| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Laboratory of Signal Transduction, Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Japan
Address all correspondence and requests for reprints to: Dr. Fumikazu Okajima, Laboratory of Signal Transduction, Institute for Molecular and Cellular Regulation, Gunma University, 3–39-15 Showa-machi, Maebashi 371, Japan.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
S), but not in its absence.
Guanosine 5'-O-(2-thiodiphosphate) inhibited the S1P-induced
GTPS-dependent activation of the enzyme. To characterize the
signaling pathways, especially receptors and G proteins involved in the
S1P-induced responses, cross-desensitization experiments were
performed. Under the conditions where homologous desensitization
occurred in S1P-, lysophosphatidic acid (LPA)-, and bradykinin-induced
induction of Ca2+ mobilization, no detectable
cross-desensitization of S1P and bradykinin was observed. This suggests
that the primary action of S1P in its activation of the
PLC-Ca2+ system was not the activation of G proteins common
to S1P and bradykinin, but the activation of a putative S1P receptor.
On the other hand, there was a significant cross-desensitization of S1P
and LPA; however, a still significant response to S1P (50–80% of the
response in the nontreated control cells) was observed depending on the
lipid dose employed after a prior LPA challenge. S1P also inhibited
cAMP accumulation in a PTX-sensitive manner. We conclude that S1P
stimulates H2O2 generation through a
PLC-Ca2+ system and also inhibits adenylyl cyclase in
FRTL-5 thyroid cells. The S1P-induced responses may be mediated partly
through a putative lipid receptor that is coupled to both PTX-sensitive
and insensitive G proteins. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In thyroid cells, hydrogen peroxide (H2O2) generation is an important process in thyroid hormone synthesis (11). A number of studies have shown that receptor agonists (such as TSH and P2-purinergic receptor agonists) and also Ca2+-mobilizing agents (such as Ca2+ ionophore and thapsigargin) stimulated H2O2 generation depending on [Ca2+]i in thyroid cells, including FRTL-5 cells (12, 13, 14, 15). Thus, agents that increase [Ca2+]i can be expected to induce H2O2 generation in the cells. In the previous studies, the effects of S1P on cell functions other than cell proliferation and cell motility have not been well characterized. In the present report, therefore, we examined the effect of S1P on H2O2 generation and its regulatory mechanism, especially focusing on the lipid-induced Ca2+-mobilizing mechanism in FRTL-5 thyroid cells. We found that exogenous S1P stimulated H2O2 generation in the mediation of Ca2+ indirectly mobilized through PLC-catalyzed IP3 production rather than Ca2+ mobilized by a direct interaction of S1P with the intracellular Ca2+ pool. We further examined the early signaling mechanisms of S1P in both intact cells and a cell-free system. Our results suggest that S1P interacts with GTP-binding protein (G protein)-coupled receptors, resulting not only in activation of PLC but also inhibition of adenylyl cyclase.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
S) and
5'-O-(2-thiodiphosphate) (GDPßS) were obtained from
Boehringer Mannheim (Indianapolis, IN); fura-2/AM was purchased from
Dojindo (Tokyo, Japan); and myo-[2-3H]inositol (23.0
Ci/mmol) was obtained from DuPont-New England Nuclear (Boston, MA). S1P
was prepared by treatment of SPC with phospholipase D as previously
described (10). Pertussis toxin (PTX) was generously provided by Dr. M.
Ui of the Institute of Physical and Chemical Research (RIKEN, Wako,
Japan), and U73122 and U73343 were purchased from Upjohn Co.
(Kalamazoo, MI). For the RIA of cAMP, a Yamasa cAMP assay kit was used,
which was a gift from Yamasa Shoyu Co. (Choshi, Chiba, Japan). The
sources of all other reagents were described previously (10, 15, 16, 17, 18, 19).
