7-10). Neutrophils are terminally differentiated cells capable of only limited protein synthesis, but HL-60 cells provide a useful system for studying formyl peptide receptor synthesis.
We affinity-labeled formyl peptide receptor of human neutrophils, monocytes, and d-HL-60 using formyl lZ5I-peptide and EGS and determined cell type differences in apparent size and charge. Glycosylation of formyl peptide receptor was examined by endo-p-N-acetylglucosaminidase F (endo F) cleavage of N-linked oligosaccharides (16) and by tunicamycin inhibition of N-glycosylation in differentiating HL-60 cells (17). We examined the requirement of glycosylation for formyl peptide binding and used papain and endo F cleavage to determine the location of N-linked oligosaccharides within the polypeptide backbone of formyl peptide receptor.
induced to express some functional and biochemical characteristics of mature neutrophils when exposed to dimethyl *This research was supported by National Institutes of Health Grants AI 18166, AI 19768, and CA 08341 and National Research Service Award AM 06589. Portions of this work have appeared in abstract form (Malech, H. L., and Gardner, J. P. (1984) Clin. Res. 32, 374A). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

N-Linked Oligosaccharides on
Formyl Peptide Receptor bovine serum, 2 mM glutamine (GIBCO Laboratories, Grand Island, NY). HL-60 induced to differentiate with 1.25% Me2S0 (Fischer) were harvested a t day 5 (6), while HL-60 induced to differentiate with 750 p~ BtZcAMP (Sigma) were harvested at day 3 (7). The d-HL-60, but not uninduced HL-60, expressed specific binding of radiolabeled formyl peptide as previously reported (7-10). The d-HL-60 were smaller in size than uninduced cells and had acquired the ability to adhere and spread on the substrate. Only d-HL-60 responded chemotactically to formyl peptide in a Boyden chamber micropore filter assay visually scored by a method described previ-For some studies, tunicamycin (Sigma) a t 0.1-2.0 pg/ml was present in cultures of HL-60 cells during three days of differentiation induced by 750 p~ BtzcAMP. Tunicamycin and other drugs inhibiting the formation of the glycolipid precursors necessary for N-glycosylation have been shown to induce HL-60 cells to differentiate (20, 21). It was therefore not possible to also examine control "uninduced cells treated with tunicamycin for 3 days.
Plasma Membrane-enriched Fraction-A plasma membrane-enriched fraction was obtained by nitrogen pressure cavitation (22) and discontinuous (15-35-40%) sucrose density gradient centrifugation as previously described (12), but with the following modifications. Before cavitation, cells were incubated a t room temperature for 30 min in buffered saline containing 5 mM diisopropyl fluorophosphate (DFP) to inactivate proteases (23). DFP (0.5 mM) and phenylmethylsulfonyl fluoride (PMSF) (0.05 mM) were added to the cavitate, and the sucrose gradient contained 1 mM DFP and 0.05 mM PMSF. The 15-35% sucrose interface contained only membranes when examined by electron microscopy, and no granule markers were detected (22, buffer (pH 6.75) containing 2 mM EGTA and 2 mM EDTA (the 24). The membranes were diluted in 10 mM potassium phosphate gradient buffer without sucrose), pelleted, resuspended a t 1 mg of protein/ml in the same buffer, and stored a t -70 "C.
Protein Determination-Protein determinations were done using a Coomassie Blue Assay (Bio-Rad).
Binding Assay for Formyl Peptide Receptor-Binding of 10 nM formyl Iz5I-peptide to cells was measured using a silicone oil assay as previously described (26).
Affinity Labeling of Formyl Peptide Receptor-Affinity labeling was done as previously described with some modifications (12). Cells before or after papain treatment (see below) were suspended in phosphate-buffered saline (pH 6.75) containing 50 nM formyl lZ5Ipeptide and incubated a t 0 "C for 15 min. Controls also contained 50 p~ nonradioactive formyl peptide. EGS (Pierce Chemical Co.) was added a t 0.3 mg/ml with incubation for 10 min a t 18 "C. An equal volume of 1 mM glycine in buffered saline was added, and the cells washed once with the glycine and three times in ice-cold buffered saline alone. Labeled cells were suspended in extraction buffer (15 mM Tris, 3 mM glycine, (pH 8 ) , 20 pg/ml chymostatin (Sigma), 5 mM DFP, 1 mM 2-mercaptoethanol, and 1% Triton X-100) at 0 "C for 10 min. After centrifugation, the extract supernatant was subjected to polyacrylamide gel electrophoresis (PAGE). Membranes before or after endo F cleavage (see below) were labeled using conditions identical to those for whole cells. However, following the addition of glycine to quench the linking reaction, the reaction mixture containing labeled membranes was layered over buffered saline containing 5% sucrose and centrifuged a t 100,000 X g for 40 min. The labeled membrane pellet was used in some endo F cleavage studies or directly subjected to PAGE analysis.
