(Received for publication, February 4, 1997, and in revised form, April 30, 1997)
From the Department of Physiology, The Johns Hopkins
University School of Medicine, Baltimore, Maryland 21205 and the
Departments of § Cell and ¶ Pharmacological and
Physiological Sciences, St. Louis University School of Medicine, St.
Louis, Missouri 63104
Modulation of the components involved in
mitogenic signaling cascades is critical to the regulation of cell
growth. GTP-binding proteins and the stimulation of phosphatidylcholine
(PC) hydrolysis have been shown to play major roles in these cascades.
One of the enzymes involved in PC hydrolysis, a PC-specific
phospholipase C (PC-PLC) has received relatively little attention. In
this paper we examined the role of a particular heterotrimeric
GTP-binding protein, Go, in the regulation of cell
growth and PC-PLC-mediated hydrolysis of PC in IIC9 fibroblasts. The
Go -subunit was ablated in IIC9 cells by stable
expression of antisense RNA. These stably transfected cells acquired a
transformed phenotype as indicated by: (a) the formation of
multiple foci in monolayer cultures, (b) the acquisition of
anchorage-independent growth in soft agar; and (c) an
increased level of thymidine incorporation in the absence of added
mitogens. These data implicate Go
as a novel tumor
suppressor. Interestingly, PC-PLC activity was constitutively active in
the Go
-ablated cells as evidenced by the chronically
elevated levels of diacylglycerol and phosphorylcholine in the absence
of growth factors. In contrast, basal activities of PC-phospholipase D, phospholipase A2, or phosphoinositol-PLC were not affected.
These data demonstrate, for the first time, a role for Go
in regulating cell growth and provide definitive evidence for the
existence of a PC-PLC in eukaryotic cells. The data further indicate
that a subunit of Go, is involved in regulating this
enzyme.
Defects in signal transduction cascades involved in the regulation
of cell growth often lead to pathological conditions, including the
development of neoplasms. Heterotrimeric GTP-binding proteins (G
proteins)1 and induced lipid metabolism are
important components in growth factor-coupled cellular signal
transduction pathways. Heterotrimeric G proteins are a family of
membrane-bound proteins composed of ,
, and
subunits, which,
in response to receptor activation, dissociate into free
subunits
and
dimers. Both the GTP-bound
subunits and
dimers
have been shown to play roles in a variety mitogenic signal
transduction cascades (1), including those involving induced lipid
metabolism (2, 3).
There is now strong evidence indicating that G proteins play crucial
roles in the regulation of mitogenic signals and that specific defects
in these proteins lead to the development of transformed phenotypes. In
addition to the observation that the activation of certain growth
factor receptors stimulates the dissociation of G proteins into
GTP-bound subunits and
dimers, mutations that reduce the
intrinsic GTPase activity in specific
-subunits transform these G
proteins into oncoproteins. For example, mutations in the
Gs
gene result in an oncogene (gsp), the
protein product of which is a Gs
with substitutions at
amino acids 201 (R201C/H) and 227 (Q227R/H/L) which have been found in
growth hormone-secreting pituitary tumors (4). Similarly, mutations in
the Gi2
gene yield another oncogene (gip2)
characterized by a substitution of amino acid 179 in Gi2
(R179C/H) which has been found in ovarian sex cord stromal tumors and
adrenal cortical tumors (5). These data provide strong support for the
notion that these G proteins are important components involved in the
regulation of mitogenic signal transduction cascades and represent
potential targets for oncogenic mutations in human tumors.
In addition to G proteins, agonist-induced lipid metabolism also plays a central role in mitogenic signaling cascades. While induced hydrolysis of phosphoinositides (PIs) has long been recognized as playing such a role (6), it is now well recognized that induced PC metabolism is often just as, if not more, important (7). Although PI and PC hydrolysis are induced by a variety of mitogens, PI hydrolysis is often transient while PC hydrolysis is usually sustained in the continuous presence of growth factors. In this regard, PC hydrolysis correlates with the requirement for the prolonged presence of growth factors for full mitogenic responses (8).