Cell culture
FRTL-5 thyroid cells, a continuous line of functional epithelial
cells from normal rat thyroid (20), were provided by Interthyr Research
Foundation (Baltimore, MD). The cells were grown in Coon’s modified
F-12 medium supplemented with 5% calf serum (Life Technologies, Grand
Island, NY) and a 6-hormone mixture (6H) on 12-well plates for
determination of H2O2 and cAMP responses,
unless otherwise stated, and on 10-cm dishes for determination of
inositol phosphate and [Ca2+]i responses and
for preparation of [3H]inositol-labeled membranes as
previously described (15, 17, 18, 19). In some experiments, for measurement
of H2O2 (Fig. 3C
), cells were also cultured on
10-cm dishes. When the cells had become 90% confluent, the culture
medium was changed to the fresh medium containing 5% calf serum and 5H
(without TSH), and the cells were cultured for another 24 h for
measurements of H2O2, cAMP, and
[Ca2+]i. For inositol phosphate response in
intact cells and PLC assay in cell-free system, the culture medium was
changed to the inositol-free DMEM medium (Life Technologies) containing
5% calf serum, 5H (without TSH), and [3H]inositol (2.5
µCi/ml), and the cells were cultured for another 24 h. Where
indicated, PTX (50 ng/ml) was added to the medium 24 h before the
experiments.
|
, 2 min
before addition of the test agents, U73122, U73343, or its vehicle
[dimethylsulfoxide (DMSO)] was added to the incubation medium. For
termination of the reaction, the cell suspensions (0.5 ml) were
transferred to the tube containing 1 ml CHCl3-methanol-HCl
(100:100:1). [3H]IP2 plus
[3H]IP3 were separated on Dowex 1x8 formate
columns as previously described (17, 18). The radioactivity of the
trichloroacetic acid (5%)-insoluble fraction was measured as the total
radioactivity incorporated into the cellular inositol lipids (17, 18).
Where indicated, the results were normalized to 105 dpm of
the total radioactivity.
|
, the cells cultured on 10-cm dishes were
harvested and treated as described for measurements of
[Ca2+]i and inositol phosphate production.
The cell suspensions were then preincubated at 37 C for the first 8 min
with the HEPES-buffered medium and for another 2 min with the same
medium containing the reaction mixture for measurement of
H2O2 generation in the presence of U73122,
U73343, or DMSO. S1P was added to the medium, and the cells were
incubated for an additional 10 min. Other experimental conditions were
described previously (15).
Measurement of cAMP content
The cells were washed once with HEPES-buffered medium, then
preincubated for 10 min at 37 C with the same medium. The medium was
replaced with fresh medium containing adenosine deaminase (0.5 U/ml),
3-isobutyl-1-methylxanthine (100 µM), TSH (10
nM), and the indicated dose of S1P, and the cells were
incubated for 10 min at 37 C. Termination of the reaction and
measurement of cAMP content were performed as previously described
(18).
Membrane preparation and assay of PLC
The membranes were prepared from the
[3H]inositol-labeled cells, and the PLC assay was
performed as previously described (19). The data were normalized to
105 dpm of the radioactivity in membranes.
Data presentation
All experiments were performed in duplicate or triplicate. The
results of multiple observations are presented as the mean ±
SE or as representative results from more than three
different batches of cells unless otherwise stated.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows a dose-dependent effect of S1P on
H2O2 generation in FRTL-5 thyroid cells.
Because of the insolubility of S1P at higher doses, we could not
examine the effect of more than 30 µM S1P in the present
study. S1P significantly stimulated H2O2
generation at the minimal dose of 0.03 µM. The effect of
PTX on the S1P-induced response was also examined in this figure. PTX
treatment was ineffective in the response to S1P at doses lower than 10
µM, whereas the toxin slightly, but significantly,
inhibited the response to a higher dose (30 µM). Because
an increase in [Ca2+]i is an important factor
for agonist-induced H2O2 generation in the
cells (15), the effect of S1P on [Ca2+]i was
examined in Fig. 1B. As expected, S1P increased
[Ca2+]i; the lipid transiently increased
[Ca2+]i followed by a sustained increase at
the level of about 50% of the peak value (Fig. 1B). In PTX-treated
cells, the peak value was slightly lower than that of control cells not
treated with the toxin (Fig. 1B). The potency of S1P for the
Ca2+ response (increment of peak level from basal level)
was comparable with that for the H2O2 response.
Furthermore, consistent with the H2O2 response,
PTX hardly affected the response to the lower dose of S1P, but
significantly inhibited the response to a higher dose of the lipid,
although the threshold of S1P for the inhibition by the toxin was
slightly lower in the Ca2+ response (
3 µM)
than the H2O2 response (30 µM).
