- Email: [email protected]
Identification of rat serotonin 5-HT2C receptors as glycoproteins containing N-linked oligosaccharides
MOLECULAR P,ESEARCH ELSEVIER
Molecular Brain Rcscarch 33 1t995) 31 t--318
Research report
Identification of rat serotonin 5-HT2c receptors as glycoproteins containing N-linked oligosaccharides Jon R. Backstrom *, Ryan S. Westphat, Herv , Canton, Elaine Sanders-Bush Department of Pharmacology, 432 Medical Research Building, Vanderbilt Unitersi~' School of Medicine, Nashtille. TN 37232-6600, USA
Accepted 21 June 1995
Abstrad Anlibodies against a portion of the rat 5-HT x. receptor third intracellular loop were generated and used to identify receptors soiubiiized from cell lines and rat brain. Western blots of CtLa,PS-soluble proteins were probed with affinity-purified anti-2C antibodies. I'he specificity of anti-2C was demonstrated with extracts prepared from NIH/3T3 fibroblasts which stably express functional rat 5-HT2c or 5-HTzA receptors. Extracts from the 5-HT2c cell line, but not the 5-HT2A cell line, contained immunoreactive proteins with masses of 51-52 kDa and 5 8 - 6 8 kDa. In the brain, immunoreactive proleins were identified from choroid plexus extracts with masses of 51 kDa and 58-62 kDa. The major 58-62 kDa and minor 51 kDa proteins were not detected in extracts prepared from the hippocampus, striatum, or frontal cortex using the same amount of CHAPS-soluble protein. These results are consistent with previous studies demonstrating that 5-HT2c receptor binding sites and mRNA are most abundant in choroid plexus. The association of asparagine-linked (N-link~,:d) oligosaccharides with the receptors was examined next. The 5-HT2c receptor cell line (3T3/2C) was grown in the presence of tunicamycin to metabolically inhibit N-linked glycosyiation. Proteins from the cell extracts were detected with masses of 40 and 41 kDa. Extracts prepared from 3 T 3 / 2 C cells (grown in the absence of tunicamycin) and from choroid plexus were incubated with N-glycosidase F to enzymatically remove available N-linked sugars. Immunoreactive proteins were detected with masses of 41 and 42 kDa from 3T3/2C cells and 41 kDa from choroid plexus. Neuraminidase, which cleaves sialic acid (N-acctylneuraminic acid) residues from glycoproteins, reduced the mass of the 51 and 58-62 kDa proteins from the choroid plexus to 50 and 54-58 kDa. In contrast, the 51-52 and 58-68 kDa proteins from 3T3/2C cells were not affected by treatment with neuraminidase. These results demonstrate that 5-H'I.,(. receptors contain N-linked sugars and suggest that sialic acid residues associate with 5-HT_,c receptors in the chc,roid plexus. The oligosaccharide moieties, which contribute up to "-- 30% of the relative mass as judged by SDS-polyacrylamide gel electrophoresis, ma3 impart functional properties to 5-HTzc receptors. Keywords: 5-HT2c receptors; Glycosylation; Sialic acid; Choroid plexus; Serotonin; Antibodies
1. Introduction The 5-hydroxytryptamin%c (5-HT2c) receptor is a member of the 5-HT 2 family of receptors that are coupled to phosphatidylinositol turnover [5]. The 5-HT2c receptor is also coupled to cyclic guanosine monophosphate (cGMP) formation [12]. Within the 5-HT 2 family, rodent 2A, 2B. and 2C receptors are approximately 50% identical at the amino acid level (reviewed in [3]). However, unique re-
Abbreviations: 5-HT, 5-hydroxytryptamine (serotonin); CHAPS, 3[(3-cholamidopropyi)dimethylammonio]- 1-propanesulfonate; SDS, sodium dodecyi sulfate; BSA, bovine serum albumin " Correspondinp, author. Fax: (1) (6i5) 343-6532; E-mail: BackstJR@ctrvax.'~anderbilt.edu0169-328X/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSD! 6169-328X(95 )00156-5
gions are present in the putative extracellular amino terminus, the third intracellular loop, and the intracellular cytoplasmic tail. These unique regions may be involved in protein-protein interactions. The distribution and relative abundance of 5-HT2c receptor mRNA in rat brain has been described [8,10,14,15]. Epithelial cells of the choroid plexus contained the highest density of receptor mRNA using in situ hybridization techniques (reviewed in [3]). Intermediate levels of mRNA were found in the hippocampus, amygdala, frontal cortex, caudate putamen, lateral habenula, subthalamic nuclei, substantia nigra (pars compacta), and regions of the olfactory system and hypothalamus. The cerebellum contained the lowest density of 5-HT2c receptor mRNA in the brain [8]. The levels of 5-HTzc receptor mRNA relative to total mRNA were investigated using Northern blots [10]. The
312
J.R. Backstrom et al. / Molecular Brain Re,search 33 (1995) 311-318
choroid plexus contained the highest amount of receptor mRNA. The second highest amount was detected in the hippocampus, which was at least 10-fold less abundant than in the choroid plexus [10]. Receptor mRNA was even less abundant in the basal ganglia, hypothalamus, ponsmedulla, and olfactory bulb. Also consistent with the in situ results, receptor mRNA was not detected in the cerebellum. Several studies have compared the sites of rat 5-HT, c receptor mRNA accumulation with sites of radioligand binding (reviewed in [9]). Binding sites of the 5-HT2c receptor are localized with radiolabeled ligands such as 5-HT (serotonin), mesulergine, and LSD. Although most brain regions which contain mRNA also bind radioligand, there appear to be some discrepancies. For examrle, the density of [3H]mesulergine sites in the lateral habenula and the subthalamic nuclei was lower than expected from., the corresponding density of in situ hybridization sites, suggesting that synthesis of the receptor may be regulated in a cell-specific manner [13]. Furthermore, down-regulation of 5-HT2c receptor binding sites by drug treatment has been shown to be independent of changes in mRNA levels [1]. Biochemical properties of 5-HT2c receptors have not been established due to the lack of specific high affinity probes. Here, we generate antibodies against a domain of the rat 5-HT2c receptor and utilize them to characterize receptors synthesized in vivo. Western blots of membrane extracts were probed with affinity-purified antibodies. First, the masses of immunoreactive receptors were determined from extracts of cells that stably express cDNA for the rat 5-HT2c receptor [2,18]. Second, the presence of immunoreactive receptor was examined in several regions of rat brain previously analyzed for tb.e relative abundance of 5-HT2c receptor mRNA using Northern blots [10]. Third, the presence of asparagine-linked (N-linked) oligosaccharides attached to the protein moiety was examined using two strategies. In one set of experiments, the 5-HT2c receptor cell line was cultured with the antibiotic tunicamycin, which inhibits the biosynthetic addition of N-linked sugars (reviewed in [6]). In the other set of experiments, extracts prepared from the 5-HT2c receptor cell line and the choroid plexus were treated with N-glycosidase F to digest N-linked oligosaccharides and with neuraminidase to digest terminal sialic acid residues. Evidence is presented for the presence of N-linked sugars associated with 5-HT2c receptors in the rat brain.
2. Materials and methods
•N H 3
Membrane -~87v
~
J,.,,~~
intrace,lular
2B Peptide: AHr.DOSHKDK• ~PNSTDZT 2A Peptide: SL$SEKLFQIRSiFIREFGS¥
COO
Fig. 1. Schematic representation of rat 5-HT2c receptor (adapted from [10]). The 5-HTzc receptor peptide (2C peptide) used to generate anli-2C antibodies corresponds to a domain of the putative third intracellular loop. Also shown are sequences of the similar domains from the related 5-HT2B and 5-HTzA receptors (2B and 2A peptide,~; respecdvcly). Asparagine residues that fulfill the N-linked consensus sequence asn-Y.-ser/thr are denoted with an arrowhead; numbers indicate their amino ack" positions.
tion (Rockville, MD). The 5-HT:A receptor cell line GF6, referred to as 3T3/2A, was a kind gift of Dr. David Julius, University of California at S.~n Francisco [11]. These cells express 5-HT2A receptors at 3 pmol/mg protein [7]. The 5-HT2c receptor cell line B3-1G, referred to as 3T3/2C, expresses 5-HT2c receptors at 5 pmol/mg protein [i8]. Cells were grown in Dulbecco's modified Eagle's medium (DMEM) containing 9% fetal bovine serum, 5 U/ml penicillin, and 5 /zg/ml streptomycin. The serum was obtained from Hyclone Laboratories (Logan, UT) and the penicillin and streptomycin were from Gibco/BRL Life Technologies (Grand Island, NY).
2.2. Generation of 5-HT2c receptor antibodies The 5-HTEc peptide (NH 2-CKKNGGEEENAPNPNPDQK-COOH) corresponds to amino acids 270-288 of the rat [10] and mouse [19] sequences. The peptide is located in the putative third intracellular loop [10] and has no significant homology to the corresponding regions of the rat 5-HTEA and 5-HT2B receptors (Fig. 1). The 5-HT2c peptide was synthesized, coupled to a carrier protein, and used to immunize rabbits and purify specific antibodies. The peptide was conjugated to keyhole limpet hemocyanin (KLH) and 1 mg suspended in Freund's complete adjuvant and injected s.c. in New Zealand rabbits (East Acres, Southbridge, MA). Antiserum was collected one week after each boost injection of 0.5 mg KLH-peptide prepared in Freund's incomplete adjuvant.
