Glycosylation sites in the atrial natriuretic peptide receptor

Oligosaccharide structures are not required for hormone binding


K. S. Misono, Department of Molecular Cardiology, Lerner Research Institute, NB50, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USA. Fax: + 216 444 9263, Tel.: + 216 444 2054, E-mail:


Atrial natriuretic peptide (ANP) is a hormone involved in cardiovascular homeostasis through its natriuretic and vasodilator actions. The ANP receptor that mediates these actions is a glycosylated transmembrane protein coupled to guanylate cyclase. The role of glycosylation in receptor signaling remains unresolved. In this study, we determined, by a combination of HPLC/MS and Edman sequencing, the glycosylation sites in the extracellular domain of ANP receptor (NPR-ECD) from rat expressed in COS-1 cells. HPLC/MS analysis of a tryptic digest of NPR-ECD identified five glycosylated peptide fragments, which were then sequenced by Edman degradation to determine the glycosylation sites. The data revealed Asn-linked glycosylation at five of six potential sites. The type of oligosaccharide structure attached at each site was deduced from the observed masses of the glycosylated peptides as follows: Asn13 (high-mannose), Asn180 (complex), Asn306 (complex), Asn347 (complex), and Asn395 (high-mannose and hybrid types). Glycosylation at Asn180 and Asn347 was partial. The role of glycosyl moieties in ANP binding was examined by enzymatic deglycosylation of NPR-ECD followed by binding assay. NPR-ECD deglycosylated with endoglycosidase F2 and endoglycosidase H retained ANP-binding activity and showed an affinity for ANP similar to that of untreated NPR-ECD. Endoglycosidase treatment of the full-length ANP receptor expressed in COS-1 cells also had no detectable effect on ANP binding. These results suggest that, although glycosylation may be required for folding and transport of the newly synthesized ANP receptor to the cell surface, the oligosaccharide moieties themselves are not involved in hormone binding.


atrial natriuretic peptide


extracellular domain of the ANP receptor


A-type natriuretic peptide receptor


B-type natriuretic peptide receptor


natriuretic peptide clearance receptor





Atrial natriuretic peptide (ANP) is a peptide hormone secreted from the atrium of the heart in response to blood volume expansion. It stimulates renal salt excretion, dilates blood vessels, and suppresses aldosterone and renin secretion, thus lowering blood pressure and volume [1]. These ANP actions are mediated by a family of transmembrane receptors coupled to guanylate cyclase, which include A-type and B-type natriuretic peptide receptor (NPR-A and NPR-B, respectively). NPR-A mediates most of the known actions of ANP and brain natriuretic peptide. NPR-B is thought to mediate the action of C-type natriuretic peptide in the central nervous system [2]. Both NPR-A and NPR-B receptors consist of a single polypeptide that contains an extracellular ANP-binding domain, a single transmembrane sequence, and an intracellular domain consisting of a kinase-homologous domain and a guanylate cyclase domain. Binding of the hormone to the extracellular domain activates the intracellular guanylate cyclase domain, generating cGMP as the intracellular second messenger. The mechanism of this guanylate cyclase activation by hormone binding remains largely unknown.

Both NPR-A and NPR-B receptors occur as a glycoprotein containing Asn-linked oligosaccharides [1,3–5]. However, neither the sites of glycosylation nor the structures of the attached oligosaccharides in these receptors have been identified by direct biochemical determinations. Stults et al. [2] have identified and characterized the N-glycosylation sites in the ANP clearance receptor (NPR-C), which is the receptor that is not coupled to guanylate cyclase and is thought to play a role in metabolic clearance of ANP from the circulation [3]. However, none of the glycosylation sites in NPR-C are conserved in NPR-A or NPR-B as N-glycosylation consensus sites [2,4,5]. Therefore, the state of glycosylation in the guanylate cyclase-coupled NPR-A and NPR-B receptors remains unknown.

Lowe and Fendly [4] studied the role of glycosylation in the function of NPR-A using 293 cells stably transfected with an NPR-A cDNA expression construct. These cells produced two receptor species: a 135-kDa protein representing a fully glycosylated and mature form of the receptor and a 125-kDa protein apparently representing a partially glycosylated immature form. Of these two species, only the 135-kDa protein could be affinity-cross-linked with 125I-ANP by a cross-linker. On the basis of these findings, these investigators have suggested that glycosylation of the receptor is required for the ANP-binding activity. Fenrick et al. [6] reported similar findings with the NPR-B receptor. They further examined the role of glycosylation by site-directed mutagenesis [7], the results of which showed that elimination of potential N-glycosylation sites in NPR-B by mutagenesis causes an almost complete loss of the binding activity expressed on the transfected cells. These studies together suggest that glycosylation is critical for the ligand-binding activities of NPR-A and NPR-B. It is possible that glycosylation is merely required for the proper folding of a newly synthesized receptor polypeptide and its transport to the cell surface, as is often the case in the biosynthesis of glycoproteins. Alternatively, it is possible that the glycosylated structure itself plays a role in ligand binding. To date, this distinction has not been clearly made.

