Professional Documents
Culture Documents
Abstract: Serotonin binding protein (SBP) is present in is lost during secretion as a result of exocytosis ; however,
all neurectodermally derived cells that store serotonin (5- a fraction of the 45-kDa SBP remains bound to the luminal
HT). Three forms of SBP have been detected (68, 56, and surface of the membrane of secretory vesicles . This pro-
45 kDa), and antibodies to SBP that interfere with the tein is exposed to the ambient medium as a consequence
binding of 5-HT react with each of these proteins . The of exocytosis, but is reinternalized when the vesicular
current experiments test two hypotheses : (a) that the 56- membrane is recaptured during vesicle recycling. Key
and 45-kDa forms of SBP are produced by posttransla- Words: Serotonin-Serotonin binding protein-Medul-
tional cleavage of a 68-kDa precursor molecule; and (b) lary thyroid carcinoma-Pulse-chase labeling-Stimu-
that 45-kDa SBP is a constituent of serotonergic secre- lated release-Vesicle recycling.
tory vesicles . Pulse-chase experiments were carried out J. Neurochem. 63, 97-107 (1994) .
using medullary thyroid carcinoma cells as a model.
These neurectodermally derived cells produce 5-HT and
all three forms of SBP. Following pulse labeling for 20
min with L-[s5S]methionine, the cells were incubated in Serotonin binding protein (SBP) is found in seroto-
the presence of an excess of unlabeled L-methionine for nergic neurons of both the central and enteric nervous
0, 30, 60, or 90 min at 37°C . Alternatively, the chase
systems (Gershon and Tamir, 1985 ; Tamir and Gers-
was performed under conditions (20°C, inhibition of ATP
generation) that delay or stop transport of newly synthe- hon, 1990) and is also present in parafollicular cells
sized proteins from the rough endoplasmic reticulum of the thyroid gland (Bernd et al ., 1981 ; Barasch et
through the Golgi apparatus. Following incubation, the al ., 1987, 1988) . These cells are all derived, ontogeneti-
cells were washed and solubilized, and SBP was immu- cally, from neurectoderm . In contrast, SBP is not found
noprecipitated . Radioactive proteins in the immunopre- in other cells that store serotonin (5-HT), but are not
cipitate were electrophoretically resolved and quantified . derived from neurectoderm, such as mast cells, plate-
Immediately after the pulse, each of the three forms of lets, and enterochromaffin cells (Gershon and Tamir,
SBP was found to be labeled with 35 S. The relative pro- 1985). SBP binds 5-HT with great specificity and high
portions of 35 S-labeled 68-, 56-, and 45-kDa SBP re-
affinity in the presence of Fe 2+ ions (KD, - 10 -y M;
mained the same at each interval of chase. These propor-
K2 _ 10-7 M; Liu et al ., 1985). There is evidence
tions were not changed when the chase was carried out
at 20°C or under conditions that blocked the biosynthesis that SBP is present in synaptic and secretory vesicles
of ATP . These observations suggest that each form of (Tamir and Gershon, 1979). In neurons, SBP is subject
SBP is a primary product of translation, that the smaller to fast axonal transport, is concentrated at terminals,
forms of SBP are not produced by cleavage from a larger and some at least is released by nerve stimulation (Jo-
molecule, and that the size of the primary products of nakait et al ., 1979 ; Gershon et al ., 1983) . In parafollicu-
translation is not altered by passage to the Golgi appara- lar cells, SBP immunoreactivity is found in secretory
tus or a post-Golgi compartment. When secretion was
induced, 45-kDa SBP, but not 56- or 68-kDa SBP, was
released to the medium . When antibodies to 45-kDa SBP Resubmitted manuscript received November 12, 1993 ; accepted
were added to the medium at the time secretion was November 24, 1993 .
induced, antibody binding sites appeared as patches on Address correspondence and reprint requests to Dr . H . Tamir at
the cell surfaces . Because of these sites, cells were lysed Division of Neuroscience, New York State Psychiatric Institute, 722
West 168th Street, New York, NY 10032, U .S .A .
when they were stimulated to secrete in the presence of
Abbreviations used : [Ca 2- ], , external calcium ; CCCP, carbonyl
antibodies to 45-kDa SBP and guinea pig complement . cyanide in-chlorophenyl hydrazone ; 5-HT, serotonin ; MEM, mini-
Antibody binding sites disappeared from cell surfaces mum essential medium ; MTC, medullary thyroid carcinoma ; PAGE,
after 20 min, at which time antibodies to SBP were found polyacrylamide gel electrophoresis ; PBS, phosphate-buffered saline;
inside the cells. It is suggested that 45-kDa SBP is pack- RER, rough endoplasmic reticulum ; SBP, serotonin binding protein ;
aged with 5-HT in secretory vesicles . Some 45-kDa SBP SDS, sodium dodecyl sulfate .