2.1. Cell cultures 2.3. Purification of 5-HT2c receptor antibodies Mouse NIH/3T3 fibroblasts and stable clones derived from transfection of the fibroblasts with rat 5-HT2c or 5-HT2A receptor cDNA were used for this study. Fibroblasts were obtained from American Type Culture Collec-
Anti-2C antibodies were purified from immune serum using a three step procedure. Serum was passed sequentially through a column containing BSA and then a column
J.R. Backstrom et al. / Molecular Bram Research 33 (1995~ 311-318
containing BSA coupled to the 2C peptide (BSA-2C peptide). In the third step, proteins eluted from the BSApeptide column were incubated with blots containing soluble proteins extracted from NIH/3T3 fibrobiasts. To construct the columns for the first two purification steps. BSA or BSA-2C peptide was added to an activated matrix (Reacti-gei 6x, Pierce Chemical Co., Rockford, IL) at a concentration of 2 mg protein per ml of gel. Excess reactive sites of the matrices were blocked with I M Tris, pH 9.0 (neutralization buffer). Two column~ containing i ml of immobilized BSA or BSA-2C peptide were washed extensively with TBS (50 mM Tris, pHT.6 containing 150 mM NaCI, 0.05% NaN 3) and elution buffer (50 mM glycine, pH 2.7 containing 0.05% NaN 3) before being equilibrated in_ TBS. Immune serum was thawed, centrifuged, and diluted with an equal volume of TBS. 10-15 ml of the diluted serum was passed through the BSA column at room temperature, collected, and applied to the BSA-peptide column. The BSA-peptide column was washed with ~ 25 ml TBS and bound proteins were eluted from the column with 10 mt of elution buffer. The material was collected in 15 ml tubes containing 0.3 ml of neutralization buffer. Protein concentrations were determined with the BCA protein assay (Pierce) using bovine serum albumin (BSA) as the protein standard. 2.5 ml of 5% BSA and 0.6 M NaCI in TBS was added to the antibody solution to make final concentrations of 1% BSA and 0.15 M NaCt. The final step in the purification procedure was employed to remove non-specific antibodies from the antibody solution that recognized soluble fibroblast proteins. The eluate from the BSA-peptide column (antibody solution) was incubated with a Western blot prepared from Tris-soluble extracts of NIH/3T3 cells. (The cross-reactive proteins were not detected in CHAPS- or SDS-soluble membrane fractions). Briefly, 2.0 mg Tris-soluble protein (see Section 2.5) was electrophoresed in a 12.5 ( h ) x 14 (w) X 1.5 cm preparatory gel and the regions corresponding to ~ 50-200 kDa were blotted to nitrocellulose (Hoefer Scientific Instruments). The membrane was blocked with BSA, washed sequentially with glycine elution buffer and blot buffer, and then equilibrated in blot buffer. The blot was incubated with the antibody solution obtained from the BSA-2C peptide column. After 0.5 h, the solution was collected and the presence of cross-reactive antibodies was monitored with Western blots prepared from soluble 3T3 proteins. The purified antibodies, refe.n-.ed to as anti-2C, were stored at 4°C. Initial specificity of anti-2C was screened cytochemically with NIH/3T3 fibroblasts and clones stably expressing 5-HT2A eDNA (3T3/2A) or 5-HT2c eDNA (3T3/2C). Cells were plated in wells containing type I collagen (Sigma). Sub-confluent cultures were washed twice with phosphate buffered saline (PBS; 15 mM phosphate, pH 7.4 containing 140 mM NaCI, 0.05% NaN 3} and fixed for 0.5 h with freshly prepared 4% formaldehyde, 4% sucrose in PBS. The cells were treated with 50 mM NH4CI in PBS
313
for 5 rain and then permeabiiized fer 5 rain with 0.2% Triton X-i00 in PBS containing 1 mM phenylmethylsulfonyl fluoride (PMSF). Non-specific sites were blockea with 3% BSA (Sigma) and i% normal goat serum (Dako) in PBS overnight at 4°C. Anti-2C or non-immune rabL'it IgG was added at a concentration of 10 # g / m l and incu0ated for 1 h~ Non-immune IgG was purified from normal ra~bit serum (Dako) using immobilized Protein A (Repligem Cambridge, MA). After extensive washing with PBS, alkaline phosphatase-conjugated goat anti-rabbit (Dako, diluted i:350 in PBS containing 1% BSA) was added and incubated for 0.5 h. After washing, the reaction product was developed with 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium in phosphatase buffer (see Section 2.5). Anti-2C antibodies labeled the 5-HT2c receptor cell line but not the 5-HT2A receptor cell line or the parental 3T3 fibroblasts (Backstrom and Sanders-Bush, unpublished observations). 2.4. Extractio~t of proteins from ceil lines and rat brain
Proteins were extracted from cell cultures and various brain regions to examine properties of the 5-HT2c receptor. Cell extracts were prepared from mouse NIH/3T3 fibroblasts and fibroblasts which stably express rat 5-HT2c (3T3/2C) or 5-HT2A (3T3/2A) receptors. For these experiments, each cell line was grown in 100 mm 2 plates. Each plate was washed twice with PBS and then treated with 1 ml of Tris extraction buffer (50 mM Tris, pH 7.6 containing I mM EDTA, 1 mM PMSF, 2 mM diisopropyl fluorophosphate (DIFP), and 5 #M leupeptin). For the tunicamycin and glycosidase experiments, cells were treated with 50 mM NaH2PO4/Na3PO4, pH 7.2 containing 5 mM EDTA, 1 mM PMSF, 2 mM DIFP, and 5 p,M leupeptin. The cells were scraped and transferred to 1.5 ml tubes. After sonication and centrifugation at 18000 × g for 15 min at 4°C, the Tris-soluble supernatant was discarded. Each pellet (membrane fraction) was treated with 0.2 ml of CHAPS extraction buffer (50 mM Tris, pH 7.6 containing I0 mM CHAPS, 0.05 mM EDTA, 1 mM PMSF, and 5 /.tM leupeptin) and the suspensions were sonicated. For the tunicamycin at, d glycosidase experiments, membranes were treated with 10 mM CHAPS prepared in 50 mM NaH2PO4/Na3PO4, pH 7.2, containing 1 mM EDTA, 1 mM PMSF, and 5 /.LM leupeptin. After sonicatio,a, the tubes were centrifuged and the CHAPS-soluble supernatant, which contained immunoreactive receptors, was collected. To determine the optimal CHAPS concentration for solubilizing immunoreactive receptors, various concentrations of CHAPS (1.25, 2.5, 5, 10, and 20 mM) were added to membrane fractions prepared in Tris buffer. For these experiments, the CHAPS-insoluble pellets obtained after centrif-ugation were subsequently treated with 0.2 ml of 1 × SDS-PAGE electrophoresis buffer. After sonication and centrifugation, the SDS-soluble supematant was col-
314
J.R. Backstrom et aL / Molecular Brain Research 33 (1995) 311-318
lected. Both CHAPS- and SDS-soluble fractions were examined for immunoreactive receptor to ensure that the total amount of recovered receptor was similar between treatments. Rat brain regions were dissected from decapitated, 250-300 g Sprague-Dawley male rats (Harlan Industries, Indianapolis, IN). Tissues were placed in 1.5 ml tubes containing 1 ml PBS on ice and then pelleted at 4°C. The supernatant was discarded and the pellet was washed with cold PBS and spun again. Soluble membrane proteins were prepared from the tissue as described for the cell extracts. Protein concentrations were determined with the BCA protein assay.
2.5. Western blots of proteins Extracts prepared from cultured ,zlls and various brain regions were mixed with SDS-PAGE sample buffer containing 2-mo,rcaptoethanol at a final concentration of 1%. The samples (20-150 /.tg of protein) were electrophoresed in 10% SDS-polyacrylamide gels. Proteins were transferred from acrylamide to nitrocellulose membranes in a modified Towbin transfer buffer (25 mM Tris, 192 mM glycine, pH--8.4, containing 0.05% 2-mercaptoethano!). The nitrocellulose was blocked with blot buffer (20 mM Tris, pH7.6, 150 mM NaCi, 0.05% Tween-20, 0.05% NaN 3) containing 3% BSA for 1-2 days at 4°C. The paper was probed at room temperature for 1 h with anti-2C (2-5 /xg/ml), washed with blot buffer, and then incubated for 1 h with alkaline phosphatase-conjugated goat anti-rabbit antibody (Dako) diluted 1:1000 in blot buffer containing 1% BSA. After washing with blot buffer, the paper was developed with 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium in phosphatase buffer (50 mM Tris, pH 9.5 containing 100 mM NaCI and 5 mM MgCI2). The masses of immunoreactive proteins were determined from standards (36-205 kDa, Sigma) which were electrophoresed and transferred to nitrocellulose strips. The strips were stained with 0.1% ponceau S in 5% acetic acid and destained with 5% acetic acid.
2.6. Metabolic inhibition of N-linked glycosylation Tunicamycin was added to 3T3/2C cells to inhibit the de novo addition of asparagine-linked (N-linked) oligosaccharides. Cells were plated at 1.5 x 106 cells/10 ml medium. After 2 days, the medium was replaced with 10 ml of fresh DMEM containing fetal bovine serum, penicillin, and streptomycin. After 1 h at 37°C, 20 p,l of vehicle or vehicle containing various concentrations of tunicamycin was added to the medium. Tunicamycin (Boehringer-Mannheim) was prepared at a concentration of 1.0 mg/ml in 0.01 N NaOH [4]. The treated cells were incubated at 37°C for 24 h. CHAPS-soluble protein was prepared from the cells and used for subsequent Western blot analysis of immunoreactive proteins.
2. Z Enzymatic remocal of oligosaccharides Aliquots of CHAPS-soluble material prepared from the 5-HT2c cell line and rat choroid plexus were incubated with glycosidases to cleave various oligosaccharide moieties from proteins. All glycosidases were obtained from Boehringer Maanheim. Recombinant N-glycosidase F (PNGase F, 0.2 U/ptl) was added to a final concentration of 0.5-1.0 U/0.1 mg protein. Neuraminidase (Clostridium perfringens, 3.5 Ll/mg) was dissolved in 50 mM NaH2PO4/Na3PO 4, pH 7.2 containing 10 mM EDTA and added to a final concentration of 4-20 mU/0.1 mg protein. O-Glycosidase (0.5 mU/p.l) was added to a final concentration of 2-6 mU/0.1 mg protein. All glycosidase and control buffer treatments were incubated at 37°C. The samples were treated with SDS-PAGE sample buffer and used for Western blot analysis of immunoreactive proteins.