HS-142-1, a novel polysaccharide of microbial origin, has been characterized as a specific antagonist for guanylate cyclase-coupled natriuretic peptide receptors [8–10]. The antagonist activity of the polysaccharide HS-142-1 suggests that its saccharide moiety may interact with the binding site and modulate hormone binding. It is therefore also possible that the oligosaccharide structure in the ANP receptor itself may either directly or indirectly be involved in ligand binding. Thus, clarification of the role of the oligosaccharide structure in ANP binding is necessary to understand the mechanism of ANP–receptor interaction as well as for designing ANP receptor agonist and antagonist compounds.

To facilitate structure–function analysis of the ANP receptor, we previously expressed the extracellular ligand-binding domain of NPR-A (NPR-ECD) in a soluble form and purified it to homogeneity by ANP affinity chromatography [11]. NPR-ECD binds ANP with an affinity and specificity similar to that of the naturally occurring full-length ANP receptor. It binds ANP with 1 : 1 stoichiometry, and this binding causes the formation of 2 : 2 complexes [11]. These findings show that the purified NPR-ECD retains the functional characteristics expected for the extracellular domain of the ANP receptor and is useful for structure and function studies. In this study, we determined the glycosylation sites in the receptor by direct biochemical analysis of the purified NPR-ECD. Furthermore, we examined the role of the glycosyl structure in ligand binding by enzymatic deglycosylation of NPR-ECD followed by ANP-binding assays. Our results suggest that the glycosyl structures in the ANP receptor are not involved in hormone binding.

Materials and methods


NPR-ECD was expressed in COS-1 cells and purified by ANP affinity chromatography as described [11]. Tos-Phe-CH2Cl-treated trypsin and Staphylococcus aureus V8 protease were purchased from Worthington Biochemical, and endoglycosidase F2 and endoglycosidase H were from Oxford Glyco Sciences ( Abingdon, Oxon, UK). Iodoacetic acid was obtained from Sigma and recrystallized from chloroform before use. Molecular-mass standards for SDS/PAGE were obtained from BioRad Laboratories. All other chemicals used were of analytical grade or the highest quality available.

Peptide mapping by HPLC/MS

NPR-ECD (500 pmol) in 2 m Tris/HCl buffer, pH 8.0, containing 5 m guanidine hydrochloride was reduced with 2-mercaptoethanol and allowed to react with iodoacetic acid as described by Crestfield et al. [12]. Reduced and S-carboxymethylated NPR-ECD was freed from the reagents using an ultracentrifuge filter tube (molecular weight 5000 cut-off ; Millipore, Bedford, MA, USA) by repeated concentration and dilution with 0.1 m ammonium bicarbonate, and then lyophilized. The material was digested with trypsin in 0.1 m ammonium bicarbonate at 37 °C for 14 h at a molar enzyme to substrate ratio of 1 : 100.

The digest was chromatographed on a Vydac C18 column (Separations Group, Hesperia, CA, USA; 2.1 × 250 mm) using a linear gradient of acetonitrile from 2% to 82% in water in the presence of 0.02% trifluoroacetic acid over a period of 80 min at a flow rate of 160 µL·min−1 in a Hewlett–Packard model 1100 HPLC system. The column effluent was split at a 9 : 1 ratio using a microsplitter valve (Upchurch Scientific, Oak Harbor, WA, USA), and one-tenth of the effluent was passed into a triple quadrupole mass spectrometer equipped with an electrospray ion source. The remainder of the effluent was collected manually while the mass spectra were simultaneously monitored. The total ion current was measured in the m/z range from 350 to 2400 at a 35-V cone voltage with a sweep time of 6 s in the positive ion mode. To identify glycopeptides in the effluent, two carbohydrate-specific marker ions (the oxonium ions of N-acetylhexosamine and N-acetylneuraminic acid with m/z 204 and 292, respectively [13,14]) generated after collision fragmentation at the ion source were monitored using a 200-ms dwell time for each ion at 180-V cone voltage in the positive ion mode. The electrospray voltage was 3500 V. The source temperature was set at 70 °C, the nebulizer gas flow was 10 L·h−1, and the drying gas flow was 300 L·h−1. The resolution was set to obtain sufficient separation of the isotope peaks of doubly charged ions. All MS analyses in this study were conducted using a Micromass Quattro II triple quadrupole MS (Manchester, UK) at the Cleveland Mass Spectrometry Facility, Department of Chemistry, Cleveland State University (OH, USA). MassLynx 3.02 software was used for instrument control, data acquisition, and data processing.