This difference in the threshold of PTX effect may reflect the
difference in the time when the S1P-induced responses were measured,
i.e. at 10 min for the H2O2 response
and at about 30 sec for the Ca2+ response.
|
, we examined the S1P effect on inositol
polyphosphate production. S1P clearly increased it in a time (Fig. 2A)-
and dose (Fig. 2B)-dependent manner, although more than 3
µM S1P was necessary for a significant production. In
this case as well, the PTX effect depended on the S1P dose employed;
the response to the lower dose (3 µM) was insensitive,
but those to higher doses (10–30 µM) were sensitive to
the toxin.
S1P-induced
[Ca2+]i increase and
H2O2 generation are
dependent on PLC activity
Thus, S1P stimulated H2O2 generation
and increased [Ca2+]i and inositol
polyphosphate production in parallel in a dose-dependent manner. This
suggests that S1P-induced stimulation of PLC-catalyzed production of
IP3 may be responsible for mobilization of Ca2+
and subsequent H2O2 generation. To further
confirm this point, we performed the following two kinds of
experiments. Firstly, we examined the effects of PLC inhibitor on the
S1P-induced responses (Fig. 3). As expected, the
S1P-induced inositol polyphosphate production was markedly inhibited by
U73122, a PLC inhibitor (Fig. 3A). In parallel with the inositol
phosphate response, the S1P-induced increases in
[Ca2+]i (Fig. 3B) and
H2O2 generation (Fig. 3C) were markedly
inhibited by this enzyme inhibitor. U73343, an inactive derivative of
U73122 for the enzyme, was ineffective on all of these responses,
suggesting the specific effect of the inhibitor (Fig. 3).
Secondly, we examined the effect of an A1-adenosine
receptor agonist on the S1P-induced responses. We have previously shown
that A1-agonists, through PTX-sensitive
Gi/Go proteins, can enhance the PLC activation
and subsequent Ca2+ mobilization induced by
Ca2+-mobilizing receptor agonists, including
1-adrenergic agonists (21), TSH (22), and
P2-purinergic agonists (18, 23). Even though PIA alone
hardly affected inositol polyphosphate production, the adenosine analog
significantly enhanced the S1P-induced response (Fig. 4A). The enhancement of the PLC response was accompanied
by enhancement of the Ca2+ response (Fig. 4C) and the
H2O2 response (Fig. 4E). Thus, PIA, which alone
exerted only a small effect, synergistically or permissively increased
the S1P-induced responses. In accordance with the cases of
A1-agonist effects on other Ca2+-mobilizing
agonist-induced actions (21, 22, 23), PIA-induced responses in
collaboration with S1P were completely inhibited by prior treatment of
the cells with PTX (Fig. 4, B, D, and F), suggesting an involvement of
Gi/Go proteins in PIA signaling. These results,
shown in Figs. 3 and 4, clearly indicated an involvement of PLC in the
S1P-induced increase in [Ca2+]i and
H2O2 generation.
|
, S1P-induced activation of PLC was partially
inhibited by PTX treatment, suggesting that
Gi/Go proteins are somehow involved in the
lipid signaling. If Gi/Go proteins are involved
in the S1P signaling, one can expect that the lipid may also affect
other Gi/Go protein-mediated responses. Here,
we examined the S1P effect on cAMP accumulation. As shown in Fig. 5, S1P, in a dose-dependent manner, inhibited
TSH-induced cAMP accumulation. In this experiment, because
3-isobutyl-1-methylxanthine, a potent phosphodiesterase inhibitor, was
included in the incubation medium, a change in the cAMP content may,
therefore, reflect adenylyl cyclase activity. As expected, PTX
treatment of the cells almost completely suppressed the S1P-induced
inhibition of cAMP accumulation (Fig. 5).
|
). S1P was effective in the
presence of GTPS, a hydrolysis-resistant derivative of GTP, but was
ineffective in its absence; the lipid significantly enhanced
GTPS-induced activation of the enzyme (Fig. 6A). The enhancement by
S1P of the enzyme activity was dependent on the concentration of
GTPS employed; S1P potentiated the action of GTPS at lower doses
of the nucleotide, but not at the maximally effective dose of GTPS
(Fig. 6B). Thus, S1P increased the apparent affinity for GTPS to
activate the enzyme. The same guanine nucleotide dependency was
observed for other Ca2+-mobilizing agonist-induced enzyme
activations in FRTL-5 cells (19) and in other cell membrane
preparations (24). This phenomenon has been recognized in such a way
that a receptor ligand induces coupling of its receptor to G proteins
and thereby accelerates the exchange of GDP with GTP (or GTPS) on
the G protein molecules (24, 25). Thus, S1P seems to activate the
enzyme through a G protein-mediated mechanism. The involvement of G
proteins in the S1P-induced activation of the enzyme was further
supported by the observation that GDPßS almost completely inhibited
S1P-induced GTPS-dependent enzyme activation (Fig. 6A).