3. Results
3.1. Identification of rat 5-HT2c receptors from cell cultures Immune serum was generated against a 19-mer 5-HT2c peptide coupled to KLH. The 2C peptide immunogen corresponds to a region of the putative third intracellular loop of the receptor (Fig. 1). Anti-2C antibodies were purified from the serum using an affinity chromatography matrix containing BSA-2C peptide. Membrane fractions were prepared from parental NIH/3T3 fibroblasts and fibroblasts which stably express rat 5-HT2c receptors (3T3/2C) or rat 5-HT2A receptors (3T3/2A). Proteins were solubilized from the membrane fraction with the detergent CHAPS, electrophoresed in SDS-polyacrylamide gels, and Western blotted to nitrocellulose. Anti-2C antibodies recognized predominant proteins with masses of 51-52 kDa and 58-68 kDa from 3T3/2C cells (Fig. 2, lane 3). The proteins were not present in the CHAPS-soluble extracts prepared from 3T3/2A cells (lane 2) or parental NIH/3T3 cells (lane 1). These results demonstrate that anti-2C recognizes rat 5-HT2c receptors expressed in a heterologous system. The optimal concentration of CHAPS required to solubilize receptors from the 5-HT:c receptor cell line was examined. Fig. 3 illustrates that the amount of immunoreactive receptor increased with increasing amounts of detergent until approximately 5 mM CHAPS per mg protein/ml (lanes 1-3). Thereafter, the amount of solubilized receptor remained constant (lanes 4-6). CHAPS-insoluble pellets were subsequently treated with SDS sample buffer to solubilize remaining receptors. Immunoreactive protein in the SDS-soluble fractions was detected only from CHAPS-insoluble pellets treated with 1.25 or 2.5 mM CHAPS (data not illustrated). Immunoreactive protein in the SDS-soluble fractions was not detected from pellets
315
J.R. Backstrom et al. ,/Molecular Brain Research 33 (t995) 311-318
eg (a qa
~
eea cea gea
c~ g.
2
a:
e
65 62 51
N
1
~
b=
t
~
N a s s (kDa~
m
5 51
3
Fig. 2. Idertificafion of immunoreactive proteins from cultured cells. CHAPS-soluble extracts were prepared and 80 p.g of protein per lane was used for Western blots. Protein was analyzed from N[H/3T3 fibroblasts (lane l/ and fibroNask,, stabl) expressing cDNA tot 5-HT~ ~3-f3/2A, lane 2) or 5-HT,c (3T3/2C, lane 3) receptors. Relative masses of predominant immunoreactive proteins from the 5-HT:c receptor cetl line 3T3/2C are indicated.
previously treated with 5.0, 7.9, or 15 mM CHAPS per mg protein/ml (data not illustrated). Since 5 mM CHAPS per mg protein/ml solubilized as many immunoreactive 5HT2c receptors as the sample buffer containing 2% SDS, all subsequent membrane preparations were treated with 10 mM CHAPS, which resulted in a final ratio of 5-7 mM CHAPS per mg protein/mE
CHAPS (raM):
1.2 2.5 5,0 10 20
mM CHAPS m g protein/mE
1.4 2.7 5.0 7.9 15
!_.o=ol.
m 2
3
4
5
6
Fig. 4. Western blot of CHAPS-soluble protein prepared from various regions of rat brain. ,~n equivalent amount of protein (50 p.g) was analyzed from cerebellum (lane lk frontal cortex (lane 2), striatum ~lane 3), hippocampus (lane 4), choroid plexus (lane 5k and 3T3/2C cells (lane 6).
3.2. Distribution of 5-HT:c receptors in rat brain Adult rats were used to determine the relative abundance of immunoreactive 5-HT_,c receptors from various brain regions. Equal amounts of protein soiubiiized from membrane fractions were Western blotted and probed with anti-2C, lmmunoreactive protein was detected from the choroid plexus (Fig. 4, lane 5), but not from the cerebellum, striatum, frontal cortex, and hippocampus (lanes 1-4, respectively). Immunoreactive proteins extracted from the choroid plexus had masses of 51 and 56-62 kDa (lane 5), which was similar, but not identical, to proteins extracted from the heterologous 5-HTzc receptor cell line (51-52 and 58-68 kDa, lane 6). Since differences in masses between the cell line and the choroid plexus may be due to species and/or cell-.D'pe specific glycosylation, the 5-HTzc receptor was analyzed for the presence of oligosacchar,.'des.
3.3. Glycosylation of 5-HT2c receptors
Lane:
1
2
3
4
5
Fig. 3. Optimization of the detergent concentration for extracting 5-HT, c receptors from 3T'3/2C cells. A membrane preparation was prepared and divided into five equal portions. Pellets were treated with Tfis buffer containing pmteinase inhibitors and the detergent CHAPS varying in concentration from 1.25 to 20 raM. The resulting supernatants were subjected to a protein assay and an equivalent volume of extract was analyzed for immunoreactive protein. The data are presented as the concentration of CHAPS (raM) relative to the concentration of protein (mg/ml)..
The presence of asparagine-linked (N-linked) oligosaccharides was examined metabolically and enzymatically. To inhibit de novo addition of sugar to polypeptides, cultures of the 5-HTzc receptor cell line were grown in the presence of 0.2 or 2.0 p.g tunicamycin/ml medium for 24 h. Fig. 5 illustrates that the predominant immunoreactive proteins from tunicamycin-treated cells had masses of 40 and 41 kDa (lanes 2 and 3).
316
J.R. Backstrom et ai. / Molecular Brain Research 33 (1995) 311-318
3T3/2C
Tunicamycin N-Glycosidase Neuraminidase:
~"
o~,"©
¢.)
c~
N
¢,)
N - G i y c o s i d a s e F:
"= ~
r"
Choroid
-
+
-
-
+
-
_
_
+
-
_
+
4
5
6
Mass (kD_ 66--
Mass (kDa) 65--
=
mi
51-41 .-40
51-41m
RMg
Htw
~
42 - - 41
w
1
Lane:
2
3
Fig. 6. Analysis of N-linked oligosaccharides associated with 5-H'r2c
1 Iir-,.~
~
A
I~,
"~
T 1:
2
3
4
5
6
1.
,,-,na,~,st,, of N-,,n,,ed ohgo..,a~.~.ha,,t,~s " - . . . . . :'~- associai~d wiih 5-HT2c 1"1~, receptors from 3 T 3 / 2 C cells. Cells were grown for 24 h in the presence of vehicle (lane 1) or tunicamycin at 0.2 /.tg/ml (lane 2) or 2 /.tg/ml (lane 3) to inhibit glycosylation. Cells were also grown in the absence of tunicamycin and the CHAPS-soluble extracts treated with N-glycosidase F for 1 h (lane 5) or 16 h (lane 6) to remove endogenous sugars. The control extract (lane 4) was incubated in the absence of enzyme for 16 h. .P.
To enzymatically remove endogenous oligosaccharides from glycoproteins, CHAPS-soluble extracts were prepared from 3T3/2C cells grown in the absence of tunicamycin. The extracts were treated with N-glycosidase F for 1 or 16 h and then subjected to SDS-PAGE. Immunoreactive proteins on the Western blots were identified with masses of 41 and 42 kDa (Fig. 5, lanes 5 and 6). A protein with a mass of 35 kDa was detected after longer incubations (lane 6), and may represent a proteolytic product of the 5-HT2c receptor. Extracts prepared from the choroid plexus were also treated with N-glycosidase F. lmmunoreactive protein with a mass of 41 kDa was detected from choroid plexus extracts (Fig. 6, lane 6). Longer incubations increased the relative amount of the 41-42 kDa proteins (Fig. 5 and data not illustrated). Since sialic acid (N-acetylneuraminic acid) may be associated with complex N-linked oligosaccharides, extracts prepared from the 5-HT2c receptor cell line and choroid plexus were treated with neuraminidase. The migration of the immunoreactive proteins from the cell line were not noticeably altered after incubation with the enzyme (Fig. 6, lanes 1 and 2). In contrast, treatment of the choroid plexus extracts with neuraminidase reduced the mass of the 51 and 58-62 kDa proteins (lane 4) to 50 and 54-58 kDa (lane 5). It is unlikely that this reduction in mass was due to proteinase activity rather than glycosidase activity since neuraminidase did not affect the migration of
receptors from 3 T 3 / 2 C cells and the choroid plexus. CHAPS-soluble extracts were prepared from 3 T 3 / 2 C cells (lanes 1-3) and the choroid plexus (lanes 4-6) and incubated for 2 h with neuraminidase (lanes 2 and 5) or N-glycosidase F (lanes 3 and 6). The control extracts (lanes 1 and 4) were incubated for 2 h in the absence of enzymes. Identical results to those of lanes 2 and 5 were obtained from samples incubated for 16 h with neuraminidase (data not illustrated).
immunoreactive receptors from 3T3/2C cells (Fig. 6) or the 41 kDa deglycosylated (N-linked) protein from the choroid plexus (Fig. 7, lane 2). The presence of serine/threonine-linked (O-linked) sugar associated with the 5-HT2c receptor was examined with the enzyme O-glycosidase. CHAPS-soluble extracts prepared from the 5-HT2c receptor cell line and choroid plexus were treated with N-glycosidase F to generate 41-42 kDa proteins lacking N-linked sugars (Fig. 7, lane 1). hnmunoreactive proteins with masses of 41-42 kDa were detected from extracts incubated with neuraminidase (lane 2). Similar proteins were also detected from extracts treated with neuraminidase and O-glycosidase (lane 3). Therefore, neither neuraminidase nor neuraminidase a~ad O-glycosidase altered the mass of the immunoreactive proteins lacking N-linked oligosaccharides in this assay.
N - G i y c o s i d a s e F:
+
+
+
Neuraminidase:
-
+
+
-
-
+
O-Glycosidase:
Mass(kDa)
4a
3 T 3 / 2 C cells: Choroid plexus: Lane:
1
2
3
Fig. 7. Analysis of O-linked oligosaccharides associated with 5-HTzc receptors from 3 T 3 / 2 C cells and the choroid plexus. CHAPS-soluble extracts were prepared and incubated with N-glycosidase F for 16 h. Some samples also included neuraminidase (lane 2) or neuraminidase and O-glycosidase (lane 3).