Further fragmentation of a tryptic peptide with S. aureus V8 protease was carried out in 0.1 m ammonium bicarbonate at 37 °C for 14 h at a molar enzyme to substrate ratio of 1 : 100.

Edman degradation of glycosylated peptide fragments

Edman sequence analysis was performed using an Applied Biosystems model 477 Protein Sequencer at the Macromolecular Structure Analysis Facility of the University of Kentucky, Lexington, KY, USA.

Expression of the ANP receptor in COS-1 cells

The full-length rat ANP receptor was expressed in COS-1 cells, and cell membrane fractions were prepared from transfected cells as described [11].

Endoglycosidase treatment

NPR-ECD (200 pmol) was deglycosylated by incubation with 3.2 mU endoglycosidase F2 in 100 µL 0.1 m sodium acetate buffer, pH 6.0, at 23 °C. At various time intervals, aliquots were taken and used for ANP-binding assay, photoaffinity labeling, and SDS/PAGE separation. Deglycosylation with endoglycosidase H was performed under the same conditions except that 8 mU endoglycosidase H was used. Deglycosylation by double digestion with endoglycosidase F2 and endoglycosidase H was carried out by incubating 200 pmol NPR-ECD with 3.2 mU and 8.0 mU of the enzymes, respectively, in 100 µL 0.1 m sodium acetate buffer, pH 6.0, at 23 °C for 15 h.

Deglycosylation of the full-length ANP receptor in the COS-1 cell membranes (320 µg protein) was carried out by incubating the membranes with 4 mU endoglycosidase F2 and 10 mU endoglycosidase H in 100 µL 0.1 m sodium acetate buffer, pH 6.0, at 23 °C for 15 h.

Assay of ANP-binding activity

NPR-ECD (0.5 pmol) was incubated with 125I-ANP(4–28) (100 000 c.p.m.) and 10 n m unmodified ANP(4–28) in 250 µL 50 m m Tris/HCl buffer (pH 7.5) containing 0.15 m NaCl, 0.1% BSA, and 0.05% bacitracin at room temperature for 30 min. A 200-µL aliquot of the mixture was then applied to a column of Sephadex G-50 (3 mL total gel-bed volume) equilibrated with 10 m m sodium phosphate buffer (pH 7.5) containing 0.15 m NaCl, 0.05% BSA, and 0.01% bacitracin. The first 1.3-mL fraction, which contained protein-bound 125I-ANP(4–28), was collected and counted in a γ-counter. Unbound 125I-ANP was eluted at a volume greater than 1.7 mL. The level of non-specific binding was determined by carrying out the incubation in the presence of a 100-fold excess (1 µm) of unmodified ANP(4–28).

ANP-binding activity associated with cell membrane preparations was measured as previously described [15].

SDS/PAGE analysis

NPR-ECD or endoglycosidase-treated NPR-ECD (10 pmol in 5 µL aliquot) was boiled for 3 min in SDS/PAGE sample buffer containing a reducing reagent and electrophoresed in a 10% polyacrylamide precast gel (NuPAGE Bis-Tris gel; Novex, San Diego, CA, USA) using the Mops buffer according to the manufacturer’s protocol. Molecular-mass standards were obtained from BioRad Laboratories. After electrophoresis, the gel was stained with Coomassie Blue R250.

Photoaffinity labeling

Photoaffinity labeling of the ANP receptor was carried out using N-azidobenzoyl-125I-ANP(4–28) (N3Bz-125I-ANP) as previously described [15]. After SDS/PAGE in a 7% polyacrylamide precast gel (NuPAGE Tris-Acetate gel; Novex), photolabeled bands were detected by autoradiography.

Electrospray MS of deglycosylated NPR-ECD

NPR-ECD treated with both endoglycosidase F2 and endoglycosidase H was applied to a Vydac C4 column (1.0 × 150 mm; Separations Group) and was eluted with a linear gradient of acetonitrile (from 2% to 82%) in the presence of 0.02% trifluoroacetic acid over 40 min at a flow rate of 40 µL·min−1 in an Applied Biosystems model 130A HPLC. The protein fraction was collected by monitoring absorption at 214 nm and then lyophilized. The material was dissolved in 0.025% formic acid in 50% acetonitrile and introduced directly into the mass spectrometer through a fused silica tube (100 µm internal diameter) at a flow rate of 2 µL·min−1. The spectrum was acquired in the mass range of m/z 800–1600 with a scan time of 6 s in the positive ion mode. Electrospray voltage and cone-voltage were 3500 V and 40 V, respectively. The ion source temperature was set at 70 °C. The nebulizer gas flow and drying gas flow were 10 L·h−1 and 250 L·h−1, respectively.