|
. S1P alone did not exert any effect at least up to 30
µM, but the lipid did activate the enzyme in the presence
of GTPS with a potency similar to that in intact cells (Fig. 2B).
The enzyme activity was also examined in the membranes prepared from
PTX-treated cells; the toxin treatment hardly affected the basal
activity or GTPS- and S1P-induced enzyme activation (Fig. 6C).
No cross-desensitization of S1P and bradykinin for induction of
Ca2+ mobilization
The foregoing results clearly suggested the involvement of some G
proteins in the S1P signaling pathways. However, it is still unclear
whether S1P activates these G proteins directly or indirectly through
receptors as do other G protein-coupled receptor agonists. To clarify
this point, we examined the cross-desensitization of bradykinin, whose
receptor is known to couple to G proteins, and S1P for induction of
Ca2+ mobilization (Fig. 7). To
eliminate the possible involvement of Gi/Go
protein-induced permissive stimulation of PLC (18, 21, 22, 23), the cells
were pretreated with PTX. The toxin treatment did not significantly
affected bradykinin-induced [Ca2+]i increase;
the bradykinin (1 µM)-induced
[Ca2+]i increase (nM) was
708 ± 55 and 699 ± 102 for control cells and PTX-treated
cells, respectively (n = 4 observation). When the cells were first
challenged with S1P, the second S1P challenge only slightly increased
[Ca2+]i. Likewise, the second bradykinin
challenge hardly increased [Ca2+]i. Thus,
S1P- and bradykinin-induced Ca2+ mobilization were
homologously desensitized. Under these conditions, there was no
cross-desensitization of S1P and bradykinin for induction of the
Ca2+ response; the response to either S1P or bradykinin
after a bradykinin challenge or a S1P challenge was unchanged compared
with the respective control response (Fig. 7).
|
) and was
almost completely by U73122 (data not shown), suggesting mediation
through a G protein-regulated PLC activation. To examine the possible
involvement of a LPA receptor in the S1P-induced Ca2+
response, cross-desensitization experiments were performed. In these
experiments as well, PTX-treated cells were used to eliminate the
involvement of Gi/Go protein-induced permissive
stimulation of the PLC-Ca2+ system. When the cells were
pretreated with S1P, the Ca2+ responses to LPA (Fig. 8, B
and C) as well as those to S1P (see Fig. 7) were markedly inhibited.
Similarly, the effects of LPA and S1P were inhibited after a LPA
challenge (Fig. 8, B and C). In the case of LPA, its effect was
markedly inhibited. However, in the case of S1P, only 20–50% of the
effect was inhibited depending on the S1P dose employed; if the lipid
dose was increased, the inhibition rate increased (Fig. 8, B and C).
Under these conditions, the bradykinin-induced action was unchanged
after a prior challenge with either S1P (Fig. 7) or LPA (data not
shown), indicating that the loss of the response was not due to the
depletion of Ca2+ in the pool.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
G Proteins have been assumed to be involved in the S1P-induced activation of PLC based on PTX sensitivity in the previous studies; the toxin treatment of cells markedly inhibited lipid-induced enzyme activation in 3T3 fibroblasts (3) and HL-60 cells (10). However, in the previous studies, there was no evidence showing that guanine nucleotides actually regulate the lipid-induced enzyme activation. In the present study, we demonstrated for the first time that the lipid-induced activation of the enzyme is guanine nucleotide dependent in the membrane preparations of FRTL-5 cells. S1P-induced guanine nucleotide-dependent activation of PLC in a cell-free system was totally independent of PTX at any dose of S1P employed, suggesting that lipid-induced enzyme activation may be mediated by PTX-insensitive G proteins.