Endo F Treatment-Endo F was prepared from cultures of Flauobacterium meningosepticum (American Type Culture Collection, Rockville, MD) following the method of Elder and Alexander (16) with some modifications as previously described (27). The enzyme was stored in 0.1 M Na phosphate (pH 6.1), 50 mM EDTA, and 50% glycerol at -20 "C at a concentration of 1 unit/pl.
Cell membranes (100 pg of protein in 100 p1 of phosphate-buffered saline (pH 6.5)) were mixed with 30 pl of endo F and incubated 5 h a t 37 "C. Membranes were then affinity-labeled.
were then added 40 p1 of endo F. This mixture was incubated at 37 "C for 2-5 h.
In some studies, affinity-labeled formyl peptide receptor or formyl peptide receptor fragments were separated by PAGE, located by autoradiography, and cut from dried gels. The gel pieces were washed for 2 h in several changes of 20% methanol in water and then frozen and lyophilized overnight. The gel pieces were rehydrated with digestion buffer containing endo F as above and incubated a t 37 "C for 16-24 h.
Papain Treatment-Fifteen million cells were treated with papain a t 30 units/ml (Sigma) in 1.5 ml of buffered saline (pH 6.65) containing 1 mM dithiothreitol (Pierce Chemical Co.). After a 10-min incubation a t 37 "C, the cells were washed twice in 25 ml of cold buffered saline containing 0.1% lactalbumin and washed twice more in buffered saline alone before affinity labeling. With neutrophils or d-HL-60, there was little aggregation of cells and no detectable loss of cell viability during this treatment. The monocytes tolerated this treatment less well in that there was considerable cell aggregation and a variable decrease in the number of viable cells.
Polyacrylamide Gel Electrophoresis and Autoradiography-Materials to be analyzed by SDS-PAGE were solubilized in 6% SDS, 6 M urea, 10% 2-mercaptoethanol, 150 mM Tris, 4 mM EDTA (pH 6.9). Some solubilized samples were heated a t 100 "C for 60 s, while other samples were solubilized without heating. It should be noted that the affinity-labeled high molecular weight material above M, 100,000 seen in some of the studies was an artifact of heating the solubilized formyl peptide receptor before SDS-PAGE. This aggregate was greatly diminished when SDS-solubilized samples were not heated. Heating of SDS-solubilized whole cell extracts was necessary to eliminate proteolytic activity which persisted despite extensive use of protease inhibitors.
Solubilized material was analyzed on 10% polyacrylamide slab gels prepared and run according to the method of Laemmli (28), but with 3.5% stacking gels and containing 0.2% SDS. Red blood cell membrane ghosts (29) were used as molecular weight markers to calculate M, on gels.
Two-dimensional PAGE was carried out according to the method of O'Farrell(30) using pH 3-10 ampholine (Bio-Rad) in the isoelectric focusing (IEF) tube gels for the first separation. The focused proteins in the tube gel were separated by SDS-PAGE in slab gels as described above. Slab gels were fixed and stained with Coomassie Brilliant Blue. The gels were dried, and autoradiographs were obtained with Kodak X-AR-5 film using fluorescence intensifying screens and exposure a t -70 "C.

Heterogeneity of Formyl Peptide Receptor from Different
Cell Types-Plasma membrane-enriched fractions from neutrophils, monocytes, and d-HL-60 were affinity-labeled with formyl lZ51-peptide and EGS and subjected to two-dimensional PAGE analysis and autoradiography (Fig. 1). Formyl peptide receptor from neutrophils (Fig. la) migrated as two spots centered at PI 5.8 and 6.2 and having M , values of 55,000-70,000 and 58,000-70,000 respectively. Formyl peptide receptor from monocytes (Fig. l b ) consisted of two spots centered at PI 5.6 and 6.0 and having M , values of 60,000-75,000 and 62,000-75,000 respectively. Formyl peptide receptor from Me2S0 d-HL-60 (Fig. IC) resolved as two major spots centered at PI 5.6 and 6.0, with M , values of 62,000-85,000 and 64,000-85,000 respectively. With d-HL-60 membranes, a number of minor specifically labeled spots were also consistently detected. Three days of differentiation with Bt2cAMP resulted in greater expression of formyl peptide receptor by d-HL-60 than seen at 5 days with Me2S0, but affinity-labeled formyl peptide receptor expressed by cells differentiated with either inducer had identical patterns after two-dimensional PAGE. Uninduced HL-60 cells showed no specific labeling with formyl 1251-peptide and EGS (see Fig. 4). Labeling of formyl peptide receptor from all cell types was markedly diminished when nonradioactive formyl peptide was present at a 1000fold excess over the concentration of formyl '"I-peptide a t the time of cross-linking, indicating that the labeling pattern was specific (not shown).