Determination of glycosylation sites

Glycosylation in NPR-ECD was characterized by HPLC/MS analysis of a tryptic digest of reduced and S-carboxymethylated NPR-ECD. The sequence of NPR-ECD contains six potential N-glycosylation sites (Asn13, Asn180, Asn306, Asn347, Asn354, and Asn395) which have the consensus sequence Asn-Xxx-Ser/Thr [16]. Based on the sequence of NPR-ECD, no more than one potential N-glycosylation site is expected to occur in a single tryptic fragment. RP-HPLC separation of the tryptic digest was monitored by total ion current scanning the m/z range from 350 to 2400 ( Fig. 1A). The separation was also monitored for two carbohydrate-specific marker ions: the oxonium ion with m/z at 204 derived from the N-acetylhexosamine moiety ( Fig. 1B) and the oxonium ion with m/z at 292 derived from N-acetylneuraminic acid ( Fig. 1C). Tryptic peptide fragments were assigned in the primary sequence of NPR-ECD on the basis of their molecular masses as summarized in Table 1 and Fig. 2. Selected-ion monitoring at m/z 204 for detection of N-acetylhexosamine identified five peaks at 27.6 min, 29.4 min, 30.3 min, 40.1 min, and 46.1 min ( Fig. 1B) which corresponded to the glycosylated forms of peptides T16 + 17 and T17, T33 + 34, T41, T1, and T35, respectively ( Table 2). Of these five peaks, four were also found by selected-ion monitoring at m/z 292, indicating the presence of N-acetylneuraminic acid ( Fig. 1C). The glycopeptide derived from T1, on the other hand, did not give the fragment ion with m/z 292, suggesting that it lacked N-acetylneuraminic acid. A peak at 35.9 min was observed in the m/z 292 scan but not the m/z 204 scan. The absence of m/z 204 ions from this peak suggests that the material did not contain oligosaccharide. The m/z 292 ion apparently represented the acyl ion of an internal fragment Tyr-Lys generated from the disintegration of peptide T29 + 30.

Figure 1.

HPLC/MS analysis of the tryptic digest of the reduced and S-carboxymethylated NPR-ECD. Tryptic peptide fragments are numbered starting from the N-terminus (see Fig. 2). (A) Total ion current from scanning over the m/z range from 350 to 2400. Glycosylated peptides are underlined. (B) and (C) Selective ion monitoring at m/z 204 and 292, respectively.

Table 1.  Molecular masses observed for the tryptic peptides from the reduced and S-carboxymethylated NPR-ECD. Peptides are numbered in order from the N-terminus of the protein as shown in Fig. 2. The average molecular-mass values were calculated on the basis of the amino-acid sequence deduced from the cDNA sequence. Peptides T17# and T35# are the unglycosylated forms of T17 and T35, respectively. ND, not determined; NS, not scanned (outside of the scanned mass range).
TrypticAmino acidRetentionMolecular mass (Da)
peptideresiduestime (min)CalculatedObserved
  • a

    Glycosylated peptides,

  • b 

    peptides containing S-carboxymethylcysteine,

  • c

    c peptides resulting from incomplete digestion by trypsin.

T1 a  1–2240.12476.8see Table 2
T2 23–3331.81095.31095.0
T3 34–35NS245.3NS
T4 36–4734.21380.61380.3
T5 b 48–7334.22548.92548.2
T6 b 74–9535.22358.72357.9
T7 96–10125.4816.9816.5
T8+9 c102–12537.22426.82426.4
T12+13 c143–15733.81827.11826.7
T14 b158–17441.92089.42088.7
T16+17 a,c177–19827.62637.8see Table 2
T17 a179–19827.62352.5see Table 2
T22+23+24 b,c207–24958.14850.64849.6
T23 b209–22032.51428.61428.0
T23+24 b,c209–24959.74637.44636.4
T29+30 c279–29535.92112.42111.9
T32+33 c299–30418.0686.9686.6
T33+34 a,c304–31429.41297.5see Table 2
T35 a315–35146.13943.4see Table 2
T39+40 c369–38938.42445.62445.6
T41 a390–40830.32119.4see Table 2
T43 b423–43523.31543.51543.4
Figure 2.

Amino-acid sequence of the NPR-ECD and fragments identified by HPLC/MS mapping. Peptides assigned in the MS are shown by horizontal lines. Glycosylated peptides are indicated by double lines. Peptides are numbered starting from the N-terminus. Cys60 and Cys86, Cys164 and Cys213, and Cys423 and Cys432 are disulfide bonded [18]. Glycosylated Asn13, Asn180, Asn306, Asn347 and Asn395 residues are shown in bold. Non-glycosylated Asn354 is shown underlined.