Consistent with the results in a cell-free system, S1P-induced actions on PLC and its cascade reactions at lower doses (<3 µM) were hardly affected by PTX treatment in intact cells; however, those at higher doses (>10 µM) were partly, but significantly, inhibited by the toxin treatment. The reversal of S1P-induced inhibition of cAMP accumulation by PTX treatment suggests that the partial inhibition of PLC activation by the toxin may not be due to incomplete ADP ribosylation of Gi/Go proteins by the toxin. In fact, [32P]ADP ribosylation of Gi/Go proteins by PTX of the membrane prepared from the cells pretreated with the toxin in a similar way to the present experiments was almost completely lost, reflecting the consumption of PTX substrate proteins by prior exposure of the cells to the toxin (17). Complete ADP ribosylation of Gi/Go proteins was also supported by the finding that PIA-induced responses in collaboration with S1P were completely inhibited by the same treatment of the cells with the toxin. Furthermore, increasing the PTX dose from 50 ng/ml to 1 µg/ml never elevated the rate of the inhibition of S1P-induced Ca2+ response (data not shown). These results suggest the involvement of a PTX-insensitive mechanism as well as a toxin-sensitive one in S1P-induced PLC activation.
We cannot explain clearly the discrepancy in the results between intact
cells and cell-free systems with respect to PTX sensitivity; however,
we observed similar phenomena in A1-adenosine
receptor-induced activation of PLC. Adenosine or
A1-receptor agonists, although they have a very low or
undetectable effect on PLC activity, markedly enhanced the enzyme
activation induced by Ca2+-mobilizing agonists such as
1-adrenergic agonists (21), TSH (22), and
P2-purinergic agonists (18, 23) in intact FRTL-5 cells. The
A1-agonist-induced actions were completely suppressed by
PTX treatment (18, 21, 22, 23). Thus, A1-receptor-mediated
activation of PLC is detected only when Ca2+-mobilizing
agonists coexist. In a cell-free system, however, the PTX-sensitive
A1-receptor agonist-induced PLC activation was not
detected, whereas the PTX-insensitive Ca2+-mobilizing
agonist-induced enzyme activation was clearly detected (19). This
suggests that in our FRTL-5 thyroid cell-free system, the indirect or
permissive stimulation of the enzyme through
Gi/Go proteins cannot be detected. This also
suggests that S1P-induced activation of PLC at higher doses involves
such an indirect or permissive activation by
Gi/Go proteins of the enzyme. The activation of
Gi/Go proteins by S1P was suggested based on
the finding that S1P at higher doses of more than 10 µM
inhibited cAMP accumulation, probably reflecting the inhibition of
adenylyl cyclase, in a PTX-sensitive manner. If S1P, like PIA, has the
ability to activate Gi/Go proteins, one might
wonder why PIA could enhance S1P-induced PLC activation. The ability of
S1P to activate Gi/Go proteins is much less
than that of A1-agonists, as evidenced by only about 50%
inhibition by S1P vs. more than 95% inhibition by PIA (18, 22) of TSH-induced cAMP accumulation. This discrepancy may explain why
PIA can enhance S1P-induced activation of PLC and its cascade
reactions. Although further experiments are required to understand more
precisely the mechanism underlying S1P-induced activation of PLC,
especially regarding the species of G proteins involved and their roles
in enzyme activation, the results of the present study suggested that
at least two types of G proteins may be involved in the S1P signaling
in FRTL-5 cells.
In addition to PLC activation, S1P has recently been reported to influence effector systems, such as adenylyl cyclase (3, 26), K+ channel (26, 27), and mitogen-activated protein kinase (4), that are involved in the early signal transduction of cell surface receptor agonists. Similar stimulation of these effector systems has been reported by lysosphingolipids other than S1P, such as SPC and psychosine (16, 26, 28, 29, 30). As mentioned above, we found that S1P induced inhibition of adenylyl cyclase in addition to PLC activation in FRTL-5 cells. In this case, PTX almost completely suppressed the S1P-induced action. Thus, S1P stimulated at least two signaling pathways, leading to activation of PLC and inhibition of adenylyl cyclase, probably through distinct G proteins in FRTL-5 thyroid cells, i.e. the former pathway may be mediated predominantly through PTX-insensitive G proteins such as Gq/G11, and the latter pathway through PTX-sensitive G proteins such as Gi/Go proteins. Gi/Go proteins may also be involved in the activation of PLC in an indirect or permissive manner.