When intact cells were affinity-labeled and extracts were subjected to two-dimensional PAGE analysis, results were similar to that seen with plasma membranes except that the two distinct isoelectric forms were less well resolved.
Structure and Function of Formyl Peptide Receptor after Endo F Cleavage-In order to examine the effect of deglycosylation upon affinity labeling, plasma membranes from neutrophils, monocytes, and d-HLSO were first treated with endo F for 5 h at 37 "C as described under "Materials and Methods." Treated and control (similarly incubated, but without endo F) membranes were then affinity-labeled and analyzed by SDS-PAGE as shown in Fig. 2  In other studies, formyl peptide receptor in plasma membranes was first affinity-labeled. The membranes were then solubilized in digestion buffer and subjected to endo F cleavage. The autoradiographic pattern after SDS-PAGE was identical to that seen when intact membranes were cleaved with endo F before affinity labeling as described above. These results suggest that endo F cleavage does not abolish ligand binding or affinity labeling. It also suggests that formyl peptide receptors from all three cell types have core polypeptides of similar size.
Heterogeneity of Neutrophil Formyl Peptide Receptor after Endo F Cleavage-Since the conditions used in the above studies did not result in cleavage of all the formyl peptide receptor by endo F, further studies were done to examine separately the carbohydrate contents and endo F cleavage profiles of the two PI forms of neutrophil formyl peptide receptor. Affinity-labeled neutrophil plasma membranes were solubilized in detergent, allowed to bind to wheat germ agglutinin linked to Sepharose, and eluted with N-acetylglucosamine as previously described (13). The autoradiographic pattern of the column eluate after two-dimensional PAGE (not shown) was identical to that seen in Fig. l a with unfractionated neutrophil membranes, demonstrating that both major PI forms of formyl peptide receptor from neutrophils were glycoproteins.
The endo F treatment of each PI form of neutrophil formyl peptide receptor under conditions of complete cleavage was examined. The two PI forms of affinity-labeled formyl peptide receptor from neutrophil membranes were separated by twodimensional PAGE and cut from the gel shown in Fig. 3a were subsequently analyzed by two-dimensional PAGE as shown in Fig. 3, b and c. Cleavage of the PI 5.8 and 6.2 neutrophil formyl peptide receptor component+ yields fragments of M, 33,000 only, but with distinctly different ranges of PI. Thus, PI differences of these components were not eliminated by endo F cleavage and are not likely the result of N-linked carbohydrate differences exclusively. In other studies, affinity-labeled neutrophil membranes were treated with alkaline phosphatase in Tris buffered saline (pH 10) for 5 h at 37 "C to see if differences in phosphorylation were responsible for the different PI forms of neutrophil formyl peptide receptor. This treatment had no effect upon the two-dimensional PAGE pattern (not shown). Effect of Tunicamycin upon Formyl Peptide Receptor Expression by d-HL-60"When tunicamycin was present during the 3 days of differentiation of HL-60 cells with 750 p~ BhcAMP, a dose-dependent decrease in the expression of formyl peptide binding was observed up to 0.5 pg/ml tunicamycin. Between 0.5 and 2 pg/ml tunicamycin, the receptor binding did not decrease further, remaining at about 25% of that seen without tunicamycin. At concentrations of tunicamycin up to 2 pg/ml, cell viability was unaffected and cell proliferation was about 75% of that seen without tunicamycin, a result consistent with previous observations (20). At all concentrations of tunicamycin tested, similar morphological differentiation had occurred, including decreased size and ability to adhere and migrate.
When HL-60 cells differentiated in the presence of tunicamycin were affinity-labeled to examine formyl peptide receptor, there was a marked dose-related effect upon the apparent molecular weight of the labeled receptor (Fig. 4). At 0.5 pg/ ml tunicamycin, the major product of affinity labeling is 38-48 kDa with a small amount of formyl peptide receptor of normal molecular weight. At 1 pg/ml tunicamycin, no formyl peptide receptor of normal molecular weight is seen, and there is equal labeling of a 38-48-kDa peptide and a 32-kDa peptide.