J.R. Backstrom e~ al. / Afolecular Brain Research 33 (1995) 311-318
4. Discussion Antibodies were raised against a region of the puta~:ive third intracellular loop of the rat 5-HT2c receptor and used to identify immunoreactive protein from a 5-HTx. receptor cell line (3T3/2C) and rat brain. The hydrophilic. 19 amino acid peptide used for production and purification of antibodies is not present in the other known 5-HT 2 receptor subtypes, 5-HT2A and 5-HT2~. The anti-2C antibodies were purified from serum using affinity chromatography. Cell lines were used to examine the specificity of anti-2C. In Western blots, anti-2C recognized proteins from CHAPS-soluble extracts of 3T3/2C cells, but not from extracts of 3T3/2A or parental NIH/3T3 cells. The immunoreactive proteins migrated as two distinct regions with masses of 51-52 and 58-68 kDa. Solubilization of immunoreactive receptors from 3T3,/2C cells was saturable and found to be maximal with 5 mM CHAPS per mg protein/ml. The relative abundance of 5-HT2c receptors in rat brain extracts was examined with Western blots. An equal amount of CHAPS-soluble protein was analyzed from selected regions that were previously examined for relative abundance of 5-HT2c receptor rnRNA using Northern blots [10]. The choroid plexus contained the greatest amount of receptor protein, consistent with the results from Northern blots. Immunoreactive receptor was not detected in the cerebellum, frontal cortex, striatum, or hippocampus using the same conditions. The hippocampus, which contains at least 10-fold less 5-HT2c receptor mRNA than the choroid plexus [10], contained at least 7.5-fold less CHAPS-soluble 5-HT2c receptor protein since immunoreactive receptor was detected with 20 /xg protein from choroid plexus extracts, but not with 150 /.tg protein from hippocampus extracts (data not illustrated). In conclusion, the results from Western blots are consistent with the results from Northern blots which suggest that the relative abundance of 5-HT2c receptors is greatest in the choroid plexus. Different masses were found for imrnunoreactive 5-HT2c receptors in extracts from choroid plexus and 3T3/2C fibroblasts. Immunoreactive protein from choroid plexus extracts was detected with masses of 51 and 54-58 kDa, while protein from 3T3/2C cells had masses of 51-52 and 58-68 kDa. These differences between the two sources may be due to cell-type and/or species specific glycosylation (reviewed in [17]). Thus, experiments were undertaken to investigate glycosylation of 5-HT2c receptors. Two experimental approaches were utilized to examine the association of N-linked oligosaccharides with 5-HT2c receptors. In the first set of experiments, 3T3/2C cells were grown in the presence of the nucleoside antibiotic tunicamycin, which inhibits the first step of ,he biosynthetic pathway and prevents the down:stream addition of oligosaccharides to asparagine (reviewed in [6]). Extracts prepared from cells grown in the presence of tunicamycin
317
contained immunoreactive proteins with masses of 40 and 41 kDa, which was significantly less than the 51-52 and 58-68 kDa receptors from control 3T3/2C cells. In the second set of experiments, extracts prepared from cells grown in the absence of tunicamycin were treated with N-glycosidase F to enzymatically remove available N-linked sugars from giycoproteins. N-Glycosidase F (PNGase F) cleaves the bond between asparagine and N-acetylglucosamine [16]. Predominant 41 and 42 kDa proteins were detected from 3T3/2C extracts treated with N-glycosidase F. Differences between the masses of the 40-42 kDa proteins from tunicamycin-treated cells (40-41 kDa) and N-glycosida~c-treated extracts (41-42 kDa) could be due to post-translational modifications of the protein. These variations may occur in the cell between the time of oligosaccharide addition (inhibited with tunicamycin) and functional expression. A 41 kDa immunoreactive protein was also detected from the choroid plexus after treatment with N-glycosidase F. In conclusion, these experiments establish that the immunoreactive 5-HT2c receptors with masses of 51-52 and 58-68 kDa from 3T3/2C cell extracts and 51 and 54-58 kDa from choroid plexus contain N-linked oligosaccharides. Extracts prepared from 3T3/2C cells and choroid plexus were treated with neuraminidase to investigate the presence of sialic acid residues. Neuraminidase did not significantly alter the mass of the 5-HT2c receptors from 3T3/2C cells, but significantly reduced the mass of the 5-HT2c receptors from choroid plexus. The mass of these receptors decreased from 51 and 58-62 kDa to 50 and 54-58 kDa. To determine whether sialic acid was associated with N- or O-linked oligosaccharide chains of the 5-HT2c receptors, extracts were treated with N-glycosidase to remove N-linked sugars and neuraminidase. The 41 kDa protein lacking N-linked sugar migrated with the same mass after treatment with neuraminidase, which suggests that sialic acid was associated with N-linked, rather than O-linked, sugars. Although O-glycosidase did not significantly alter the mass of the deglycosylated (N-linked) receptor, the possibility that 5-HT2c receptors contain O-linked sugar cannot be ruled out. For example, the sugar may not be accessible to O-giycosidase or the decrease in mass after deglycosylation (O-linked) may not be detectable in this assay. In conclusion, data are presented which demonstrate that rat 5-ttT2c receptors from cell and brain extracts contain N-linked oligosaccharides and suggest that 5-HT2c receptors in the choroid plexus associate with sialic acid residues. The rat 5-HT2c receptor contains six sites that fit the consensus sequence of asparagine-linked glycosylation sites. Three of these sites are predicted to be exposed to the extracellular milieu, and therefore more likely to contain sugars. At least one of the extracellular sites may be utilized in vivo since the mass of immunoreactive receptors decreased when cells were grown in the presence of tunicamycin or extracts were incubated with N-glycosidase F. The presence of two potential sites at residues 204 and
318
J.R. Backstrom et at. / Mo!ecular Brain Research 33 (1995) 311-318
205 may ensure that this domain is available for N-linked glyco.,;ylation. Another domain, possibly at asparagine 39, may also be glycosylated since 3T3/2C cells and choroid plexus extracts contained immunoreactive receptors in two distinct regions of SDS-polyacrylamide gels. We are currently developing antibodies to the amino- and carboxyterminal regions of the 5-HT2c receptor and generating point mutants to determine the role of N-linked glycosylation on 5-aT2c receptor function.
Acknowledgements The authors wish to thank Ann Westpl~,l for her expert technical assistance with the cell cultures. This work was supported by Research Grant MH 34007 from the Alcohol, Drug Abuse, and Mental Health Administration and by a grant from Bristol Myers Squibb Corp.
References [1] Barker, E.L. and Sanders-Bush, E., 5-Hydroxytryptaminelc receptor density and mRNA levels in choroid plexus epithelial cells after treatment with mianserin and ( - ) - l - ( 4 - b r o m o - 2 , 5 dimethoxyphenyl)-2-aminopropane, Mol. Pharm., 44 (1993) 725730. [2] Barker, E.L., Westphal, R.S., Schmidt, D. and Sanders-Bush, E., Constitutively active 5-hydroxytryptamine 2C receptors reveal novel inverse agonist activity of receptor ligands, J. Biol. Chem., 269 (1994) 11687-11690. [3] Boess, F.G. and Martin, I.L., Molecular biology of 5-HT receptors, Neuropharmacology, 33 (1994) 275-317. [4] Chatterjee, S., Kwiterovich, P.O. Jr. and Sekerke, C.S., Effects of tunicamycin on the binding and degradation of !o~, density lipoproteins and glycoprotein synthesis in cultured human fibroblasts, J. Biol. Chem., 254 (1979) 3704-3707. [5] Conn, P.J., Sanders-Bush, E., Hoffman, B.J. and Hartig, P.R., A unique serotonin receptor in choroid plexus is linked to phosphafidylinositol turnover, Proc. Natl. Acad. Sci. USA, 83 (1986) 40864088.
[6] Elbein, A.D., Inhibitors of the biosynthesis and processing of N-linked oligosaccharide chains, Annu. Ret.. Biochem., 56 (1987) 497534. [7] Grotewiel, M.S. and Sanders-Bush, E., Regulation of serotonin~A receptors in heterologous expression systems, J. Neurochem., 63 (1994) 1255-1260. [8] Hoffman, B.J. and Mezey, E., Distribution of serotonin 5-HTIc receptor mRNA in adult rat brain, FEBS Lett., 247 (1989) 453-462. [9] Hoyer, D., Clarke, D.E., Fozard, J.R., Hartig, P.R.. Martin, G.R., Mylecharane, E.J., Saxena, P.R. and Humphrey, P.P.A., International union of pharmacology classification of receptors for 5-hydroxytryptamine (serotonin), Pharmacol. Ret,., 46 (1994) 157-203. [10] Julius, D., MacDermott, A.B., Axel, R. and Je.,isell, T.M., Molecular characterization of a functional cDNA encoding the serotonin lc receptor, Science, 241 (1988)558-564. [11] Julius, D., Huang, K.N., Livelli, T.J., Axel, R. and JesseU, T.M., The 5HT2 receptor defines a famil', of structurally distinct but functionally conserved serotonin receptors, Proc. Natl. Acad. Sci. USA, 87 (1990) 928-932. [12] Kaufman, M.J., Hartig, PR. and Hoffman, B.J., Serotonin 5-l-lT2c receptor stimulates cyclic GMP formation in choroid plexus, J. Neurochem., 64 (1995) 199-205. [13] Mengod, G., Nguyen, H.L., Waeber C, L:dbb,:li, H. and Pa!actos, J.M., The di.~t;ibution and cellular localization of the serotonin 1C receptor mRNA in the rodent brain examined bv in situ hybridization histochemistry. Comparison with receptor binding distribution, Neuroscience, 35 (1990)577-591. [14] Molineaux, S.M., Jessell, T.M., Axel, R. and Julius, D., 5-HTlc receptor is a prominent serotonin receptor subtype in the central nervous system, Proc. Natl. Acad. Sci. USA, 86 (1989) 6793-6797. [15] Pompeiano, M., Palacios, J.M. and Mengod, G., Distribution of the serotonin 5-HT 2 receptor family mRNAs: comparison between 5HT2A and 5-HT2c receptors, Mot. Brain Res., 23 (1994) 163-178. [16] Tarentino, A.L., Gomez, C.M. and Plummer, T.H. Jr., Deglycosy!ation of asparagine-linked glycans by peptide:N-glycosidase F, Biochemistr), 24 (1985) 4665-4671. [17] Varki, A., Biological roles of oligosaccharides: all of the theories are correct, G~ycobiology, 3 (1993)97-130. [18] Westphal, R.S. and Sanders-Bush, E., Reciprocal binding properties of 5-hydroxytryptamine type 2C receptor agonists and inverse agonists, Mot. Pharm., 46 (1994) 937-942. [19] Yu, L., Nguyen, H., Lc, H., Bloem, L.J.. Kozak. C.A., Hoffman, B.J., Snutch, T.P., Lester, H.A., Davidson, N. and Liibbert, H., The mouse 5-H~:c receptor contains eight hydrophobic domains and is X-linked, biol. Brain Res., 11 (1991) 143-149.