The involvement of G proteins in S1P signaling pathways raises the possibility of the existence of G protein-coupled receptors for the lipid. In fact, although no direct evidence has been presented, cell surface receptors for S1P have recently been proposed in Xenopus oocytes (30), HL-60 cells (10), and HEK293 cells (26). Bradykinin activates PLC through PTX-insensitive G proteins, probably Gq/G11 proteins (31). In FRTL-5 cells as well, the bradykinin-induced [Ca2+]i increase was insensitive to PTX, suggesting an involvement of Gq/G11 proteins in the PLC-Ca2+ system, although the possibility cannot be ruled out that S1P and bradykinin act through a different Gq/G11 family of proteins. Lack of cross-desensitization between S1P and bradykinin in the induction of Ca2+ mobilization in FRTL-5 cells suggests that S1P-induced desensitization seems to occur at the level before G protein in the lipid signaling. In our preliminary experiments, we found that S1P did not significantly activate the PLC-Ca2+ system in either PC12 cells or GH3 cells, in which Gq/G11 proteins are demonstrated to be involved in the receptor-mediated PLC activation (32, 33). This also supports the idea that Gq/G11 proteins may not be the site of S1P. These results imply the existence of the putative receptor as a primary action site of S1P. To conclude the existence of the putative S1P receptor, however, S1P binding experiments or receptor cloning studies are necessary.
In relation to the putative S1P receptor, it has also been proposed
that S1P acts through a recently identified receptor for LPA, one of
the lysoglycerolipids, which has a chemical structure similar to that
of S1P, in some cell types (34, 35). In FRTL-5 cells, LPA also
increased [Ca2+]i and induced
H2O2 generation (data not shown); however, in
this case, the toxin treatment markedly (70–80%; in the case of S1P,
this value was at most 50%) suppressed these LPA-induced responses.
Furthermore, under conditions where the LPA-induced
[Ca2+]i increase was almost completely
desensitized, S1P was still effective in increasing
[Ca2+]i depending on the concentration of S1P
employed, suggesting that S1P acts at least partly through a LPA
receptor-independent mechanism, although S1P and LPA may in part share
the same receptor, as evidenced from the partial cross-desensitization
of these lipids. The inhibition rate of the S1P-induced response after
a LPA challenge increased as the dose of S1P increased. This suggests
that S1P may interact preferentially with a S1P-specific receptor at
the lower dose, but if the dose of S1P is increased, the lipid may
interact with, in addition to the S1P specific receptor, a LPA receptor
or a similar receptor whose signaling pathway converges with that of
the LPA receptor.
At present, we cannot present evidence showing that extracellular S1P induces in vivo the thyroid functional responses observed in the present study. However, a recent study showed that S1P can be released from platelets into the extracellular medium in response to physiological agonists such as thrombin (36); about 5% of S1P in platelets (1.42 nmol/109 platelets) could be released in response to agonists. If this occurs in extracellular environment of thyroid cells in vivo, the concentration of S1P would reach about 20 nM; this value (nanomoles per liter) was estimated as the following equation, 0.05 x (1.42/109) x (3 x 1011), where the amount of platelets in blood was assumed to be 3 x 1011 platelets/liter (37). A minimum effective doses of S1P to induce a [Ca2+]i increase and H2O2 generation in the present study was about 30 nM, which is not very different from the concentration estimated under conditions where platelets are activated. S1P is a form of sphingosine phosphorylated by sphingosine kinase. Because sphingosine and the enzyme seem to be distributed in many types of cells and tissues (38), we can also imagine that S1P is produced in thyroid cells and released outside the cells as an autocrine factor. This hypothesis is our next project of investigation.