97
98 IL TAMIR ET AL .
vesicles (Barasch et al ., 1987; Tamir et al ., 1990) . Two chased for various periods of time in the presence of
forms of SBP, 45 and 56 kDa, have been purified unlabeled methionine. After each period of chase, the
from rat brain (Liu et al ., 1985) . Each of 12 different amount of radioactivity incorporated into the various
monoclonal antibodies raised against these proteins re- forms of SBP was ascertained . In addition, parafollicu-
acts with both forms (Liu et al ., 1990b) ; however, some lar and MTC cells were stimulated to secrete to deter-
polyclonal antisera have been raised, which are type mine whether any of the three forms of SBP are se-
specific (Liu et al ., 1985) . When the 5-HT binding creted. After evidence was obtained that 45-kDa SBP
sites on 45- or 56-kDa SBP are labeled with a photoaf- is secreted with 5-HT, studies were done to determine
finity probe and peptide maps are prepared, similar whether some 45-kDa SBP might also be retained on
patterns are obtained (Liu et al., 1990a). In addition, the plasma membrane of cells and recycled. The data
anti-idiotypic antibodies that recognize 5-HT binding suggest that 45-, 56-, and 68-kDa SBP are translated
proteins react with both 45- and 56-kDa SBP (Tamir independently and that only the 45-kDa SBP is se-
et al., 1991). These observations indicate that 45- and creted ; however, a portion of the vesicular store of
56-kDa forms of SBP share sequence homology, par- 45-kDa SBP remains membrane bound and can be
ticularly in domains that bind 5-HT; however, the recaptured by endocytosis .
forms are not identical. More recently, a third 5-HT
binding protein (68 kDa) was identified that reacts with MATERIALS AND METHODS
each of the anti-SBP monoclonal antibodies and with
some of the polyclonal anti-SBP sera. The 45- and 56- Cells
kDa forms of SBP appear to be highly restricted to 5- Sheep thyroid glands were dissociated with trypsin and
HT-containing cells ; however, the 68-kDa protein is chromatographically purified as described previously (Bernd
also found in neurons that are not known to be seroto- et al., 1981). Essentially, dissociated cells were exposed to
nergic (Barasch et al., 1987; Kirchgessner et al., 1988) . thyroid stimulating hormone and passed through columns of
The relationships of the three forms of SBP to one thyroglobulin coupled to Sepharose beads . This method
takes advantage of the fact that stimulated thyroid follicular
another are not clear. It is possible that a larger precur- cells attempt to phagocytize thyroglobulin, but parafollicular
sor protein is synthesized first and then cleaved after cells do not. The follicular cells are thus retained on the
translation to yield smaller products . The 68- and/or columns while parafollicular cells pass through. MTC cells
56-kDa proteins could thus be precursors of 45-kDa were obtained from Dr . Barry Nelkin (Johns Hopkins Uni-
SBP . Alternatively, each of these proteins may be syn- versity, Baltimore, MD, U.S.A.) .
thesized independently . There is evidence that 56- and
45-kDa SBP may be located in different intracellular Cell culture, radiolabeling, and SBP extraction
compartments (Gershon et al ., 1983). For example, MTC cells were propagated in vitro as described pre-
newly taken up [3H]5-HT forms a complex in situ viously (Tamir et al., 1989) . The amounts of total SBP in
MTC cells were quantified using an enzyme-linked immuno-
mainly with 45-kDa SBP; moreover, 45-kDa SBP is sorbent assay, carried out as described previously (Liu et al.,
the form that predominates in the spinal cord, which 1985). Cultures were metabolically labeled within 3 days
contains serotonergic axons and synapses, whereas 56- after reaching confluence (approximately 1 week in culture).
kDa SBP predominates in the region of the raphe nu- For metabolic labeling of SBP, confluent cultures (5 X 10`'
clei, which is rich in serotonergic nerve cell bodies cells) were grown in low-methionine media and labeled with
(Gershon et al., 1983) . During ontogeny, 56-kDa SBP [ 35 S]methionine (50 pCi/ml) for 20 min . Chase periods were
appears before 45-kDa SBP, but the acquisition of the initiated by removing the labeling medium and briefly wash-
45-kDa protein is coincident with the formation of ing the nnonolayers twice with cold phosphate-buffered sa-
serotonergic synapses (Liu et al ., 1987; Lauder et al., line (PBS) before adding medium containing 50 Ng/ml of
1988; lvgy-May et al., 1994) . nonradioactive i.-methionine . Cells were chased in this me-
dium for 30, 60, or 90 min . At the end of the chase period,
The current study was undertaken to determine cells were placed on ice, rinsed three times with cold PBS,
whether 45-kDa SBP is synthesized independently or and lysed with 400 pl of lysis buffer 11% Triton X-100,
is derived posttranslationally from the larger forms of 0.5% sodium deoxycholate, and 0.1% sodium dodecyl sul-
SBP ; the hypothesis that 45-kDa SBP is packaged with fate (SDS)] in PBS containing proteolytic enzyme inhibitors:
5-HT in secretory vesicles was also tested. Parafollicu- phenylmethyl sulfonyl fluoride (1 mM; Sigma, St. Louis,
lar cells and medullary thyroid carcinoma (MTC) cells, MO, U.S.A.) and leupeptin, pepstatin, antipain, and aprotinin
which are derived from them, were investigated as all at 10 Ng/ml (Sigma) . SBP in cell lysates was immunopre-
model systems . MTC cells have been characterized cipitated as described below . In some experiments, the chase
extensively and shown to be useful models for the was performed in glucose-free media that contained the H+
study of mechanisms relevant to serotonergic neurons ionophore, carbonyl cyanide m-chlorophenyl hydrazone
(CCCP ; 20 pM; Sigma), or at 20°C. The lower temperature
(Tamir et al ., 1989, 1990, 1992) . These cells produce is known to delay or stop transport of newly synthesized
all three forms of SBP, and they synthesize, store, protein from the rough endoplasmic reticulum (RER) to the
secrete, and take up 5-HT. A pulse-chase technique Golgi apparatus (Burgess and Kelly, 1987) . Because this
was used to follow the intracellular processing and process is ATP dependent, it is also inhibited by CCCP.