Depending on the cell type and specific mitogen, three enzymes have been implicated in mitogen-induced PC metabolism: PLA2, PC-PLD, and PC-PLC. PLA2 removes the fatty acid esterified at sn-2 of the glycerol backbone in PC resulting in the liberation of a free fatty acid, often arachidonic acid, and a lysophospholipid. PLD-mediated hydrolysis of PC results in the production of phosphatidic acid (PA) and free choline. This generated PA is often, but not always, hydrolyzed by phosphatidic acid phosphohydrolase (PAPH) leading to the production of diacylglycerol (DAG). Alternatively, in some systems (9-15), PC-derived DAGs, in addition to phosphorylcholine, are produced from a PC-PLC-mediated hydrolysis of PC. Most studies have focused on PLA2 and PLD while very little attention has been given to PC-PLC. Indeed, the existence of a eukaryotic PC-PLC remains somewhat controversial.
Given this lack of attention to eukaryotic PC-PLC, it is not surprising
that the cellular components involved in its regulation have not been
identified. The observation that activation of receptors, such as the
thrombin receptor (16), which are known to couple to G proteins (17)
lead to an increase in PC-PLC activity (15, 18) suggest that this
enzyme is regulated by a G protein. That a G protein couples to PC-PLC
is consistent with the fact that these proteins are known to regulate
other specific phospholipases. PI-PLC is regulated by a pertussis
toxin-sensitive G protein (19), involving both
q and
dimers (20-22). In a similar manner, heterotrimeric G proteins
have been implicated in the regulation of a high molecular weight
PLA2 (23), while PC-PLD has been shown to be regulated by
small molecular weight GTP-binding proteins (24). In view of these
data, it is reasonable to hypothesize that PC-PLC may also be regulated
by a G protein.
In this report, we demonstrate that the ablation of Go
results in a transformed phenotype. Furthermore, in these
Go
-ablated cells, PC-PLC is significantly elevated
providing definitive evidence for a PC-PLC and implicating
Go, Go
in particular, in the regulation of
this enzyme in vivo. The relationship between
Go
, the transformed phenotype and the constitutive
activation of PC-PLC is discussed.
Tissue culture media components, Lipofectin
reagents, Geneticin (G418), and calf alkaline phosphatase (1000 units/µg) were purchased from Boehringer Mannheim. Plastic culture
dishes were purchased from Falcon Labware. Highly purified human
thrombin (4000 NIH units/ml) and bovine serum albumin
(radioimmunoassay grade, fraction V) were purchased from Sigma.
Escherichia coli diacylglycerol kinase was obtained from
Lipidex or CalBiochem. AG1X8 Resin (200-400 mesh, formate form) was
from Bio-Rad. TLC plates were purchased from EM Diagnostics, Analabs,
and Analtech. CytoScint scintillation counting fluid was obtained from
ICN. Radioactive materials were purchased from Amersham. Molecular biology enzymes were purchased from Stratagene, Life Technologies, Inc., New England Biolabs, and Boehringer Mannheim.
[methyl-3H]Choline (83 Ci/mmol) and
[9,10-3H]myristic acid (53 Ci/mol) were purchased from
Amersham. [
-32P]ATP (3000 Ci/mmol),
[methyl-3H]thymidine (6.7 Ci/mmol), and
[5,6,8,9,11,12,14,15-3H]arachidonic acid (180-240
Ci/mmol) were purchased from NEN Life Sciences Products. Rat
Go
cDNA plasmid (pGEM-2/Go
) was
generously provided by Dr. Randy Reed (Howard Hughes Medical Institute,
Johns Hopkins Medical Institutes, Baltimore MD).