In conclusion, in FRTL-5 thyroid cells, exogenous S1P can induce H2O2 generation, a pivotal process for expression of thyroid functions, probably through IP3-dependent Ca2+ mobilization. S1P not only activates PLC, resulting in the production of IP3, but also inhibits adenylyl cyclase, resulting in a decrease in the cAMP content. The S1P-induced stimulation of the signaling pathways may be mediated through a receptor(s) coupling to G proteins. Thus, in addition to the second messenger role inside the cells, S1P seems to affect cellular functions, such as the lipid mediators PGs and leukotrienes, from outside the cells.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received May 21, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-adrenergic
receptor-mediated activation of phospholipase C and Ca2+
mobilization in a pertussis toxin-sensitive manner in FRTL-5 thyroid
cells. FEBS Lett 248:145–149[CrossRef][Medline]
q subfamily of guanine
nucleotide-binding regulatory protein subunits attenuate activation
of phosphatidylinositol 4,5-bisphosphate hydrolysis by hormones. J
Biol Chem 266:20519–20524q and G11 proteins couple the
thyrotropin-releasing hormone receptor to phospholipase C in
GH3 rat pituitary cells. J Biol Chem 267:24983–24988This article has been cited by other articles:
 |
|
 |
|
  N. Kanamoto, T. Akamizu, H. Hosoda, Y. Hataya, H. Ariyasu, K. Takaya, K. Hosoda, M. Saijo, K. Moriyama, A. Shimatsu, et al. Substantial Production of Ghrelin by a Human Medullary Thyroid Carcinoma Cell Line J. Clin. Endocrinol. Metab., October 1, 2001; 86(10): 4984 - 4990. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Xu, Y.-j. Xiao, L. M. Baudhuin, and B. M. Schwartz The Role and Clinical Applications of Bioactive Lysolipids in Ovarian Cancer Reproductive Sciences, January 1, 2001; 8(1): 1 - 13. [Abstract] [PDF]
|
|
|
|
 |
|
 Y. S. Bae, J.-Y. Sung, O.-S. Kim, Y. J. Kim, K. C. Hur, A. Kazlauskas, and S. G. Rhee Platelet-derived Growth Factor-induced H2O2 Production Requires the Activation of Phosphatidylinositol 3-Kinase J. Biol. Chem., March 31, 2000; 275(14): 10527 - 10531. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Kon, K. Sato, T. Watanabe, H. Tomura, A. Kuwabara, T. Kimura, K.-i. Tamama, T. Ishizuka, N. Murata, T. Kanda, et al. Comparison of Intrinsic Activities of the Putative Sphingosine 1-Phosphate Receptor Subtypes to Regulate Several Signaling Pathways in Their cDNA-transfected Chinese Hamster Ovary Cells J. Biol. Chem., August 20, 1999; 274(34): 23940 - 23947. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. An, T. Bleu, and Y. Zheng Transduction of Intracellular Calcium Signals through G Protein-Mediated Activation of Phospholipase C by Recombinant Sphingosine 1-Phosphate Receptors Mol. Pharmacol., May 1, 1999; 55(5): 787 - 794. [Abstract] [Full Text]
|
|
|
|
|
|
K. Tornquist, A.-M. Malm, M. Pasternack, R. Kronqvist, S. Bjorklund, R. Tuominen, and J. P. Slotte Tumor Necrosis Factor-alpha , Sphingomyelinase, and Ceramide Inhibit Store-operated Calcium Entry in Thyroid FRTL-5 Cells J. Biol. Chem., April 2, 1999; 274(14): 9370 - 9377. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Sato, H. Tomura, Y. Igarashi, M. Ui, and F. Okajima Possible Involvement of Cell Surface Receptors in Sphingosine 1-Phosphate-Induced Activation of Extracellular Signal-Regulated Kinase in C6 Glioma Cells Mol. Pharmacol., January 1, 1999; 55(1): 126 - 133. [Abstract] [Full Text]
|
|
|
|
 |
|
 J. R. Van Brocklyn, M.-J. Lee, R. Menzeleev, A. Olivera, L. Edsall, O. Cuvillier, D. M. Thomas, P. J.P. Coopman, S. Thangada, C. H. Liu, et al. Dual Actions of Sphingosine-1-Phosphate: Extracellular through the Gi-coupled Receptor Edg-1 and Intracellular to Regulate Proliferation and Survival J. Cell Biol., July 13, 1998; 142(1): 229 - 240. |
and •) or
PTX-treated cells (
and 
) were incubated for the indicated time
with (closed) or without (open) 10 µM S1P. The production
of IP2 plus IP3 was measured. Results are
expressed as percentages of the respective initial values. Normalized
initial values (see Materials and Methods) were 636
± 8 dpm for control cells and 669 ± 12 dpm for PTX-treated
cells, respectively. Data were the mean ± SE of three
separate experiments for control and PTX-treated cells. In B, control
cells (•) or PTX-treated cells (