transport of each of the forms of SBP. MTC cells were SDS-polyacrylamide gel electrophoresis (PAGE) was car-
metabolically labeled with [ 35 S]methionine and then ried out as described by Laemmli (1970) on 8.5% acrylamide
slab gels . Equal amounts of protein were layered in each min at 37°C), the treated proteins were separated by 10%
lane . The positions of labeled bands were visualized by fluo- SDS-PAGE, transferred to nitrocellulose sheets, and stained
rography . Gels were fixed, dried, affixed to Kodak XAR-2 with Coomassie Blue R-250 (Sigma) as described previously
film, and exposed at -70°C. The resulting fluorograms were (Tamir et al ., 1989).
scanned densitometrically (using a UMax 600 scanner) and
analyzed using the NIH image program (version 1 .49) on a Measurement of [35 S]SBP release from
Macintosh Quadra 800 computer . Peaks were identified on metabolically labeled MTC cells
the scans and the proportion of the radiolabeled proteins that MTC cells were labeled with 135 S]methionine by incuba-
migrated with the three forms of SBP was determined by tion overnight as described above. Medium was aspirated
measuring the area under each peak . A portion of each gel and the monolayers were rinsed twice with Ca 21 -free mini-
was electrophoretically transferred to nitrocellulose . Lanes mum essential medium (MEM) . Cells were then incubated
in these blots were either stained with Amido Black IOB at 37°C in Hanks salt solution containing 2 or 5 mM Ca`+.
(Bio-Rad, Richmond, CA, U .S .A .) or used for the immunolo- The medium was collected after 30 min and the radioactivity
calization of SBP. The blots were treated for 1 h with 2 .5% was determined . The harvested cells were washed exten-
nonfat dry milk (Carnation) in PBS (pH 7 .5) containing sively and lysed . [35 S] SBP was immunoprecipitated as de-
0.05% Tween-20, rinsed, and immunostained with mono- scribed below and fractionated by SDS-PACE .
clonal antibodies that recognize all three forms of SBP (Liu
et al ., 1990b) . Immunochemistry
Immunoprecipitation was used to isolate 35 S-labeled SBP.
Effect of CCCP on transport of proteins from the Cell lysates were passed five times through a 19-gauge nee-
dle and centrifuged at 12,000 g for 10 min at 4°C. The
RER through the Golgi complex
supernatant fluid was stored at -70°C until analyzed . For
To confirm that CCCP prevents transport of newly synthe-
the precipitation of labeled SBP released into the media,
sized proteins from RER to Golgi, the effect of CCCP on
the conversion of the high-mannose residues of asparagine- aspirated media were centrifuged (12,000 g) and the superna-
tant fluids were concentrated to a volume of 1 ml (using a
linked glycoproteins to complex oligosaccharides was inves-
tigated. Newly synthesized glycoproteins were pulse labeled Centricon microconcentrator; Amicon, Danvers, MA,
with ['H]mannose and chased in the presence or absence of U.S .A .) . The solutions were then incubated at 4°C overnight
with a monoclonal antibody that recognizes all three forms
CCCP. High-mannose sugars are added to asparagine resi-
dues in the RER and then trimmed and processed further to of SBP (44-10 ; Liu et al ., 1990b) . The resulting immune
complex oligosaccharides in the Golgi apparatus (Kornfeld complexes were collected by precipitation with secondary
and Kornfeld, 1985 ; Kornfeld, 1987). The [3H]mannose la- antibodies (rabbit anti-mouse IgM; 2 h at 4°C) . Precipitates
bel, therefore, is added only in the RER and cannot render were washed twice with lysis buffer. The final pellets were
a complex oligosaccharide radioactive unless the labeled gly- suspended in SDS sample buffer, resolved by SDS-PAGE,
coprotein moves into the Golgi apparatus, which is where and quantified .
the terminal sugars of complex oligosaccharides are added. Immunocytochemistry
Wheat germ agglutinin binds to complex oligosaccharides, SBP immunoreactivity was located as described pre-
but not to high-mannose sugars ; wheat germ agglutinin bind- viously (Tamir et al ., 1989) . In brief, cells were fixed with
ing, therefore, was utilized to isolate glycoproteins with com- 4% formaldehyde and permeabilized with 0.2% Triton X-
plex oligosaccharides . MTC cells (6 x 106) were incubated 100 (Sigma). Polyclonal antibodies to 45-kDa SBP were
for 30 min at 37°C with 50 kCi/ml of [3H]mannose (15 Ci/ applied and sites of immunoreactivity were visualized using
mmol ; American Radiolabel Co ., St. Louis, MO, U.S .A .) . goat anti-rabbit IgG coupled to rabbit horseradish peroxidase
Cells were then rinsed with PBS and chased for 90 and 180 (Kirkegaard and Perry, Gaithersburg, MD, U .S .A .) . Peroxi-
min with or without CCCP (20 yM). Cells were washed dase activity was visualized with 3,3'-diaminobenzidine and
again with PBS, collected by centrifugation, and lysed (30 H2O, . In control experiments, nonimmune serum was substi-
min at 4°C) in buffer (25 mM HEPES, pH 7.0 ; 1 % Triton tuted for primary antibodies .