IIC9 cells, a subclone of Chinese
hamster embryo fibroblasts (25), were grown, maintained, and
serum-deprived as described previously (8). Briefly, cultures were
grown and maintained in minimal essential medium-/Ham's F-12 medium
(1:1, v/v) containing 5% (v/v) fetal calf serum, 100 units of
penicillin/ml of 100 mg of streptomycin/ml, and 2 mM
L-glutamine (complete media). Subconfluent cultures were
serum-deprived by washing three times with Dulbecco's modified
Eagle's medium containing 1 mg/ml bovine serum albumin (radioimmunoassay grade), 100 units of penicillin, 100 mg of
streptomycin/ml, 2 mM L-glutamine, and 20 mM NaHepes, pH 7.4. The cultures were then incubated in
this media supplemented with 5 mg/ml human transferrin (serum-free
medium) and incubated for 2 days at 37 °C. Cultures were washed
twice and equilibrated in fresh serum-free media for at least 30 min
prior to addition of each experiment.
EcoRI fragment of Go cDNA
from pGEM-2/Go
was subcloned into a vector plasmid,
pcDNAI, in a antisense orientation, i.e. the 3
end of
the Go
cDNA was immediately adjacent to
cytomegalovirus promoter under which control the antisense sequence is
transcribed. Plasmids were sequenced and transfected into IIC9 cells
using a Lipofectamine protocol (Life Technologies, Inc.). pcDNAI
without inserts were transfected into IIC9s as a control. Briefly,
subconfluent cells in 60-mm culture dishes were washed with Opti-MEM
(Life Technologies, Inc.) and transfected by incubation with 5-µg
plasmids and Lipofectamine for 24 h at 37 °C. The Opti-MEM
media was then replaced with complete media and the cells were grown
for 48 h at 37 °C to allow expression of the neomycin
resistance gene products. The transfected cells were subcultured and
grown for several weeks in selection medium (complete medium
supplemented with 500 µg/ml G418). G418-resistant clones were
isolated with cloning cylinders and the transfected clones were
maintained in complete medium supplemented with 250 µg/ml G418.
All other assays, including growth in soft agar, [3H]thymidine incorporation, Western blot analysis, and quantification of DAG mass, choline metabolites, and PLD activation were performed as described previously (8, 15, 26-29) as indicated in the figure legends.
To investigate the physiological role of
Go, we stably transfected IIC9 cells with a
Go
antisense construct (Fig.
1A). Western blot analysis demonstrated that
Go
was absent in the transfected cells while other G
protein
subunits, Gi1
, Gi2
,
Gs
, and Gq
, were present (Fig.
1B). This has been observed in at least three independently
isolated Go
-ablated clones (data not shown).
In contrast to wild type IIC9 cells which are flat and extended (Fig.
2A), Go-ablated cells appear
round, retracted, and form multiple foci in confluent monolayer
cultures (Fig. 2B). This morphology, observed in three
independently isolated clones, suggest that the
Go
-ablated cells have lost contact inhibition and
acquired a transformed phenotype.
An important characteristic of transformed fibroblasts is their ability
to grow in an anchorage independent manner. In view of this and the
above data, we assessed the ability of the Go-ablated cells to grow in soft agar. As shown in Table I,
Goa1 cells formed 20-30-fold more colonies in soft agar
than wild type cells and this has been observed in a second,
independently isolated, clone (data not shown). Furthermore, each of
the colonies formed by the ablated cells were much larger and more
dense (Fig. 3, B and D) than the
colonies formed by the wild type cells (Fig. 3, A and
C). Cells transfected with control vectors (vectors without inserts) formed colonies similar to those seen with wild type cells
(data not shown).
|
To further investigate the possibility that the ablated cells were
transformed, we assessed the "basal" level of thymidine incorporation. As shown in Fig. 4, wild type IIC9 cells
became quiescent after serum deprivation for 48 h and the level of
[3H]thymidine incorporation was low. In contrast,
serum-deprived Go-ablated cells displayed a 10-fold
higher level of [3H]thymidine incorporation than the
quiescent wild type cells (Fig. 4). FCS (10%) or thrombin (2 NIH
units/ml) stimulated only a modest increase in
[3H]thymidine incorporation in the ablated cells while
these treatments stimulated a 10-fold increase in
[3H]thymidine incorporation in the quiescent wild type
cells (Fig. 4). These data have been observed in three independently
isolated clones. Consistent with these data, the
Go
-ablated cells survive in serum-free media for an
extended period of time while the wild type cells do not (Fig. 2,
C and D). These data indicate that the
Go
-ablated cells are not growth arrested in serum-free
medium and are consistent with the transformed phenotype of these
cells.