Molecular Brain Rcscarch 33 1t995) 31 t--318
Research report
Identification of rat serotonin 5-HT2c receptors as glycoproteins containing N-linked oligosaccharides Jon R. Backstrom *, Ryan S. Westphat, Herv , Canton, Elaine Sanders-Bush Department of Pharmacology, 432 Medical Research Building, Vanderbilt Unitersi~' School of Medicine, Nashtille. TN 37232-6600, USA
Accepted 21 June 1995
Abstrad Anlibodies against a portion of the rat 5-HT x. receptor third intracellular loop were generated and used to identify receptors soiubiiized from cell lines and rat brain. Western blots of CtLa,PS-soluble proteins were probed with affinity-purified anti-2C antibodies. I'he specificity of anti-2C was demonstrated with extracts prepared from NIH/3T3 fibroblasts which stably express functional rat 5-HT2c or 5-HTzA receptors. Extracts from the 5-HT2c cell line, but not the 5-HT2A cell line, contained immunoreactive proteins with masses of 51-52 kDa and 5 8 - 6 8 kDa. In the brain, immunoreactive proleins were identified from choroid plexus extracts with masses of 51 kDa and 58-62 kDa. The major 58-62 kDa and minor 51 kDa proteins were not detected in extracts prepared from the hippocampus, striatum, or frontal cortex using the same amount of CHAPS-soluble protein. These results are consistent with previous studies demonstrating that 5-HT2c receptor binding sites and mRNA are most abundant in choroid plexus. The association of asparagine-linked (N-link~,:d) oligosaccharides with the receptors was examined next. The 5-HT2c receptor cell line (3T3/2C) was grown in the presence of tunicamycin to metabolically inhibit N-linked glycosyiation. Proteins from the cell extracts were detected with masses of 40 and 41 kDa. Extracts prepared from 3 T 3 / 2 C cells (grown in the absence of tunicamycin) and from choroid plexus were incubated with N-glycosidase F to enzymatically remove available N-linked sugars. Immunoreactive proteins were detected with masses of 41 and 42 kDa from 3T3/2C cells and 41 kDa from choroid plexus. Neuraminidase, which cleaves sialic acid (N-acctylneuraminic acid) residues from glycoproteins, reduced the mass of the 51 and 58-62 kDa proteins from the choroid plexus to 50 and 54-58 kDa. In contrast, the 51-52 and 58-68 kDa proteins from 3T3/2C cells were not affected by treatment with neuraminidase. These results demonstrate that 5-H'I.,(. receptors contain N-linked sugars and suggest that sialic acid residues associate with 5-HT_,c receptors in the chc,roid plexus. The oligosaccharide moieties, which contribute up to "-- 30% of the relative mass as judged by SDS-polyacrylamide gel electrophoresis, ma3 impart functional properties to 5-HTzc receptors. Keywords: 5-HT2c receptors; Glycosylation; Sialic acid; Choroid plexus; Serotonin; Antibodies
1. Introduction The 5-hydroxytryptamin%c (5-HT2c) receptor is a member of the 5-HT 2 family of receptors that are coupled to phosphatidylinositol turnover [5]. The 5-HT2c receptor is also coupled to cyclic guanosine monophosphate (cGMP) formation [12]. Within the 5-HT 2 family, rodent 2A, 2B. and 2C receptors are approximately 50% identical at the amino acid level (reviewed in [3]). However, unique re-
Abbreviations: 5-HT, 5-hydroxytryptamine (serotonin); CHAPS, 3[(3-cholamidopropyi)dimethylammonio]- 1-propanesulfonate; SDS, sodium dodecyi sulfate; BSA, bovine serum albumin " Correspondinp, author. Fax: (1) (6i5) 343-6532; E-mail: BackstJR@ctrvax.'~anderbilt.edu0169-328X/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSD! 6169-328X(95 )00156-5
gions are present in the putative extracellular amino terminus, the third intracellular loop, and the intracellular cytoplasmic tail. These unique regions may be involved in protein-protein interactions. The distribution and relative abundance of 5-HT2c receptor mRNA in rat brain has been described [8,10,14,15]. Epithelial cells of the choroid plexus contained the highest density of receptor mRNA using in situ hybridization techniques (reviewed in [3]). Intermediate levels of mRNA were found in the hippocampus, amygdala, frontal cortex, caudate putamen, lateral habenula, subthalamic nuclei, substantia nigra (pars compacta), and regions of the olfactory system and hypothalamus. The cerebellum contained the lowest density of 5-HT2c receptor mRNA in the brain [8]. The levels of 5-HTzc receptor mRNA relative to total mRNA were investigated using Northern blots [10]. The
312
J.R. Backstrom et al. / Molecular Brain Re,search 33 (1995) 311-318
choroid plexus contained the highest amount of receptor mRNA. The second highest amount was detected in the hippocampus, which was at least 10-fold less abundant than in the choroid plexus [10]. Receptor mRNA was even less abundant in the basal ganglia, hypothalamus, ponsmedulla, and olfactory bulb. Also consistent with the in situ results, receptor mRNA was not detected in the cerebellum. Several studies have compared the sites of rat 5-HT, c receptor mRNA accumulation with sites of radioligand binding (reviewed in [9]). Binding sites of the 5-HT2c receptor are localized with radiolabeled ligands such as 5-HT (serotonin), mesulergine, and LSD. Although most brain regions which contain mRNA also bind radioligand, there appear to be some discrepancies. For examrle, the density of [3H]mesulergine sites in the lateral habenula and the subthalamic nuclei was lower than expected from., the corresponding density of in situ hybridization sites, suggesting that synthesis of the receptor may be regulated in a cell-specific manner [13]. Furthermore, down-regulation of 5-HT2c receptor binding sites by drug treatment has been shown to be independent of changes in mRNA levels [1]. Biochemical properties of 5-HT2c receptors have not been established due to the lack of specific high affinity probes. Here, we generate antibodies against a domain of the rat 5-HT2c receptor and utilize them to characterize receptors synthesized in vivo. Western blots of membrane extracts were probed with affinity-purified antibodies. First, the masses of immunoreactive receptors were determined from extracts of cells that stably express cDNA for the rat 5-HT2c receptor [2,18]. Second, the presence of immunoreactive receptor was examined in several regions of rat brain previously analyzed for tb.e relative abundance of 5-HT2c receptor mRNA using Northern blots [10]. Third, the presence of asparagine-linked (N-linked) oligosaccharides attached to the protein moiety was examined using two strategies. In one set of experiments, the 5-HT2c receptor cell line was cultured with the antibiotic tunicamycin, which inhibits the biosynthetic addition of N-linked sugars (reviewed in [6]). In the other set of experiments, extracts prepared from the 5-HT2c receptor cell line and the choroid plexus were treated with N-glycosidase F to digest N-linked oligosaccharides and with neuraminidase to digest terminal sialic acid residues. Evidence is presented for the presence of N-linked sugars associated with 5-HT2c receptors in the rat brain.
2. Materials and methods
•N H 3
Membrane -~87v
~
J,.,,~~
intrace,lular
2B Peptide: AHr.DOSHKDK• ~PNSTDZT 2A Peptide: SL$SEKLFQIRSiFIREFGS¥
COO
Fig. 1. Schematic representation of rat 5-HT2c receptor (adapted from [10]). The 5-HTzc receptor peptide (2C peptide) used to generate anli-2C antibodies corresponds to a domain of the putative third intracellular loop. Also shown are sequences of the similar domains from the related 5-HT2B and 5-HTzA receptors (2B and 2A peptide,~; respecdvcly). Asparagine residues that fulfill the N-linked consensus sequence asn-Y.-ser/thr are denoted with an arrowhead; numbers indicate their amino ack" positions.
tion (Rockville, MD). The 5-HT:A receptor cell line GF6, referred to as 3T3/2A, was a kind gift of Dr. David Julius, University of California at S.~n Francisco [11]. These cells express 5-HT2A receptors at 3 pmol/mg protein [7]. The 5-HT2c receptor cell line B3-1G, referred to as 3T3/2C, expresses 5-HT2c receptors at 5 pmol/mg protein [i8]. Cells were grown in Dulbecco's modified Eagle's medium (DMEM) containing 9% fetal bovine serum, 5 U/ml penicillin, and 5 /zg/ml streptomycin. The serum was obtained from Hyclone Laboratories (Logan, UT) and the penicillin and streptomycin were from Gibco/BRL Life Technologies (Grand Island, NY).
2.2. Generation of 5-HT2c receptor antibodies The 5-HTEc peptide (NH 2-CKKNGGEEENAPNPNPDQK-COOH) corresponds to amino acids 270-288 of the rat [10] and mouse [19] sequences. The peptide is located in the putative third intracellular loop [10] and has no significant homology to the corresponding regions of the rat 5-HTEA and 5-HT2B receptors (Fig. 1). The 5-HT2c peptide was synthesized, coupled to a carrier protein, and used to immunize rabbits and purify specific antibodies. The peptide was conjugated to keyhole limpet hemocyanin (KLH) and 1 mg suspended in Freund's complete adjuvant and injected s.c. in New Zealand rabbits (East Acres, Southbridge, MA). Antiserum was collected one week after each boost injection of 0.5 mg KLH-peptide prepared in Freund's incomplete adjuvant.