X-100; 0.2% deoxycholate ; and the proteolytic enzyme in-
hibitors mentioned above) . Cell debris was removed by cen- Retention and membrane binding of SBP
trifugation (10 min; 12,000 g) . The lysate (200 Fig of protein) Isolated parafollicular cells or MTC cells were plated onto
was then incubated with wheat germ agglutinin coupled to glass cover slips coated with poly-tAysine (12.5 Ng/ml ;
agarose beads (Sigma) for 17 h at 4°C. The beads were Sigma) and cultured overnight in MEM supplemented with
collected by centrifugation and washed with 25 mM HEPES 10% fetal bovine senun (Tamir et al ., 1990) . Cells were
buffer, and the radioactivity of the glycoproteins bound to stimulated to secrete by the addition of 5 .0 mM Ca 2+ or
the beads was measured by liquid scintillation. thyroid stimulating hormone (10.0 mU/ml; NIH) . To deter-
mine whether 45-kDa SBP appears on the plasma membrane
Effect of phospholipase C on the apparent of secreting cells, secretion-dependent uptake of antibodies
molecular weight of the three forms of SBP to 45-kDa SBP was investigated . Secretogogues were added
Posttranslational modification of a single protein could to cells simultaneously with partially purified antibodies to
give rise to products of different size, for example, by adding 45-kDa SBP (y-globulin fraction) and incubated for 20 min
carbohydrate or lipid moieties . SBP is not glycosylated (Ger- at 37 °C. As a control, secretogogues were added with preim-
shon and Tamir, 1985) . To determine whether one or more mune serum and incubated similarly. The stimulated cells
forms of SBP is covalently attached to phosphatidylinositol, were washed once with PBS and allowed to incubate for an
a crude preparation that contained all three forms of SBP additional 30 min at 37°C . The cells were then washed three
(500 N,g) in 0.02 M Tris-HCI, pH 7.0, containing 0.1 M NaCl times with PBS, fixed with 4% formaldehyde in PBS for 20
and 0.0025 MMgCl2, was treated with phospholipase C (200 min at room temperature, and permeabilized with 0.2% Tri-
p,,g ; 13 U/mg ; Sigma) . Following exposure to the enzyme (60 ton X-100. A biotinytated goat anti-rabbit antibody (1 :200 ;
(Fig . 4B) . These observations suggest that 45-kDa SBP MTC cells lyse when they are stimulated to
is secreted when MTC cells are stimulated with in- secrete in the presence of antibodies to SBP plus
creased [Ca 2 ' 1 ' . complement
If, as the experiments described above suggest, anti-
A fraction of the secreted SBP remains bound to bodies bind to SBP that appears on cell surfaces as a
the plasma membrane result of secretion, then MTC cells should be subject
to lysis when they are stimulated to secrete in the
The fact that 45-kDa SBP is secreted implies that
presence of antibodies to SBP upon addition of com-
the protein is present in secretory vesicles . MTC cells
plement. This prediction was tested . The live/dead via-
secrete 5-HT as well as SBP (Tamir et al ., 1990) . Fol-
bility test was used as a probe to detect cell lysis . As
lowing secretion, recycled vesicles can be restocked
cells lyse in the presence of ethidium homodimer, they
with 5-HT, which is synthesized in the cytosol and
become permeable and intensely orange-red fluores-
can be transported into secretory vesicles . No such
cent . Resting MTC cells were found to be stable in
mechanism is available for a protein, such as SBP . The
vitro when they were exposed to antibodies to SBP
hypothesis was thus tested that a portion of the SBP
alone, to complement alone, or to the complement plus
ordinarily present in vesicles is bound to the inner face
antibodies to SBP (Fig . 6A) . Secretion was induced
of the vesicular membrane, remains bound to the cell
with increased [Ca'-+1, (5 .0 rnM) or cholera toxin (10
surface following exocytosis, and is recaptured with
the vesicular membrane by endocytosis during the re- h, g/ml) . Cholera toxin was used for these experiments
because it has been shown previously to stimulate
covery period . Antibodies to SBP were added to cul-
MTC cells potently and rapidly to secrete (Tamir et
tures of living cells as a probe to detect the appearance
al ., 1990) . Both secretogogues led within 15 min to
of surface-bound SBP . When nonstimulated MTC cells
the appearance of lysed cells (Fig . 6B and C) . When
were exposed to antibodies to SBP, no immunoreactiv-
cholera toxin was employed as the secretogogue, 15
ity was detected (Fig . 5A) . Because the cells were not
± 5% of the total number of cells was lysed . In con-
permeabilized, antibodies would bind to the cells only
trast, no lysis was detected when either secretogogue
if immunoreactive material were present on the cell
was applied in the absence of antibodies to SBP and
surface. Secretion was evoked by raising the (Ca-+1,
complement (not illustrated) .