Table 2.  Oligosaccharide compositions and deduced oligosaccharide structures. The calculated molecular mass values include the masses of N-glycans.HexNAc, N-acetylhexosamine;Hex, hexose; DeoxyHex, deoxyhexose; NeuAc, N-acetylneuraminic acid.
Tryptic GlycosylationRetentionMolecular mass (Da)Deduced saccharideDeduced oligosaccharide
peptideResiduessitetime (min)CalculatedObservedcompositionstructure
T1  1–22Asn1340.13531.8
High mannose
High mannose
High mannose
High mannose
High mannose
Complex, biantennary
Complex, triantennary
Complex, biantennary
Complex, biantennary
Complex, triantennary
Complex, biantennary
Complex, biantennary
Complex, triantennary
Complex, triantennary
Complex, biantennary
Complex, triantennary
Complex, triantennary
Complex, biantennary
Complex, triantennary
Complex, biantennary
Complex, triantennary
Complex, biantennary
Complex, triantennary
Complex, triantennary
Complex, biantennary
Complex, triantennary
Complex, triantennary
High mannose
High mannose

The identities of the glycopeptides assigned by the HPLC/MS analysis above were confirmed by Edman sequencing. The material under each of the peaks eluted at 27.6 min (containing peptides T16 + 17 and T17), 29.4 min (peptide T33 + 34), 30.3 min (peptide T41), and 40.1 min (peptide T1) was sequenced directly by Edman degradation ( Table 3). The material under the peak at 46.1 min (peptide T35) was further digested with S. aureus V8 protease. The resulting fragments were separated by RP-HPLC to obtain peptides designated T35-SP1 (residues 315 through 334) and T35-SP2 (residues 335 through 351) (data not shown). Peptide T35-SP2 was sequenced by Edman degradation ( Table 3). All peptides were completely sequenced.

Table 3.  Edman sequencing of glycosylated tryptic peptides.
time (min)
Yield of the first
Edman cycle (pmol)
site identified
  • a

    Retention time of parent peptide T35,

  • b

    b coeluted non-glycosylated peptides.

27.6T16+T17177–198ERL XITVNHQEFVEGDPDHYPK 62.8Asn180
T25 b250–262SAQGLVPQKPWER 71.6 
29.4T33+34304–314KF XFTVEDGLK 130.8Asn306
46.1 aT35-SP2335–351TLAQGGTVTDGE XITQR 13.5Asn347
30.3T41390–408VVLNY XGTSQELMAVSEHK 65.6Asn395
T11 b133–142LGDFVTALHR160.5 
T37 b356–365SFQGVTGYLK 45.1 

Edman degradation of the fraction T1 yielded the amino-acid sequence from residue 1 to 22, in which degradation cycle 13 yielded no detectable phenylthiohydantoin (PTH) amino acid, confirming N-glycosylation at residue Asn13. Edman degradation of the fraction containing peptides T16 + 17 and T17 gave two tandem sequences from residues 177–198 and residues 179–198. In both sequences, degradation cycles corresponding to Asn180 yielded no detectable PTH-Asn, confirming glycosylation at position 180. Edman degradation of peptide T33 + 34 gave the sequence residues 304–314, in which glycosylation at residue Asn306 was confirmed by the absence of PTH-Asn at the corresponding cycle. Similarly, Edman degradations of peptide T35-V2 gave a sequence from residues 335–351, with N-glycosylation being confirmed at Asn347. Degradation of peptide T41 gave the sequence from residue 390 to 408, with N-glycosylation at Asn395. These results allowed us to unequivocally assign five N-glycosylation sites in NPR-ECD at positions Asn13, Asn180, Asn306, Asn347, and Asn395.

As shown in Table 3, glycosylated peptides T16 + 17 and T17 were coeluted with peptide T25, and glycosylated peptide T41 was coeluted with peptides T11 and T37. Sequences of these coeluted peptides were seen in Edman degradation. However, the presence of these peptides did not interfere with the assignment of the sequences or the glycosylation sites.

Peptide peaks designated T17# and T35# ( Fig. 1A) had molecular masses that corresponded to the non-glycosylated form of peptide T17 and T35, respectively ( Table 1). This result indicates that glycosylation at residues Asn180 and Asn347 was partial. Residue Asn354 in peptide T36 forms an N-glycosylation consensus sequence. However, only the non-glycosylated T36 peptide was observed in the HPLC/MS analysis, suggesting that Asn354 is not glycosylated. Approximately 95% of the total sequence of NPR-ECD was accounted for by the HPLC/MS mapping ( Fig. 2). The remaining 5% of the sequence was contained in small tryptic peptides, ranging from two to six amino-acid residues in size. Such peptides either had molecular masses outside of the scan range or were eluted near the pass-through fraction and were thus not analyzed by MS.

Oligosaccharide structure

The sugar composition of the oligosaccharide attached at each glycosylation site was deduced from the difference between the observed mass and the theoretical mass of a corresponding non-glycosylated peptide, as summarized in Table 2. From the estimated carbohydrate composition, the types of oligosaccharide structures were predicted. Multiple forms of glycosyl structure were found at each site, reflecting microheterogeneity in glycosylation. Relative amounts of different glycosyl forms present at each site were estimated from the ion intensity distribution observed in MS as summarized in Table 4.