DAG Level Is Chronically Elevated in Go
Cells transformed as a consequence of a defect in a signal transduction component normally associated with the regulation of mitogenesis often show changes in the concentrations of second messengers under their control (30-32). Many of the signaling cascades known to be involved in mediating mitogenic signals involve the stimulation of lipid metabolism and G proteins are known to play a role in some of these cascades (1, 6, 7). In addition, the elevation of DAG levels plays a central early role in transducing the mitogenic signal in these cascades.
In view of this and the above observations regarding the transformed
phenotype of the Go-ablated cells, we measured the mass of DAG in the ablated cells. Subconfluent wild type and ablated cells
were incubated in serum-free medium for 2 days and DAG levels were
quantified. Interestingly, the basal DAG level in the serum-starved Go
-ablated cells was twice that of quiescent wild type
cells (Fig. 5A). Furthermore, while the
addition of
-thrombin to the wild type cells resulted in a 2-fold
increase in DAG mass level, addition of
-thrombin to the ablated
cells did not induce a significant further increase in DAG levels (Fig.
5A). These results, observed in two independently isolated
clones, indicate that the DAG level in Go
-ablated cells
was constitutively elevated even in the absence of any added
mitogens.
The Increased DAG Is Due to a Constitutively Activated PC-PLC
We have shown that PC hydrolysis is the major, if not
exclusive, source of mitogen-induced DAGs in IIC9 cells (8, 15, 26,
33-35). In view of these data, we examined the possibility that an
increase in PC hydrolysis contributed to the elevated DAG level in the
Go-ablated cells.
To determine if PC hydrolysis was affected in the
Go-ablated cells, the cells were radiolabeled to
isotopic equilibrium with [3H]choline chloride in
serum-free medium for 48 h and the intracellular [3H]choline and [3H]phosphorylcholine level
were quantified (15). TLC analysis of water-soluble head groups
indicated that the phosphorylcholine level in the
Go
-ablated cells was 5-10-fold higher than that found
in wild type cells. The level of choline in the ablated and wild type
cells, however, was identical (Fig. 5B). To ensure that the
increased level of radiolabeled phosphorylcholine was not due to
contaminant which co-migrated with the phosphorylcholine, the
radioactivity in the region of the TLC plate containing
phosphorylcholine was recovered, subjected to alkaline phosphatase
hydrolysis, and the products were identified by TLC. All of the
radioactivity that migrated with phosphorylcholine was converted to
choline indicating that co-migrating contaminants were not present
(data not shown). These data indicate that both phosphorylcholine and DAG, the two products of PC-PLC, are elevated in ablated cells and
strongly suggest that PC-PLC is constitutively activated in Go
-ablated cells. These results have been observed in
three independently isolated clones.
An alternative explanation for the above results is that PC is hydrolyzed via a PLD and the resulting PA is dephosphorylated to DAG, via PAPH, while the free choline is phosphorylated via CK. As a result of the combined action of all three enzymes, PLD, PAPH, and choline kinase (CK), an apparent PC-PLC activity would be detected similar to that observed in v-ras transformed cells (36-38).
In view of the above, PLD activity was quantified in
Go-ablated and wild type IIC9 cells by taking advantage
of the unique transphosphatidylation activity of PLD and the ability to
preferentially label PC by acute labeling with
[3H]myristate (15). In the transphosphatidylation
reaction, a small molecular weight alcohol such as ethanol is used as
the nucleophile in lieu of water resulting in the generation of
phosphatidylethanol instead of PA. As shown in Fig. 5C, PLD
activity is indistinguishable in the ablated and wild type cells in the
absence of thrombin or FCS. Furthermore, the addition of thrombin to
both cell types results in comparable increases in PLD activity. These
data indicate that both basal and thrombin activated PLD activity are
unaltered in Go
-ablated cells and have been observed in
three independently isolated clones.