FIG. 3. Analysis of endo F treatment of the two isoelectric
forms of affinity-labeled formyl peptide receptor from neutrophil plasma membranes. Neutrophil plasma membranes were affinity-labeled with formyl '%I-peptide and EGS and subjected to two-dimensional (IEF X SDS-PAGE) gel analysis. Shown in a is the autoradiograph of this gel analysis. This autoradiograph was used to locate the two labeled isoelectric forms of receptor in the corresponding gel. The two portions of that gel containing the labeled material were cut out. These gel pieces were treated with endo F for 24 h at 37 "C and again subjected to two-dimensional gel analysis. Autoradiographs of the gel analysis of the endo F-treated PI 6.2 and 5.8 forms of the receptor are shown in b and c, respectively. analysis. Shown are the autoradiographs of the gel analysis of extracts from affnity-labeled uninduced HL-60 cells ( l a n e a), HL-60 cells differentiated in the absence of tunicamycin ( l a n e b), and HL-60 cells differentiated in the presence of 0.5 pg/ml tunicamycin ( l a n e c), 1 pg/ ml tunicamycin ( l a n e d ) , or 2 pg/ml tunicamycin ( l a n e e). ceptor after papain treatment of intact cells. Intact cells were incubated with papain for 10 min at 37 "C. Cells similarly incubated, but without papain, served as controls. Treated cells were labeled with formyl '251-peptide and EGS. Triton X-100 extracts of the labeled cells were subjected to SDS-PAGE analysis. Shown are autoradiographs of the gel analysis after labeling of neutrophils incubated without ( l a n e a) or with ( l a n e b) papain, monocytes incubated without (lane c) or with ( l a n e d ) papain and Me,SO-differentiated HL-60 cells incubated without ( l a n e e) or with ( l a n e f ) papain.
At 2 pg/ml tunicamycin, the 32-kDa peptide is the major product of affinity labeling. Thus, the results obtained with tunicamycin inhibition of asparagine-linked glycosylation of formyl peptide receptor during differentiation of HL-60 suggest that there are at least two N-linked oligosaccharide chains, that these oligosaccharides are not necessary for ligand binding, and that the core peptide is about 32 kDa. This is consistent with the results obtained with endo F cleavage.

Labeling of Formyl Peptide Receptor and Effect of Endo F after Papain
Cleavage-Since it has previously been shown that papain cleavage of intact neutrophils generates a membrane-bound formyl peptide receptor fragment that retains the ability to bind formyl peptides (15), the glycosylation of this fragment was examined in neutrophils, monocytes, and d-HL-60. Whole cells were cleaved with papain, washed, and then affinity-labeled. Triton X-100 extracts of these cells were analyzed by SDS-PAGE and autoradiography as shown in Fig. 5. With all three cell types, there was a decrease in affinity labeling of formyl peptide receptor by formyl Iz5Ipeptide compared to controls, and a new band of similar (though not identical) size was seen in extracts from all three cell types. This membrane-bound fragment of formyl peptide receptor from papain-treated neutrophils migrated at M, 31,000 and from papain-treated d-HL-60 at M , 29,000. With papain-treated monocytes, a MI 29,000 fragment was detected, but a prominent band at MI 25,000 was also seen. Since the monocytes did not tolerate the papain treatment as well as the neutrophils or d-HL-60, it is possible that the second band is an artifact of further cleavage of formyl peptide receptor in disrupted monocytes. It has been shown that cleavage of neutrophil plasma membranes by papain generates a smaller fragment of formyl peptide receptor than does cleavage of intact cells (15), presumably due to increased accessibility to a papain cleavage site on the endodomain of the receptor.

on Formyl Peptide Receptor 2513
When cells were first affinity-labeled with formyl 1251-peptide and EGS and then treated with papain, labeled fragments of formyl peptide receptor with M, identical to those seen in Fig. 5 were obtained.
The portions of the gel shown in Fig. 5 containing uncleaved receptor from control neutrophils and digested receptor from papain-treated neutrophils were cut out, treated with endo F for 16 h at 37 "C, and examined by SDS-PAGE and autoradiography (not shown). Under these conditions, the intact formyl peptide receptor is completely converted by endo F treatment to a band at M, 33,000 (as previously shown in Fig.   3). In contrast, under the same conditions, papain-cleaved neutrophil formyl peptide receptor is unaffected by endo F treatment, still migrating at M , 31,000, thus suggesting the absence of N-linked oligosaccharides on this fragment.