to 5 mM. Under these conditions, MTC cells became
MTC cells were exposed to repeated cycles of se-
intensely SBP immunoreactive (Fig . 5B) . The appear-
cretogogue stimulation in the presence of antibodies
ance of the cell surface SBP immunoreactivity was
to 45-kDa SBP and complement. Following each cycle
punctate . As there was no detectable immunoreactivity
of complement-mediated lysis, the surviving cells were
on the cell surfaces when the cells were stimulated
permitted to regrow to confluency . After 10 such cy-
in the presence of preimmune serum (Fig . 5D), the
cles, immunocytochemical examination revealed that
secretion-induced immunostaining of the cell surface
the resulting cells (Fig . 7A) contained much less SBP
was specific . The stimulated cells were allowed to re- immunoreactivity than did control cells that had not
cover for 20 min . After this period, there was a pro- been subjected to cycles of immunolysis (Fig . 7B) .
nounced decrease in the immunoreactivity of the sur- Enzyme-linked immunosorbent assay was used to esti-
face (not illustrated) . Cells were fixed and permeabil-
mate the degree to which the SBP content of MTC
ized at this stage . The permeabilized cells were cells surviving after repeated cycles of immunolysis
exposed to secondary antibodies to determine whether had been reduced ; the 45-kDa SBP content of the ex-
the cells had internalized antibodies to SBP during the perimental cells was 48 - 8% of controls . The 5-
stimulation and subsequent rest periods . Immunoreac- HT concentration in the MTC cells remaining after
tivity was observed (Fig . 5C) . These experiments were repeated cycles of complement-mediated lysis was 50
repeated with primary parafollicular cells isolated from + 10% of controls . It is concluded that subjection of
sheep thyroid glands to determine whether the secre- MTC cells to repeated cycles of lysis secondary to
tion-induced appearance of cell surface SBP immuno- stimulation to secrete in the presence of antibodies to
reactivity also occurred in a nontumor cell of the type SBP and complement selects a variant line that is both
from which MTC cells are derived . Relatively little SBP and 5-HT deficient .
SBP immunoreactivity was found on the surfaces of
resting parafollicular cells . Stimulation with increased
DISCUSSION
[Ca" 1, (5 .0 mM), however, caused immunoactive SBP
to appear on the plasma membrane . When either MTC The present study was undertaken to test whether the
cells or parafollicular cells were stimulated with in- smaller forms of SBP are derived by posttranslational
creased [Ca 2 ~]~ in the presence of an unrelated anti- proteolysis from a larger precursor and to test the hy-
body, the cell surface-immunoreactive material did not pothesis that 45-kDa SBP is secreted . In view of the
appear (not illustrated) . These observations are consis- known similarities among the proteins, the data ob-
tent with the idea that membrane-bound SBP becomes tained were surprising because they suggested that
externalized as a result of secretion and is reinterna- none of the three forms of SBP is derived from i
lized during the postsecretion recovery period . larger precursor molecule. The relative proportions of'
I. New . . hem
., Vol. 63, No. l, 1994
SEROTONIN BINDING PROTEIN 10 3
radioactive 45-, 56-, and 68-kDa SBP did not change These data suggest that each of the three isoforms of
as a function of time in chase medium following their SBP is probably a primary product of translation . The
biosynthesis during a pulse-labeling incubation in the possibility that a large and still unidentified molecule
presence of [ 35S]methionine . If 68-kDa SBP or another, is cleaved to yield each of the three isoforms of SBP
still larger, molecule had been a precursor of the in less than 20 min, the time allotted for pulse labeling
smaller forms of SBP, then the proportional radioactiv- in the presence of [ 35 S]methionine, has not been ex-
ity of the smaller molecules would have been expected cluded ; however, the data imply that 68-kDa SBP does
to increase during the chase period, whereas that of not give rise to 56- or 45-kDa SBP, and that 56-kDa
the larger molecules would have been expected to de- SBP is not a precursor of the 45-kDa protein . These
crease . Instead, the relative proportions of each of the conclusions were supported by observations of the ef-
three forms of SBP remained relatively constant fects of addition of CCCP to the chase medium and
throughout a 90-min incubation in chase medium . by subjecting pulse-labeled cells to a chase at 20°C.
Each of these perturbations is known to interfere with contrast, no secretion of 35 S- labeled 56- or 68-kDa SBP
the transfer of newly synthesized protein from the RER could be detected under these conditions . These data
to the Golgi apparatus . The effectiveness of CCCP in suggest that 45-kDa SBP is probably synthesized in
this regard was verified by ascertaining that it com- the RER of MTC cells and transported via the secretory
pletely blocked the conversion of high-mannose sugars pathway of intracellular transport to be packaged in
to complex oligosaccharides, a step that depends on secretory vesicles . These results are compatible with
transport from RER to Golgi (Kornfeld and Kornfeld, prior electron microscopic immunocytochemical ob-
1985 ; Kornfeld, 1987) . Neither CCCP nor incubation servations that the immunoreactivities of 45-kDa SBP
at 20°C altered the proportions of radioactive 68-, 56-, and 5-HT are found in the same secretory vesicles in
or 45-kDa SBP found in the cells . These data suggest both MTC cells and parafollicular cells (Barasch et al.,
that the size of newly synthesized SBP is not altered 1987 ; Tamir et al., 1990) . The failure of increased
as a result of processing in a compartment reached as [Ca21], to cause the cells to secrete 56- or 68-kDa
a consequence of intracellular transport through the SBP may be due to a difference in the intracellular
vacuolar apparatus of the cells . Finally, incubation of cornpartmentation of the three isoforms of SBP. In
68-, 56-, or 45-kDa SBP with phospholipase C failed contrast to 45-kDa SBP, it seems likely that the 56-
to alter the apparent molecular weight of any of the and 68-kDa isoforms are probably not packaged in
proteins ; furthermore, none of the forms of SBP was 5-HT-containing secretory vesicles and thus are not
glycosylated . It can be concluded that the three iso- released by stimuli that induce the secretion of 5-HT .