Table 4.  Relative abundance of the oligosaccharide structures at each glycosylation site found in NPR-ECD. Relative amounts were estimated from the intensity of electrospray positive ions.
amount (%)
Asn180 64Complex-biantennary
Asn306 50Complex-biantennary
Asn347 60Complex-biantennary
Asn395 65High-mannose

Role of oligosaccharide structure in ANP binding

The role of oligosaccharide structures in ligand binding was examined by treating NPR-ECD with endoglycosidase and examining its effect on ANP-binding activity. Deglycosylation was followed by SDS/PAGE with Coomassie Blue staining. Photoaffinity labeling was also performed to detect the deglycosylated protein species that retained the ANP-binding activity. In some cases, deglycosylation was also verified by electrospray MS.

Treatment of NPR-ECD with endoglycosidase F2 had no significant effect on ANP-binding activity after the 7-h incubation period ( Fig. 3A). This treatment converted the 60-kDa NPR-ECD band into two bands (56-kDa and 54-kDa bands) which were discernible in SDS/PAGE after staining with Coomassie Blue ( Fig. 3B), confirming deglycosylation by the enzyme. Endoglycosidase F2 preferentially cleaves biantennary complex oligosaccharides [1], while triantennary complex oligosaccharides are resistant to endoglycosidase F2. As presented in Table 2, both biantennary and triantennary complex oligosaccharides occur at positions Asn180, Asn306, and Asn347 in varying relative contents. Therefore, the endoglycosidase F2 treatment is expected to yield a mixture of NPR-ECD polypeptides containing triantennary oligosaccharides in various amounts and positions. Therefore, it is likely that the 56-kDa and 54-kDa bands (as well as other additional diffused bands) in SDS/PAGE ( Fig. 3B) represent partially deglycosylated molecules. Similar changes could be seen in the bands detected by photoaffinity labeling, where untreated NPR-ECD at the 61-kDa position was converted into 58-kDa and 56-kDa bands ( Fig. 3C). These photoaffinity-labeled bands migrated more slowly than the corresponding Coomassie-stained bands. The lower mobility of the photoaffinity-labeled bands is due presumably to the addition of the covalently linked ANP affinity reagent moiety, which has a mass of ≈ 3 kDa and contains four positively charged residues (4 Arg residues). Affinity labeling observed with the deglycosylated 58-kDa and 56-kDa bands indicates that NPR-ECD deglycosylated by endoglycosidase F2 retained ANP-binding activity and therefore was radiolabeled by the affinity reagent.

Figure 3.

Effects of endoglycosidase F2 and endoglycosidase H treatments on the ANP-binding activity. NPR-ECD was treated with endoglycosidase F2 (A, B, and C) or endoglycosidase H (D, E, and F) for various incubation times. Aliquots were used for the assay of ANP-binding activity (A and D), SDS/PAGE and Coomassie Blue staining (B and E), and photoaffinity labeling by N3Bz-125I-ANP (C and F).

Similarly, treatment with endoglycosidase H did not cause a significant loss in ANP-binding activity ( Fig. 3D), while both Coomassie Blue-stained bands ( Fig. 3E) and photoaffinity-labeled bands for NPR-ECD ( Fig. 3F) were shifted to lower positions by the treatment. Endoglycosidase H is expected to cleave both high-mannose and hybrid-type oligosaccharides [17] attached at residues Asn13 and Asn395. Consistently, the material obtained after endoglycosidase H treatment was less heterogeneous than that obtained after endoglycosidase F2 treatment.

To examine further the role of oligosaccharide structure, we treated NPR-ECD with both endoglycosidase F2 and endoglycosidase H simultaneously. Most of the ANP-binding activity was retained after 7 h ( Fig. 4A). This treatment yielded a 52-kDa band, a less intense 56-kDa band, and defused bands with sizes greater than 56 kDa ( Fig. 4B). The most intense 52-kDa protein may represent NPR-ECD fully deglycosylated by both endoglycosidase F2 and endoglycosidase H. The 56-kDa band and the larger defused bands may represent a mixture of partially deglycosylated species that retain residual triantennary oligosaccharide moieties at Asn180, Asn306, and Asn347. The structure of the deglycosylated NPR-ECD was also confirmed by electrospray MS ( Fig. 4C). The observed molecular mass of 49 831 Da was in close agreement with the calculated molecular mass of NPR-ECD deglycosylated by cleavage at all chitobiose linkages by endoglycosidase F2 and endoglycosidase H, leaving the sugar moieties N-acetylglucosamine attached at Asn13 and Asn395 residues and fucosyl-N-acetylglucosamine attached at Asn180, Asn306, and Asn347 residues (calculated mass 49 828.5 Da) ( Fig. 4D). Partially deglycosylated species, such as those with triantennary complex oligosaccharides attached at positions Asn180, Asn306, and Asn347, were not evident in the MS, presumably because they are less efficiently ionized, heterogeneous, and less abundant individually.