2.1. Cell cultures 2.3. Purification of 5-HT2c receptor antibodies Mouse NIH/3T3 fibroblasts and stable clones derived from transfection of the fibroblasts with rat 5-HT2c or 5-HT2A receptor cDNA were used for this study. Fibroblasts were obtained from American Type Culture Collec-
Anti-2C antibodies were purified from immune serum using a three step procedure. Serum was passed sequentially through a column containing BSA and then a column
J.R. Backstrom et al. / Molecular Bram Research 33 (1995~ 311-318
containing BSA coupled to the 2C peptide (BSA-2C peptide). In the third step, proteins eluted from the BSApeptide column were incubated with blots containing soluble proteins extracted from NIH/3T3 fibrobiasts. To construct the columns for the first two purification steps. BSA or BSA-2C peptide was added to an activated matrix (Reacti-gei 6x, Pierce Chemical Co., Rockford, IL) at a concentration of 2 mg protein per ml of gel. Excess reactive sites of the matrices were blocked with I M Tris, pH 9.0 (neutralization buffer). Two column~ containing i ml of immobilized BSA or BSA-2C peptide were washed extensively with TBS (50 mM Tris, pHT.6 containing 150 mM NaCI, 0.05% NaN 3) and elution buffer (50 mM glycine, pH 2.7 containing 0.05% NaN 3) before being equilibrated in_ TBS. Immune serum was thawed, centrifuged, and diluted with an equal volume of TBS. 10-15 ml of the diluted serum was passed through the BSA column at room temperature, collected, and applied to the BSA-peptide column. The BSA-peptide column was washed with ~ 25 ml TBS and bound proteins were eluted from the column with 10 mt of elution buffer. The material was collected in 15 ml tubes containing 0.3 ml of neutralization buffer. Protein concentrations were determined with the BCA protein assay (Pierce) using bovine serum albumin (BSA) as the protein standard. 2.5 ml of 5% BSA and 0.6 M NaCI in TBS was added to the antibody solution to make final concentrations of 1% BSA and 0.15 M NaCt. The final step in the purification procedure was employed to remove non-specific antibodies from the antibody solution that recognized soluble fibroblast proteins. The eluate from the BSA-peptide column (antibody solution) was incubated with a Western blot prepared from Tris-soluble extracts of NIH/3T3 cells. (The cross-reactive proteins were not detected in CHAPS- or SDS-soluble membrane fractions). Briefly, 2.0 mg Tris-soluble protein (see Section 2.5) was electrophoresed in a 12.5 ( h ) x 14 (w) X 1.5 cm preparatory gel and the regions corresponding to ~ 50-200 kDa were blotted to nitrocellulose (Hoefer Scientific Instruments). The membrane was blocked with BSA, washed sequentially with glycine elution buffer and blot buffer, and then equilibrated in blot buffer. The blot was incubated with the antibody solution obtained from the BSA-2C peptide column. After 0.5 h, the solution was collected and the presence of cross-reactive antibodies was monitored with Western blots prepared from soluble 3T3 proteins. The purified antibodies, refe.n-.ed to as anti-2C, were stored at 4°C. Initial specificity of anti-2C was screened cytochemically with NIH/3T3 fibroblasts and clones stably expressing 5-HT2A eDNA (3T3/2A) or 5-HT2c eDNA (3T3/2C). Cells were plated in wells containing type I collagen (Sigma). Sub-confluent cultures were washed twice with phosphate buffered saline (PBS; 15 mM phosphate, pH 7.4 containing 140 mM NaCI, 0.05% NaN 3} and fixed for 0.5 h with freshly prepared 4% formaldehyde, 4% sucrose in PBS. The cells were treated with 50 mM NH4CI in PBS
313
for 5 rain and then permeabiiized fer 5 rain with 0.2% Triton X-i00 in PBS containing 1 mM phenylmethylsulfonyl fluoride (PMSF). Non-specific sites were blockea with 3% BSA (Sigma) and i% normal goat serum (Dako) in PBS overnight at 4°C. Anti-2C or non-immune rabL'it IgG was added at a concentration of 10 # g / m l and incu0ated for 1 h~ Non-immune IgG was purified from normal ra~bit serum (Dako) using immobilized Protein A (Repligem Cambridge, MA). After extensive washing with PBS, alkaline phosphatase-conjugated goat anti-rabbit (Dako, diluted i:350 in PBS containing 1% BSA) was added and incubated for 0.5 h. After washing, the reaction product was developed with 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium in phosphatase buffer (see Section 2.5). Anti-2C antibodies labeled the 5-HT2c receptor cell line but not the 5-HT2A receptor cell line or the parental 3T3 fibroblasts (Backstrom and Sanders-Bush, unpublished observations). 2.4. Extractio~t of proteins from ceil lines and rat brain
Proteins were extracted from cell cultures and various brain regions to examine properties of the 5-HT2c receptor. Cell extracts were prepared from mouse NIH/3T3 fibroblasts and fibroblasts which stably express rat 5-HT2c (3T3/2C) or 5-HT2A (3T3/2A) receptors. For these experiments, each cell line was grown in 100 mm 2 plates. Each plate was washed twice with PBS and then treated with 1 ml of Tris extraction buffer (50 mM Tris, pH 7.6 containing I mM EDTA, 1 mM PMSF, 2 mM diisopropyl fluorophosphate (DIFP), and 5 #M leupeptin). For the tunicamycin and glycosidase experiments, cells were treated with 50 mM NaH2PO4/Na3PO4, pH 7.2 containing 5 mM EDTA, 1 mM PMSF, 2 mM DIFP, and 5 p,M leupeptin. The cells were scraped and transferred to 1.5 ml tubes. After sonication and centrifugation at 18000 × g for 15 min at 4°C, the Tris-soluble supernatant was discarded. Each pellet (membrane fraction) was treated with 0.2 ml of CHAPS extraction buffer (50 mM Tris, pH 7.6 containing I0 mM CHAPS, 0.05 mM EDTA, 1 mM PMSF, and 5 /.tM leupeptin) and the suspensions were sonicated. For the tunicamycin at, d glycosidase experiments, membranes were treated with 10 mM CHAPS prepared in 50 mM NaH2PO4/Na3PO4, pH 7.2, containing 1 mM EDTA, 1 mM PMSF, and 5 /.LM leupeptin. After sonicatio,a, the tubes were centrifuged and the CHAPS-soluble supernatant, which contained immunoreactive receptors, was collected. To determine the optimal CHAPS concentration for solubilizing immunoreactive receptors, various concentrations of CHAPS (1.25, 2.5, 5, 10, and 20 mM) were added to membrane fractions prepared in Tris buffer. For these experiments, the CHAPS-insoluble pellets obtained after centrif-ugation were subsequently treated with 0.2 ml of 1 × SDS-PAGE electrophoresis buffer. After sonication and centrifugation, the SDS-soluble supematant was col-
314
J.R. Backstrom et aL / Molecular Brain Research 33 (1995) 311-318
lected. Both CHAPS- and SDS-soluble fractions were examined for immunoreactive receptor to ensure that the total amount of recovered receptor was similar between treatments. Rat brain regions were dissected from decapitated, 250-300 g Sprague-Dawley male rats (Harlan Industries, Indianapolis, IN). Tissues were placed in 1.5 ml tubes containing 1 ml PBS on ice and then pelleted at 4°C. The supernatant was discarded and the pellet was washed with cold PBS and spun again. Soluble membrane proteins were prepared from the tissue as described for the cell extracts. Protein concentrations were determined with the BCA protein assay.
2.5. Western blots of proteins Extracts prepared from cultured ,zlls and various brain regions were mixed with SDS-PAGE sample buffer containing 2-mo,rcaptoethanol at a final concentration of 1%. The samples (20-150 /.tg of protein) were electrophoresed in 10% SDS-polyacrylamide gels. Proteins were transferred from acrylamide to nitrocellulose membranes in a modified Towbin transfer buffer (25 mM Tris, 192 mM glycine, pH--8.4, containing 0.05% 2-mercaptoethano!). The nitrocellulose was blocked with blot buffer (20 mM Tris, pH7.6, 150 mM NaCi, 0.05% Tween-20, 0.05% NaN 3) containing 3% BSA for 1-2 days at 4°C. The paper was probed at room temperature for 1 h with anti-2C (2-5 /xg/ml), washed with blot buffer, and then incubated for 1 h with alkaline phosphatase-conjugated goat anti-rabbit antibody (Dako) diluted 1:1000 in blot buffer containing 1% BSA. After washing with blot buffer, the paper was developed with 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium in phosphatase buffer (50 mM Tris, pH 9.5 containing 100 mM NaCI and 5 mM MgCI2). The masses of immunoreactive proteins were determined from standards (36-205 kDa, Sigma) which were electrophoresed and transferred to nitrocellulose strips. The strips were stained with 0.1% ponceau S in 5% acetic acid and destained with 5% acetic acid.
2.6. Metabolic inhibition of N-linked glycosylation Tunicamycin was added to 3T3/2C cells to inhibit the de novo addition of asparagine-linked (N-linked) oligosaccharides. Cells were plated at 1.5 x 106 cells/10 ml medium. After 2 days, the medium was replaced with 10 ml of fresh DMEM containing fetal bovine serum, penicillin, and streptomycin. After 1 h at 37°C, 20 p,l of vehicle or vehicle containing various concentrations of tunicamycin was added to the medium. Tunicamycin (Boehringer-Mannheim) was prepared at a concentration of 1.0 mg/ml in 0.01 N NaOH [4]. The treated cells were incubated at 37°C for 24 h. CHAPS-soluble protein was prepared from the cells and used for subsequent Western blot analysis of immunoreactive proteins.
2. Z Enzymatic remocal of oligosaccharides Aliquots of CHAPS-soluble material prepared from the 5-HT2c cell line and rat choroid plexus were incubated with glycosidases to cleave various oligosaccharide moieties from proteins. All glycosidases were obtained from Boehringer Maanheim. Recombinant N-glycosidase F (PNGase F, 0.2 U/ptl) was added to a final concentration of 0.5-1.0 U/0.1 mg protein. Neuraminidase (Clostridium perfringens, 3.5 Ll/mg) was dissolved in 50 mM NaH2PO4/Na3PO 4, pH 7.2 containing 10 mM EDTA and added to a final concentration of 4-20 mU/0.1 mg protein. O-Glycosidase (0.5 mU/p.l) was added to a final concentration of 2-6 mU/0.1 mg protein. All glycosidase and control buffer treatments were incubated at 37°C. The samples were treated with SDS-PAGE sample buffer and used for Western blot analysis of immunoreactive proteins.