DISCUSSION
Addition of a specific high mannose oligosaccharide to selected asparagine residues of nascent polypeptides at the luminal surface of the rough endoplasmic reticulum appears to be a ubiquitous event in eucaryotic cells (31,32). Many polypeptide hormone receptors are integral membrane glycoproteins. These include insulin receptors (33), nerve growth factor receptors (34), epidermal growth factor receptors (35), and the cholecystokinin receptor in pancreas (27). Endo F has been used to show that the cholecystokinin-binding proteins (27) and 8-adrenergic receptors (36) contain N-linked oligosaccharides.
The further processing of N-linked oligosaccharide from high mannose to complex type occurs in the Golgi cisternae with cell type-specific differences in the final product (31,32).
Cell type differences in M, of 8-adrenergic receptors appear to be due to differences in N-glycosylation (36). Affinitylabeled insulin receptors from brain and from adipose tissue also differ in M , presumably due to differences in their oligosaccharide contents (37,38).
Tunicamycin prevents N-glycosylation of proteins by inhibiting the enzymes required for synthesis of the glycolipid precursor required for transfer of the high mannose oligosaccharide chain to the nascent polypeptide (17). Tunicamycin has been used to demonstrate N-glycosylation of a number of membrane receptors during synthesis including insulin receptor (39-41), acetylcholine receptor (42), and receptor for immunoglobulin E (43). Addition of N-linked oligosaccharide chains to the insulin receptor appears to be required for the expression of any insulin binding activity (39-41). Acetylcholine receptor and the receptor for immunoglobulin E do not require glycosylation for expression of ligand binding activity (42,43). Acetylcholine receptor lacking N-linked oligosaccharide chains appears to have the same affinity for ligand as glycosylated receptor but appears to be degraded more rapidly (42). Although it has previously been shown that neutrophil formyl peptide receptor binds to wheat germ agglutinin and is therefore a glycoprotein (13), the extent of this glycosylation, its requirement for ligand binding, and the differences in glycosylation of this receptor in different cell types were not known. In the present study, we have shown that affinitylabeled formyl peptide receptor from neutrophils, monocytes, and d-HL-60 migrates in two-dimensional PAGE over a broad range of molecular weight with two distinct isoelectric forms. The formyl peptide receptor polypeptide backbone is of similar molecular weight in these three cell types despite significant differences in M, of the native receptors. N-Linked oligosaccharides of the receptor contribute half or more of the apparent MI of the native receptor as determined by SDS-PAGE. Heterogeneity in N-linked oligosaccharides does not appear to be responsible for the two PI forms of receptor observed in the three cell types. Finally, these N-linked oligosaccharides appear to be localized to a very small peptide region which is at the opposite end from the membranebinding domain and does not include the peptide-binding site itself.
Endo F treatment of formyl peptide receptor from neutrophils and monocytes under conditions of incomplete cleavage generated at least one major intermediate cleavage product.
Intermediate products of endo F cleavage of formyl peptide receptor from d-HL-60 are more difficult to discern, although in some studies a t least one intermediate product can be seen.
Tunicamycin caused a dose-dependent decrease in M , of affinity-labeled formyl peptide receptor from HL-60 differentiating in the presence of Bt2cAMP. At 0.5 pg/ml tunicamycin, a M , 38,000-48,000 band was the major product seen after affinity labeling with formyl lZ5I-peptide and EGS, while at 2 pg/ml, the major product was seen a t M , 32,000. There is some suggestion that the M , 38,000-48,000 band may actually consist of two closely migrating components since the upper component of this band (see Fig. 4) shows a greater apparent decrease relative to the lower component at increasing concentrations of tunicamycin. These results are consistent with the data obtained with endo F cleavage, suggesting that formyl peptide receptor contains at least two and possibly three N-linked oligosaccharide chains on a M , 32,000 core polypeptide.
Papain cleavage of formyl peptide receptor in whole cells demonstrates that receptors from all three cell types studied have papain cleavage sites on the ectodomain of the polypeptide backbone. There are no N-linked oligosaccharide chains cleaved by endo F from the membrane-bound papain-generated formyl peptide receptor fragment, suggesting that the Nlinked oligosaccharide side chains are located on the approximately 1-2-kDa portion of the polypeptide backbone removed by papain cleavage. Neither papain, endo F, nor tunicamycin treatment prevented affinity labeling of the altered formyl peptide receptor. Thus, the ligand-binding site must lie between the intramembranous portion of the receptor and the papain cleavage site(s). Furthermore, the N-linked oligosaccharides are not required for ligand binding.
Further studies are necessary to determine the role of these oligosaccharides in targeting newly synthesized formyl peptide receptor to the correct cell compartment during differentiation (31, 32, 44, 45) and its role in degradation (42, 46) or recycling (47, 48) of this receptor after appearance at the cell surface.