forms of the SBP are products of different genes, or Whether they enter the cisternal space at all remains
that they are synthesized as the result of alternative to be determined. It has been shown previously that
splicing of mRNA transcribed from a single gene. In SBP is secreted from electrically stimulated enteric
considering these observations, it should be recalled neurons and that this release is Ca`+ dependent (Jona-
that secretion by parafollicular and MTC cells is regu- kait et al ., 1979) . Enteric neurons, moreover, contain
lated and that no secretogogue was added during any only 45-kDa SBP [the larger isoforms cannot be de-
of the three pulse-chase experiments . The proportions tected in the bowel (Gershon et al., 1983)] ; therefore,
of the three forms of SBP were thus not altered by it is likely that in serotonergic neurons, as well as in
secretion, as they would have been unless all three the MTC paraneurons, 45-kDa SBP is a component of
were secreted in equal amounts . synaptic vesicles that is secreted together with 5-HT
Studies of the regulated secretion of the three iso- as a result of exocytosis . This idea would explain the
forms of SBP provided evidence that only 45-kDa SBP observation that 45-kDa SBP and not 56-kDa SBP is
is secreted . Exposure of metabolically labeled MTC concentrated in the terminals of serotonergic neurons
cells to increased [Ca"],, which is known to induce in the brain, and explain why 45-kDa SBP is subject
the cells to secrete 5-HT (Tamir et al ., 1990), caused to fast axonal transport (Gershon et al ., 1983) .
them to secrete 35 S-labeled 45-kDa SBP as well . In Costorage of 45-kDa SBP with 5-HT in the same vesi-
cles of serotonergic neurons would also account for the crete. The predicted secretion-dependent appearance
observation that newly taken up [3H]5-HT is rapidly of 45-kDa SBP was detected by stimulating cells to
complexed with 45-kDa SBP in the CNS in situ . secrete in the presence of antibodies to the protein. The
The secretion of a large protein, such as SBP, from fact that 45-kDa SBP was bound to the cell surfaces of
a neuron raises a question that does not arise with secreting cells was demonstrated by both immunocyto-
respect to a small-molecule neurotransmitter, such as chemistry and complement-mediated lysis. Patches of
5-HT . This question is how synaptic vesicles or their 45-kDa SBP immunoreactivity appeared as a result of
membranes can be recycled . Synaptic vesicle recycling stimulation, and the cells were lysed when they were
has not been studied in serotonergic neurons, but it is stimulated in the presence of antibodies to 45-kDa SBP
known to occur in other systems (see Südhof and Jahn, and complement. Furthermore, when time was allotted
1991 ; Kelly, 1993), including the motor nerve terminal for endocytosis to occur after the cells were stimulated
at the neuromuscular junction, where it has been inves- to secrete, the antibodies to SBP that were present in
tigated extensively (Heuser and Reese, 1981 ; Koening the medium were taken up by the cells . These observa-
and Ikeda, 1989) . Vesicle recycling thus would be ex- tions are compatible with the idea that the secretory
pected to occur in serotonergic cells as well . Recycling vesicles of MTC and parafollicular cells, like synaptic
involves the recapture by endocytosis of the synaptic vesicles, recycle and that a fraction of the vesicular
vesicle membrane that has fused with the plasma mem- content of 45-kDa SBP is retained as a result of binding
brane of the nerve terminal during exocytosis. The to the luminal face of the vesicular membrane. Re-
retrieved membrane is then reutilized to form new syn- maining to be demonstrated is the proportion of 45-
aptic vesicles that have to be restocked with a neuro- kDa SBP secreted versus that retained, and the mecha-
transmitter. Integral membrane proteins that are com- nism of 45-kDa SBP binding . In several other systems
ponents of the synaptic vesicle membrane are retrieved also, components of synaptic vesicles are known to be
by endocytosis together with the membrane and thus partially released and partially retained during exo-
do not have to be resynthesized or added to the re- cytosis . These molecules include dopamine /3-hydrox-
formed vesicles. The neurotransmitter 5-HT is synthe- ylase (Viveros et al., 1968; De Potter et al ., 1969;
sized in the cytosol, but is transported into vesicles by Sabban et al., 1987), carboxypeptidase E (Fricker et
a reserpine-sensitive carrier protein that recently has al., 1990), the prehormone convertases, PC] and PC2
been cloned (Liu et al., 1992). How a soluble protein (Kirchmair et al., 1992), and chromogranins A and B
of the vesicular interior could be added to a recycled (Settleman et al ., 1985 ; Pimplikar and Huttner, 1992).