3. Results
3.1. Identification of rat 5-HT2c receptors from cell cultures Immune serum was generated against a 19-mer 5-HT2c peptide coupled to KLH. The 2C peptide immunogen corresponds to a region of the putative third intracellular loop of the receptor (Fig. 1). Anti-2C antibodies were purified from the serum using an affinity chromatography matrix containing BSA-2C peptide. Membrane fractions were prepared from parental NIH/3T3 fibroblasts and fibroblasts which stably express rat 5-HT2c receptors (3T3/2C) or rat 5-HT2A receptors (3T3/2A). Proteins were solubilized from the membrane fraction with the detergent CHAPS, electrophoresed in SDS-polyacrylamide gels, and Western blotted to nitrocellulose. Anti-2C antibodies recognized predominant proteins with masses of 51-52 kDa and 58-68 kDa from 3T3/2C cells (Fig. 2, lane 3). The proteins were not present in the CHAPS-soluble extracts prepared from 3T3/2A cells (lane 2) or parental NIH/3T3 cells (lane 1). These results demonstrate that anti-2C recognizes rat 5-HT2c receptors expressed in a heterologous system. The optimal concentration of CHAPS required to solubilize receptors from the 5-HT:c receptor cell line was examined. Fig. 3 illustrates that the amount of immunoreactive receptor increased with increasing amounts of detergent until approximately 5 mM CHAPS per mg protein/ml (lanes 1-3). Thereafter, the amount of solubilized receptor remained constant (lanes 4-6). CHAPS-insoluble pellets were subsequently treated with SDS sample buffer to solubilize remaining receptors. Immunoreactive protein in the SDS-soluble fractions was detected only from CHAPS-insoluble pellets treated with 1.25 or 2.5 mM CHAPS (data not illustrated). Immunoreactive protein in the SDS-soluble fractions was not detected from pellets
315
J.R. Backstrom et al. ,/Molecular Brain Research 33 (t995) 311-318
eg (a qa
~
eea cea gea
c~ g.
2
a:
e
65 62 51
N
1
~
b=
t
~
N a s s (kDa~
m
5 51
3
Fig. 2. Idertificafion of immunoreactive proteins from cultured cells. CHAPS-soluble extracts were prepared and 80 p.g of protein per lane was used for Western blots. Protein was analyzed from N[H/3T3 fibroblasts (lane l/ and fibroNask,, stabl) expressing cDNA tot 5-HT~ ~3-f3/2A, lane 2) or 5-HT,c (3T3/2C, lane 3) receptors. Relative masses of predominant immunoreactive proteins from the 5-HT:c receptor cetl line 3T3/2C are indicated.
previously treated with 5.0, 7.9, or 15 mM CHAPS per mg protein/ml (data not illustrated). Since 5 mM CHAPS per mg protein/ml solubilized as many immunoreactive 5HT2c receptors as the sample buffer containing 2% SDS, all subsequent membrane preparations were treated with 10 mM CHAPS, which resulted in a final ratio of 5-7 mM CHAPS per mg protein/mE
CHAPS (raM):
1.2 2.5 5,0 10 20
mM CHAPS m g protein/mE
1.4 2.7 5.0 7.9 15
!_.o=ol.
m 2
3
4
5
6
Fig. 4. Western blot of CHAPS-soluble protein prepared from various regions of rat brain. ,~n equivalent amount of protein (50 p.g) was analyzed from cerebellum (lane lk frontal cortex (lane 2), striatum ~lane 3), hippocampus (lane 4), choroid plexus (lane 5k and 3T3/2C cells (lane 6).
3.2. Distribution of 5-HT:c receptors in rat brain Adult rats were used to determine the relative abundance of immunoreactive 5-HT_,c receptors from various brain regions. Equal amounts of protein soiubiiized from membrane fractions were Western blotted and probed with anti-2C, lmmunoreactive protein was detected from the choroid plexus (Fig. 4, lane 5), but not from the cerebellum, striatum, frontal cortex, and hippocampus (lanes 1-4, respectively). Immunoreactive proteins extracted from the choroid plexus had masses of 51 and 56-62 kDa (lane 5), which was similar, but not identical, to proteins extracted from the heterologous 5-HTzc receptor cell line (51-52 and 58-68 kDa, lane 6). Since differences in masses between the cell line and the choroid plexus may be due to species and/or cell-.D'pe specific glycosylation, the 5-HTzc receptor was analyzed for the presence of oligosacchar,.'des.
3.3. Glycosylation of 5-HT2c receptors
Lane:
1
2
3
4
5
Fig. 3. Optimization of the detergent concentration for extracting 5-HT, c receptors from 3T'3/2C cells. A membrane preparation was prepared and divided into five equal portions. Pellets were treated with Tfis buffer containing pmteinase inhibitors and the detergent CHAPS varying in concentration from 1.25 to 20 raM. The resulting supernatants were subjected to a protein assay and an equivalent volume of extract was analyzed for immunoreactive protein. The data are presented as the concentration of CHAPS (raM) relative to the concentration of protein (mg/ml)..
The presence of asparagine-linked (N-linked) oligosaccharides was examined metabolically and enzymatically. To inhibit de novo addition of sugar to polypeptides, cultures of the 5-HTzc receptor cell line were grown in the presence of 0.2 or 2.0 p.g tunicamycin/ml medium for 24 h. Fig. 5 illustrates that the predominant immunoreactive proteins from tunicamycin-treated cells had masses of 40 and 41 kDa (lanes 2 and 3).
316
J.R. Backstrom et ai. / Molecular Brain Research 33 (1995) 311-318
3T3/2C
Tunicamycin N-Glycosidase Neuraminidase:
~"
o~,"©
¢.)
c~
N
¢,)
N - G i y c o s i d a s e F:
"= ~
r"
Choroid
-
+
-
-
+
-
_
_
+
-
_
+
4
5
6
Mass (kD_ 66--
Mass (kDa) 65--
=
mi
51-41 .-40
51-41m
RMg
Htw
~
42 - - 41
w
1
Lane:
2
3
Fig. 6. Analysis of N-linked oligosaccharides associated with 5-H'r2c
1 Iir-,.~
~
A
I~,
"~
T 1:
2
3
4
5
6
1.
,,-,na,~,st,, of N-,,n,,ed ohgo..,a~.~.ha,,t,~s " - . . . . . :'~- associai~d wiih 5-HT2c 1"1~, receptors from 3 T 3 / 2 C cells. Cells were grown for 24 h in the presence of vehicle (lane 1) or tunicamycin at 0.2 /.tg/ml (lane 2) or 2 /.tg/ml (lane 3) to inhibit glycosylation. Cells were also grown in the absence of tunicamycin and the CHAPS-soluble extracts treated with N-glycosidase F for 1 h (lane 5) or 16 h (lane 6) to remove endogenous sugars. The control extract (lane 4) was incubated in the absence of enzyme for 16 h. .P.
To enzymatically remove endogenous oligosaccharides from glycoproteins, CHAPS-soluble extracts were prepared from 3T3/2C cells grown in the absence of tunicamycin. The extracts were treated with N-glycosidase F for 1 or 16 h and then subjected to SDS-PAGE. Immunoreactive proteins on the Western blots were identified with masses of 41 and 42 kDa (Fig. 5, lanes 5 and 6). A protein with a mass of 35 kDa was detected after longer incubations (lane 6), and may represent a proteolytic product of the 5-HT2c receptor. Extracts prepared from the choroid plexus were also treated with N-glycosidase F. lmmunoreactive protein with a mass of 41 kDa was detected from choroid plexus extracts (Fig. 6, lane 6). Longer incubations increased the relative amount of the 41-42 kDa proteins (Fig. 5 and data not illustrated). Since sialic acid (N-acetylneuraminic acid) may be associated with complex N-linked oligosaccharides, extracts prepared from the 5-HT2c receptor cell line and choroid plexus were treated with neuraminidase. The migration of the immunoreactive proteins from the cell line were not noticeably altered after incubation with the enzyme (Fig. 6, lanes 1 and 2). In contrast, treatment of the choroid plexus extracts with neuraminidase reduced the mass of the 51 and 58-62 kDa proteins (lane 4) to 50 and 54-58 kDa (lane 5). It is unlikely that this reduction in mass was due to proteinase activity rather than glycosidase activity since neuraminidase did not affect the migration of
receptors from 3 T 3 / 2 C cells and the choroid plexus. CHAPS-soluble extracts were prepared from 3 T 3 / 2 C cells (lanes 1-3) and the choroid plexus (lanes 4-6) and incubated for 2 h with neuraminidase (lanes 2 and 5) or N-glycosidase F (lanes 3 and 6). The control extracts (lanes 1 and 4) were incubated for 2 h in the absence of enzymes. Identical results to those of lanes 2 and 5 were obtained from samples incubated for 16 h with neuraminidase (data not illustrated).
immunoreactive receptors from 3T3/2C cells (Fig. 6) or the 41 kDa deglycosylated (N-linked) protein from the choroid plexus (Fig. 7, lane 2). The presence of serine/threonine-linked (O-linked) sugar associated with the 5-HT2c receptor was examined with the enzyme O-glycosidase. CHAPS-soluble extracts prepared from the 5-HT2c receptor cell line and choroid plexus were treated with N-glycosidase F to generate 41-42 kDa proteins lacking N-linked sugars (Fig. 7, lane 1). hnmunoreactive proteins with masses of 41-42 kDa were detected from extracts incubated with neuraminidase (lane 2). Similar proteins were also detected from extracts treated with neuraminidase and O-glycosidase (lane 3). Therefore, neither neuraminidase nor neuraminidase a~ad O-glycosidase altered the mass of the immunoreactive proteins lacking N-linked oligosaccharides in this assay.
N - G i y c o s i d a s e F:
+
+
+
Neuraminidase:
-
+
+
-
-
+
O-Glycosidase:
Mass(kDa)
4a
3 T 3 / 2 C cells: Choroid plexus: Lane:
1
2
3
Fig. 7. Analysis of O-linked oligosaccharides associated with 5-HTzc receptors from 3 T 3 / 2 C cells and the choroid plexus. CHAPS-soluble extracts were prepared and incubated with N-glycosidase F for 16 h. Some samples also included neuraminidase (lane 2) or neuraminidase and O-glycosidase (lane 3).