Figure 4.

Effect of combined endoglycosidase F2 and endoglycosidase H treatment of the NPR-ECD on binding activity. NPR-ECD was treated simultaneously with both endoglycosidase F2 and endoglycosidase H. (A) Effect on ANP-binding activity. (B) Photoaffinity labeling by N3Bz-125I-ANP before (lane 1) and after (lane 2) the treatment. (C) Electrospray mass spectrum of NPR-ECD after the treatment. (D) Schematic presentation of the predicted NPR-ECD structure after deglycosylation by endoglycosidase F2 and endoglycosidase H.

Competitive binding assays with the untreated and deglycosylated NPR-ECD gave similar Kd values at ≈ 2 n m( Fig. 5), indicating that deglycosylation by combined endoglycosidase F2 and endoglycosidase H treatment did not significantly alter the binding affinity of NPR-ECD for ANP.

Figure 5.

Competitive binding assays with untreated and endoglycosidase-treated NPR-ECD. (○) Untreated NPR-ECD; (●) NPR-ECD treated with endoglycosidase F2 and endoglycosidase H.

The effect of deglycosylation on ANP binding was also tested with the full-length ANP receptor expressed in COS-1 cell membranes. The partially purified COS-1 cell membranes were treated with endoglycosidase F2 and endoglycosidase H under non-denaturing conditions. After incubation, the membranes retained 91% of the ANP-binding activity. Photoaffinity labeling experiments showed that the 130-kDa band of the ANP receptor in the untreated membranes was mostly converted into a 120-kDa band by the glycosidase treatment ( Fig. 6). These results suggest that the full-length ANP receptor deglycosylated by endoglycosidase F2 and endoglycosidase H retained ANP-binding activity.

Figure 6.

Effect of combined endoglycosidase F2 and endoglycosidase H treatment on the ANP-binding activity of the full-length receptor. Untreated (lanes 1 and 2) and treated membranes (lanes 3 and 4) were photoaffinity labeled by N3Bz-125I-ANP. Lanes 1 and 3 show non-specific labeling obtained in the presence of 0.1 µm unmodified ANP.


The glycosylation sites in the extracellular hormone-binding domain of the ANP receptor (type-A) have been determined using the recombinant NPR-ECD protein expressed in COS-1 cells and purified by ANP affinity chromatography. NPR-ECD, which binds ANP with an affinity and specificity comparable with that of the naturally occurring full-length ANP receptor, contains ≈ 16% (w/w) carbohydrate [11]. By a combination of HPLC/MS peptide mapping and Edman degradation, the sites of glycosylation in NPR-ECD were determined, and the type of the oligosaccharide structures attached at each glycosylation site was inferred from the observed fragment masses as follows: Asn13 (high-mannose type), Asn180 (biantennary complex type and triantennary complex type), Asn306 (biantennary complex type and triantennary complex type), Asn347 (biantennary complex type and triantennary complex type), and Asn395 (high-mannose type and hybrid type). One potential N-glycosylation site (Asn354) was not glycosylated. The HPLC/MS data for a tryptic digest of NPR-ECD in this study, when combined with the data for a lysylendopeptidase digest previously reported for the assignment of disulfide bonds [18], accounted for the entire 435-amino-acid sequence of NPR-ECD. Analysis of the combined data revealed no post-translational modification other than N-glycosylation and disulfide bonds.

The five N-glycosylation sites identified in NPR-ECD (from rat) are conserved in the NPR-A family of receptors from various species as the consensus sites for N-glycosylation, suggesting an important role for glycosylation. Site-directed mutagenesis studies by Fenrick et al. [7] suggest that bovine NPR-B is N-glycosylated at five of seven potential sites. Of these five putative glycosylation sites in NPR-B, two (Asn2 and Asn173) occur at the sites similar to those in NPR-A (corresponding to Asn13 and Asn180, respectively), suggesting a level of similarity in the pattern of glycosylation between NPR-A and NPR-B receptor families. On the other hand, glycosylation sites in the clearance receptor NPR-C [2] occur at positions entirely different from those in NPR-A and NPR-B receptors, suggesting that the polypeptide folding in NPR-C may differ considerably from that in NPR-A and NPR-B receptors.

Utilizing the information on the structures of the attached oligosaccharides, we examined whether the oligosaccharide structures in the receptor are involved in ligand binding. NPR-ECD was treated with endoglycosidase F2, endoglycosidase H, or both in combination, and the effect of the treatment on ANP-binding activity was measured. The extent of deglycosylation by the endoglycosidase treatment was monitored by SDS/PAGE with Coomassie Blue staining, photoaffinity labeling, and, in one instance, by molecular-mass determination by electrospray MS. We found that NPR-ECD treated with endoglycosidase F2, endoglycosidase H, or both retains nearly all of the ANP-binding activity, while significant deglycosylation was clearly evident. In addition, the binding affinity of NPR-ECD for ANP was not altered by the treatment with both endoglycosidase F2 and endoglycosidase H. These results strongly suggest that the oligosaccharide structures in the ANP receptor are not involved in hormone binding.