vesicle is not at all clear. Two possibilities can be MTC cells normally are heterogeneous with respect
envisioned. Because proteins enter the vacuolar space to their content of 45-kDa SBP. It follows, therefore,
of cells cotranslationally, 45-kDa SBP cannot, like 5- that subjection of MTC cells to repeated cycles of
HT, be transported through the vesicular membrane . secretion-dependent, complement-mediated lysis can
Addition of newly synthesized 45-kDa SBP would re- be used to select a variant cell line that is 45-kDa SBP-
quire recycling vesicles to fuse with virgin vesicles. deficient . The technique of repeated cycles of activity-
As there are no ribosomes or RER in nerve terminals, dependent, complement-mediated lysis previously has
such a mechanism would necessitate the transport of been used successfully to obtain a synaptotagmin-
vesicles containing 45-kDa SBP from the cell body. deficient variant cell line (Kasai et al., 1992) . Synapto-
An alternative mechanism involving the recycling of tagmin is a vesicular membrane protein . MTC cell
45-kDa SBP along with the vesicular membrane would variants deficient in 45-kDa SBP were obtained by
make the system more efficient . The 45-kDa SBP origi- repeating cycles of' secretion-dependent, complement-
nally present in 5-HT-storing secretory vesicles could mediated lysis. These variants were found to have
continue to be utilized if a fraction of that protein were lower than control levels of 5-HT. This observation
retained at the time of exocytosis . Retention of 45-kDa suggests that 45-kDa SBP may be important in the
SBP could be accomplished if a fraction of the total vesicular storage of 5-HT.
present in vesicles were to be bound to the luminal
face of the vesicular membrane. If this binding occurs, Acknowledgment : This study was supported by NIMH
then patches of 45-kDa SBP would be expected to grant 37575, NS grant 12969, and DK grant 19743 .
appear on the cell surface when cells are stimulated
to secrete. This exteriorized 45-kDa SBP would be REFERENCES
reinternalized when the vesicular membrane is re-
Barasch J . M ., Tamir H ., Nunez E . A ., and Gershon M . D . (1987)
trieved by endocytosis . The process would result in a Serotonin-storing secretory granules from thyroid parafollicular
progressive decline in the quantity of vesicular 45-kDa cells . J. Neurosci. 7, 4017-4033 .
SBP as vesicles recycle repeatedly ; however, vesicles Barasch J . M ., Gershon M . D ., Nunez E . A ., Tamir H ., and M-
have a finite lifetime and there is probably a limit to Awgati Q . (1988) Thyrotropin induces the acidification of the
the number of times recycling can occur. secretory granules of parafollicular cells by increasing the chlo-
ride conductance of the granular membrane . J. Cell Biol . 107,
Experiments were carried out to test the hypothesis 2137-2147 .
that 45-kDa SBP is recycled, as well as released, when Bernd P., Gershon M . D ., Nunez F_. A ., and Tamir H . (1981) Separa-
parafollicular cells or MTC cells are stimulated to se- tion of dissociated thyroid follicular and parafollicular cells:
association of serotonin binding protein . J . Cell Biol . 88, 499- Liu K .-P., Tamir H ., Hsiung S ., Adlersberg M ., and Gershon M . D.
508 . (1987) Prenatal development of serotonin binding protein in
Burgess T . L . and Kelly R . B . (1987) Constitutive and regulated relation to other transmitter-related characteristics of central
secretion of proteins . Annu. Rev. Cell Biol. 3, 243-293 . serotonergic neurons . Dev. Brain Res. 32, 31-41 .
De Potter W . P., De Schaepdryver A . F ., Moerman E. J ., and Smith Liu K.-P ., Tamir H ., and Adlersberg M . (1990a) Photoaffinity label-
A . D . (1969) Evidence for the release of vesicles-proteins to- ing of the two forms of serotonin binding protein : peptide map-
gether with adrenaline upon stimulation of the splenic nerve . ping of the binding sites . ./. Neurochern. 54, 963-970.
J. Plo,siol. (Load.) 204, 102P-104P . Liu K .-P., Yu P .-Y ., Hsiung S . H ., Kirchgessner A . L ., Gershon
Fricker L . D ., Das B ., and Angeletti R . H . (1990) Identification of M . D ., and Tamir H . (1990b) Preparation and characterization
the pH-dependent membrane anchor of carboxypeptidase E (EC of monoclonal antibodies to serotonin binding protein . ./. Neuro-
3 .4 .17 . 10) . J. Biol . Chem. 265, 2476-2482. chem . 55, 1013-1021 .
Gershon M . D . and Tamir H . (1985) Peripheral sources of serotonin Liu Y ., Peter D ., Roghani A ., Schuldiner S ., Prive G . G ., Eisenberg
and serotonin-binding proteins, in Serotonin and the Cardiovas- D ., Brecha N ., and Edwards R . H . (1992) A cDNA that sup-
cular Svstern (Vanhoutte P . M ., ed), pp . 15-26 . Raven Press, presses MPP` toxicity encodes a vesicular amine transporter .
New York . Cell 70, 539-551 .
Gershon M . D ., Liu K .-P., Karpiak S . E ., and Tamir H . (1983) Melander A . and Sundler S . (1972) Interactions between catechol-
Storage of serotonin in vivo as a complex with serotonin binding amine, 5-hydroxytryptamine and thyroid stimulating hormone.
protein in central and peripheral serotonergic neurons . J. Neu- Endocrinology 90, 188- 193 .
rosci. 3, 1901-1911 . Moore P . L ., MacCoubrey 1 . C ., and Haugland R . P . (1990) A rapid
Heuser ,1 . H . and Reese T . S . (1981) Structural changes after trans- pH-insensitive, two-color fluorescence viability (cytotoxicity)
mitter release at the frog neuromuscular junction . J. Cell Biol. assay . (Abstr .) J. Cell Biol . 111, 304, PN 10579 .