J.R. Backstrom e~ al. / Afolecular Brain Research 33 (1995) 311-318
4. Discussion Antibodies were raised against a region of the puta~:ive third intracellular loop of the rat 5-HT2c receptor and used to identify immunoreactive protein from a 5-HTx. receptor cell line (3T3/2C) and rat brain. The hydrophilic. 19 amino acid peptide used for production and purification of antibodies is not present in the other known 5-HT 2 receptor subtypes, 5-HT2A and 5-HT2~. The anti-2C antibodies were purified from serum using affinity chromatography. Cell lines were used to examine the specificity of anti-2C. In Western blots, anti-2C recognized proteins from CHAPS-soluble extracts of 3T3/2C cells, but not from extracts of 3T3/2A or parental NIH/3T3 cells. The immunoreactive proteins migrated as two distinct regions with masses of 51-52 and 58-68 kDa. Solubilization of immunoreactive receptors from 3T3,/2C cells was saturable and found to be maximal with 5 mM CHAPS per mg protein/ml. The relative abundance of 5-HT2c receptors in rat brain extracts was examined with Western blots. An equal amount of CHAPS-soluble protein was analyzed from selected regions that were previously examined for relative abundance of 5-HT2c receptor rnRNA using Northern blots [10]. The choroid plexus contained the greatest amount of receptor protein, consistent with the results from Northern blots. Immunoreactive receptor was not detected in the cerebellum, frontal cortex, striatum, or hippocampus using the same conditions. The hippocampus, which contains at least 10-fold less 5-HT2c receptor mRNA than the choroid plexus [10], contained at least 7.5-fold less CHAPS-soluble 5-HT2c receptor protein since immunoreactive receptor was detected with 20 /xg protein from choroid plexus extracts, but not with 150 /.tg protein from hippocampus extracts (data not illustrated). In conclusion, the results from Western blots are consistent with the results from Northern blots which suggest that the relative abundance of 5-HT2c receptors is greatest in the choroid plexus. Different masses were found for imrnunoreactive 5-HT2c receptors in extracts from choroid plexus and 3T3/2C fibroblasts. Immunoreactive protein from choroid plexus extracts was detected with masses of 51 and 54-58 kDa, while protein from 3T3/2C cells had masses of 51-52 and 58-68 kDa. These differences between the two sources may be due to cell-type and/or species specific glycosylation (reviewed in [17]). Thus, experiments were undertaken to investigate glycosylation of 5-HT2c receptors. Two experimental approaches were utilized to examine the association of N-linked oligosaccharides with 5-HT2c receptors. In the first set of experiments, 3T3/2C cells were grown in the presence of the nucleoside antibiotic tunicamycin, which inhibits the first step of ,he biosynthetic pathway and prevents the down:stream addition of oligosaccharides to asparagine (reviewed in [6]). Extracts prepared from cells grown in the presence of tunicamycin
317
contained immunoreactive proteins with masses of 40 and 41 kDa, which was significantly less than the 51-52 and 58-68 kDa receptors from control 3T3/2C cells. In the second set of experiments, extracts prepared from cells grown in the absence of tunicamycin were treated with N-glycosidase F to enzymatically remove available N-linked sugars from giycoproteins. N-Glycosidase F (PNGase F) cleaves the bond between asparagine and N-acetylglucosamine [16]. Predominant 41 and 42 kDa proteins were detected from 3T3/2C extracts treated with N-glycosidase F. Differences between the masses of the 40-42 kDa proteins from tunicamycin-treated cells (40-41 kDa) and N-glycosida~c-treated extracts (41-42 kDa) could be due to post-translational modifications of the protein. These variations may occur in the cell between the time of oligosaccharide addition (inhibited with tunicamycin) and functional expression. A 41 kDa immunoreactive protein was also detected from the choroid plexus after treatment with N-glycosidase F. In conclusion, these experiments establish that the immunoreactive 5-HT2c receptors with masses of 51-52 and 58-68 kDa from 3T3/2C cell extracts and 51 and 54-58 kDa from choroid plexus contain N-linked oligosaccharides. Extracts prepared from 3T3/2C cells and choroid plexus were treated with neuraminidase to investigate the presence of sialic acid residues. Neuraminidase did not significantly alter the mass of the 5-HT2c receptors from 3T3/2C cells, but significantly reduced the mass of the 5-HT2c receptors from choroid plexus. The mass of these receptors decreased from 51 and 58-62 kDa to 50 and 54-58 kDa. To determine whether sialic acid was associated with N- or O-linked oligosaccharide chains of the 5-HT2c receptors, extracts were treated with N-glycosidase to remove N-linked sugars and neuraminidase. The 41 kDa protein lacking N-linked sugar migrated with the same mass after treatment with neuraminidase, which suggests that sialic acid was associated with N-linked, rather than O-linked, sugars. Although O-glycosidase did not significantly alter the mass of the deglycosylated (N-linked) receptor, the possibility that 5-HT2c receptors contain O-linked sugar cannot be ruled out. For example, the sugar may not be accessible to O-giycosidase or the decrease in mass after deglycosylation (O-linked) may not be detectable in this assay. In conclusion, data are presented which demonstrate that rat 5-ttT2c receptors from cell and brain extracts contain N-linked oligosaccharides and suggest that 5-HT2c receptors in the choroid plexus associate with sialic acid residues. The rat 5-HT2c receptor contains six sites that fit the consensus sequence of asparagine-linked glycosylation sites. Three of these sites are predicted to be exposed to the extracellular milieu, and therefore more likely to contain sugars. At least one of the extracellular sites may be utilized in vivo since the mass of immunoreactive receptors decreased when cells were grown in the presence of tunicamycin or extracts were incubated with N-glycosidase F. The presence of two potential sites at residues 204 and
318
J.R. Backstrom et at. / Mo!ecular Brain Research 33 (1995) 311-318
205 may ensure that this domain is available for N-linked glyco.,;ylation. Another domain, possibly at asparagine 39, may also be glycosylated since 3T3/2C cells and choroid plexus extracts contained immunoreactive receptors in two distinct regions of SDS-polyacrylamide gels. We are currently developing antibodies to the amino- and carboxyterminal regions of the 5-HT2c receptor and generating point mutants to determine the role of N-linked glycosylation on 5-aT2c receptor function.
Acknowledgements The authors wish to thank Ann Westpl~,l for her expert technical assistance with the cell cultures. This work was supported by Research Grant MH 34007 from the Alcohol, Drug Abuse, and Mental Health Administration and by a grant from Bristol Myers Squibb Corp.
References [1] Barker, E.L. and Sanders-Bush, E., 5-Hydroxytryptaminelc receptor density and mRNA levels in choroid plexus epithelial cells after treatment with mianserin and ( - ) - l - ( 4 - b r o m o - 2 , 5 dimethoxyphenyl)-2-aminopropane, Mol. Pharm., 44 (1993) 725730. [2] Barker, E.L., Westphal, R.S., Schmidt, D. and Sanders-Bush, E., Constitutively active 5-hydroxytryptamine 2C receptors reveal novel inverse agonist activity of receptor ligands, J. Biol. Chem., 269 (1994) 11687-11690. [3] Boess, F.G. and Martin, I.L., Molecular biology of 5-HT receptors, Neuropharmacology, 33 (1994) 275-317. [4] Chatterjee, S., Kwiterovich, P.O. Jr. and Sekerke, C.S., Effects of tunicamycin on the binding and degradation of !o~, density lipoproteins and glycoprotein synthesis in cultured human fibroblasts, J. Biol. Chem., 254 (1979) 3704-3707. [5] Conn, P.J., Sanders-Bush, E., Hoffman, B.J. and Hartig, P.R., A unique serotonin receptor in choroid plexus is linked to phosphafidylinositol turnover, Proc. Natl. Acad. Sci. USA, 83 (1986) 40864088.
[6] Elbein, A.D., Inhibitors of the biosynthesis and processing of N-linked oligosaccharide chains, Annu. Ret.. Biochem., 56 (1987) 497534. [7] Grotewiel, M.S. and Sanders-Bush, E., Regulation of serotonin~A receptors in heterologous expression systems, J. Neurochem., 63 (1994) 1255-1260. [8] Hoffman, B.J. and Mezey, E., Distribution of serotonin 5-HTIc receptor mRNA in adult rat brain, FEBS Lett., 247 (1989) 453-462. [9] Hoyer, D., Clarke, D.E., Fozard, J.R., Hartig, P.R.. Martin, G.R., Mylecharane, E.J., Saxena, P.R. and Humphrey, P.P.A., International union of pharmacology classification of receptors for 5-hydroxytryptamine (serotonin), Pharmacol. Ret,., 46 (1994) 157-203. [10] Julius, D., MacDermott, A.B., Axel, R. and Je.,isell, T.M., Molecular characterization of a functional cDNA encoding the serotonin lc receptor, Science, 241 (1988)558-564. [11] Julius, D., Huang, K.N., Livelli, T.J., Axel, R. and JesseU, T.M., The 5HT2 receptor defines a famil', of structurally distinct but functionally conserved serotonin receptors, Proc. Natl. Acad. Sci. USA, 87 (1990) 928-932. [12] Kaufman, M.J., Hartig, PR. and Hoffman, B.J., Serotonin 5-l-lT2c receptor stimulates cyclic GMP formation in choroid plexus, J. Neurochem., 64 (1995) 199-205. [13] Mengod, G., Nguyen, H.L., Waeber C, L:dbb,:li, H. and Pa!actos, J.M., The di.~t;ibution and cellular localization of the serotonin 1C receptor mRNA in the rodent brain examined bv in situ hybridization histochemistry. Comparison with receptor binding distribution, Neuroscience, 35 (1990)577-591. [14] Molineaux, S.M., Jessell, T.M., Axel, R. and Julius, D., 5-HTlc receptor is a prominent serotonin receptor subtype in the central nervous system, Proc. Natl. Acad. Sci. USA, 86 (1989) 6793-6797. [15] Pompeiano, M., Palacios, J.M. and Mengod, G., Distribution of the serotonin 5-HT 2 receptor family mRNAs: comparison between 5HT2A and 5-HT2c receptors, Mot. Brain Res., 23 (1994) 163-178. [16] Tarentino, A.L., Gomez, C.M. and Plummer, T.H. Jr., Deglycosy!ation of asparagine-linked glycans by peptide:N-glycosidase F, Biochemistr), 24 (1985) 4665-4671. [17] Varki, A., Biological roles of oligosaccharides: all of the theories are correct, G~ycobiology, 3 (1993)97-130. [18] Westphal, R.S. and Sanders-Bush, E., Reciprocal binding properties of 5-hydroxytryptamine type 2C receptor agonists and inverse agonists, Mot. Pharm., 46 (1994) 937-942. [19] Yu, L., Nguyen, H., Lc, H., Bloem, L.J.. Kozak. C.A., Hoffman, B.J., Snutch, T.P., Lester, H.A., Davidson, N. and Liibbert, H., The mouse 5-H~:c receptor contains eight hydrophobic domains and is X-linked, biol. Brain Res., 11 (1991) 143-149.