The effect of deglycosylation was also examined with the full-length ANP receptor expressed in COS-1 cell membranes. The full-length receptor deglycosylated by endoglycosidase F2 and endoglycosidase H in combination retained the ANP-binding activity. In our earlier studies using bovine adrenal cortex plasma membranes, we found that the ANP receptor deglycosylated by endoglycosidase F2 could still be photoaffinity-labeled by N3Bz-125I-ANP [19], suggesting also that glycosyl structures are not involved in ANP binding. The conclusion that oligosaccharide structures are not required for ANP binding is consistent with the three-dimensional structure of the apoNPR-ECD dimer that we have recently determined by X-ray crystallography [20]. In the crystal structure of the apoNPR-ECD dimer, the five N-glycosylation sites identified here are positioned away from the putative ANP-binding site identified by affinity labeling and site-directed mutagenesis.

Lowe and Fendly [4] report that the ANP receptor expressed in 293 cells, when metabolically labeled with 35S and isolated by immunoprecipitation, yields two 35S-labeled protein bands at 135 kDa and 125 kDa. Of these two bands, only the 135-kDa band representing a fully glycosylated mature form of the receptor could be affinity-cross-linked with 125I-ANP by a cross-linker, suggesting that glycosylation of the receptor is necessary for the ANP-binding activity.

Fenrick et al. [7] examined the effect of glycosylation in NPR-B by eliminating potential N-glycosylation sites by site-directed mutagenesis and measuring levels of ANP-binding activity expressed on the transfected COS cells. Elimination of potential glycosylation sites resulted in a marked loss of ANP-binding activity, indicating that glycosylation is essential for the ANP binding activity of the expressed protein. On the other hand, the results of the present study strongly suggest that the oligosaccharide structures are not involved in ANP binding. Therefore, it appears that glycosylation is required for proper folding of a newly synthesized receptor polypeptide and/or its transport to the cell surface and that, once properly folded, the receptor does not utilize the glycosyl structure for ligand binding. In addition, a recent report found that elimination of potential N-glycosylation sites in the heat-stable enterotoxin receptor by site-directed mutagenesis renders the protein more readily denatured by urea [21]. Therefore, it is possible that glycosylation may also contribute to the stability of the folded protein.

Koller et al. [5] found that only the fully glycosylated form of the receptor binds ANP and is also phosphorylated. Phosphorylation of the receptor is essential for signaling, and dephosphorylation leads to receptor desensitization [22,23]. It appears then that proper folding of the N-terminal extracellular domain of the receptor, which is dependent on glycosylation, is required for formation of the intracellular domain into its functional and phosphorylated structure.

The purified recombinant NPR-ECD binds ANP with an affinity and specificity similar to that of the native full-length ANP receptor [11], suggesting that the NPR-ECD polypeptide is folded into the form that reflects the structure of the native ANP receptor. Because glycosylation is apparently necessary for correct folding of the ANP receptor to a form with proper ANP-binding activity, it is likely that the glycosylation sites in NPR-ECD determined here reflect those occurring in the native ANP receptor. On the other hand, the type of oligosaccharides found in NPR-ECD (both complex-type and high-mannose-type oligosaccharides as deduced in this study) are considerably different from those in the native ANP receptor from bovine aorta and bovine adrenal plasma membranes which contain mostly complex-type oligosaccharides [19]. The presence of high-mannose oligosaccharides in NPR-ECD suggests that the expressed protein may not be fully mature with respect to the processing of the oligosaccharide structure during biosynthesis. This difference in the oligosaccharide structure may have resulted from rapid synthesis of NPR-ECD by the transfected COS-1 cells.

To our knowledge, this study represents the first direct biochemical determination of glycosylation sites for any guanylate cyclase-coupled receptor. The absence of the effect of endoglycosidase treatment on ANP binding suggests that the oligosaccharides in the receptor are not involved in ligand binding. We propose that glycosylation plays a role in folding of the nascent receptor polypeptide and its transport to the cell surface and that, once the receptor is properly folded, the oligosaccharide structures are not required for hormone binding.


This work was supported by grants from the National Institutes of Health (HL54329) and the American Heart Association National Center (no. 95012310). We thank Amy Moore for editorial assistance and Robin Lewis for her assistance in preparing the manuscript.


  1. Enzymes: atrial natriuretic peptide receptor type-A (guanylate cyclase-A; accession number P18910; EC; endoglycosidase H (EC; endoglycosidase F2 (EC