88, 564-580 . Pimplikar S . W . and Huttner W . B . (1992) Chromogranin B (secreto-
Hirsch P . F. and Munson P. L. (1969) Thyrocalcitonin . Phvsiol. Rev. granin 1), a secretory protein of the regulated pathway, is also
49, 548-622 . present in a tightly membrane-associated form in PC12 cells .
Ivgy-May N ., Tamir H ., and Gershon M . D . (1994) Synaptic proper- ./. Biol. Chem. 267, 4110-4118 .
ties of serotonergic growth cones in developing rat brain . J. Sabban E. L ., Kuhn L . J ., and Levin B . E . (1987) ln vivo biosynthesis
Neurosci. 14, 1011-1029. of two subunit forms of dopamine ß-hydroxylase in rat brain .
Jonakait G . M ., Tamir H ., Gintzler A . R ., and Gershon M . D . (1979) Neuroscience 7, 192-200 .
Release of serotonin and its binding protein from enteric neu- Settleman J ., Nolan J ., and Angeletti R . (1985) Chromogranin, an
rons . Brain Res. 174, 55-69 . integral membrane protein . ,/. Biol. Chem . 260, 1641-1644 .
Kasai Y . S ., Yoshida A ., Sato K ., Hoshino T ., Ogura A ., Kondo S ., Südhof T . C . and Jahn R . (1991) Proteins of synaptic vesicles in-
Fujimoto Y ., Kuwahara R ., Kato R ., and Takahashi M . (1992) volved in exocytosis and membrane recycling . Neuron 6, 665-
Neurotransmitter release from synaptotagrnin-deficient clonal 677 .
variant of PC 12 cells . Science 256, 1820-1823 . Tamir H . and Gershon M . D . (1979) Storage of serotonin and seroto-
Kelly R . B . (1993) Storage and release of neurotransmitters . Cell nin binding protein in synaptic vesicles . J. Neurochem. 33, 35-
72, 43-53 . 44 .
Kirchgessner A . L ., Gershon M . D ., Liu K .-P ., and Tamir H . (1988) Tamir H . and Gershon M . D . (1990) Serotonin-storing secretory
Storage of serotonin binding protein with serotonin in the rat vesicles, in The Neuropharmacology of Serotonin (Whittaker-
CNS . J. Neurosci. 8, 3879-3890. Azmitia P . M . and Peroutka S . J ., eds), pp . 53-67 . New York
Kirchmair R ., Gee P., Angeletti R . H ., Laslop A ., Fischer-Colbrie Academy of Sciences, New York.
R ., and Winkler H . (1992) Immunological characterization of Tamir H ., Liu K .-P ., Payette R . F ., Hsiung S . H ., Adlersberg M .,
the endoproteases PC I and PC2 in adrenal ehromaffin granules Nunez E. A ., and Gershon M . D. (1989) Human medullary
and in the pituitary gland . FEBS Lett . 297, 302-305 . thyroid carcinoma : characterization of the serotonergic and neu
Kocning J . H . and Ikeda K . (1989) Disappearance and reformation ronal properties of a neurectodermally derived cell line . J. New
of synaptic vesicle membrane upon transmitter release observed rosci. 9, 1199- 1212 .
under reversible blockage of membrane retrieval . ./ . Neurosci. Tamir H ., Liu K .-P ., Hsiung S .-C ., Adlersberg M ., Nunez E . A ., and
9, 3844-3860 . Gershon M . D . (1990) Multiple signal transduction mechanisms
Kornfeld R . and Kornfeld S . (1985) Assembly of asparagine-linked leading to the secretion of 5-hydroxytryptamine by MTC cells,
oIigosaccharides . Annu. Rev. Biochem. 54, 631-664. a neurectodermally derived cell line . J. Neurosci. 10, 3743-
Kornfeld S . (1987) Trafficking of lysosomal enzymes. FASEB J . l, 3753 .
462-468 . Tamir H ., Liu K .-P ., Hsiung S .-C ., Yu P . Y ., Kirchgessner A . L.,
Laemmli U . K . (1970) Cleavage of structural proteins during the and Gershon M . D . (1991) Identification of serotonin receptors
assembly of the head of bacteriophage T4 . Nature 227, 680- recognized by anti-idiotypic antibodies . .l. Neurochem. 57, 930-
685 . 942 .
Lauder J . M., "Tamir H ., and Sadler T . W . (1988) Serotonin and Tamir H ., Hsiung S . H ., Yu P . Y ., Liu K .-P ., Adlersberg M ., Nunez
morphogenesis . 1 . Sites of 5-HT uptake and binding protein A . E ., and Gershon M . D . (1992) Serotonergic signalling be-
immunoreactivity in the mid gestation mouse embryo . Develop- tween thyroid cells : protein kinase C and 5-HT, receptors in
ment 102, 709-720 . secretion and action of serotonin . Synapse 12, 155-158 .
Liu K .-P ., Gershon M . D ., and Tamir H . (1985) Identification, puri- Viveros D . H ., Arqueros L ., and Kirshner N . (1968) Release of
fication, and characterization of two forms of serotonin binding catecholamine and dopamine 0-hydroxylase from the adrenal
protein from rat brain . J . Neurochem. 44, 1289-1301 . medulla . Lifé Sci. 7, 609-618 .