To further examine the possible involvement of PLD/PAPH/CK activities,
CK activity was quantified in wild type and Go-ablated serum-deprived cells. These cells were incubated with
[3H]choline for 15 and 30 min and the level of
radiolabeled phosphorylcholine was quantified. As shown in Fig.
5D, the conversion of choline to phosphorylcholine was
essentially identical in both cell types demonstrating that the CK
activity was not elevated in the Go
-ablated cells. These
results have been observed in two independently isolated clones.
We should note that sphingomyelinase was also not contributing to the
increased phosphorylcholine levels in Go-ablated cells. If this choline metabolite was generated from a
sphingomyelinase-mediated hydrolysis of sphingomyelin ceramide, in
addition to phosphorylcholine, would be generated. Ceramide levels were
quantified, therefore, in wild type and Go
-ablated cells
and was found to be identical in both cell types (data not shown).
Taken together, the above data eliminate the involvement of PLD/PAPH/CK
as a mechanism for the chronic elevation of DAG and phosphorylcholine
levels in Go-ablated cells. In addition, they indicate
that Go
ablation-induced transformation is different from v-ras-induced transformation, since the later involves
PLD/PAPH/CK activities (36-38).
Another mitogen-activated PC hydrolyzing enzyme is
PLA2. In IIC9 cells, thrombin and FCS stimulate
PLA2 activity which hydrolyze PC to lysophosphocholine and
arachidonic acid (27), both of which have been implicated in mitogenic
signaling cascades (7). Basal and -thrombin-induced PLA2
activities in Go
-ablated and wild type cells was
assessed by quantifying the release of arachidonic acid and its
metabolites. As observed for PLD, basal and
-thrombin-activated PLA2 activity was not affected by the ablation of
Go
(Fig. 6A). Consistent with
these data, glycerolphosphocholine, another metabolite produced by the
hydrolysis of PLA2-generated lysophospholipid, was also at
similar levels in both cell types (data not shown). These data,
observed in three independently isolated clones, indicated that
PLA2 activity is unaffected in the
Go
-ablated cells.
PI Hydrolysis Is Suppressed in Go
Induced PI hydrolysis has been observed in response to a
wide variety of mitogens and defects in this metabolism have been implicated in cellular transformation. -Thrombin induces the hydrolysis of PIs and contributes to some of the thrombin-induced increase in DAG in IIC9s at early times (39). We therefore examined PI
metabolism in Go
-ablated cells to determine if this
hydrolysis contributed to the increased level of DAGs in the ablated
cells. PI metabolism was quantified in cells radiolabeled with
myo-[3H]inositol as described previously (27).
As shown in Fig. 6B, basal PI metabolism remains unaffected
in the ablated cells. As in wild type cells, thrombin induced PI
hydrolysis in the Go
-ablated cells. Interestingly,
however, the level of PI metabolism, assessed as IP3
production, is attenuated in the ablated cells (11-fold increase in
wild type cells and 2.5-fold in Go
-ablated cells). Similar results were obtained for the production of IP1 and
IP2 and have been observed in two independently isolated
clones (data not shown).
The above data demonstrate that
PC-PLC is constitutively active in Go-ablated cells. To
determine if this activation results from an increase in
dimers,
Goa1 cells were transfected with the
subunit of transducin
(Gt
) (see Ref. 66) which has been used to sequester
these dimers (40, 41). As shown in Fig. 7, transfecting
Goa1 cells with Gt
reduces the PC-PLC activity, indicated by the phosphorylcholine level, to that observed in wild type
IIC9s. These data suggest that an increase in
dimers is involved
in the increased PC-PLC activity observed in Go
-ablated cells and support the notion that these dimers are involved in the
receptor-mediated regulation of this enzyme.
The data in this paper demonstrate for the first time that the subunit of the Go GTP-binding protein is involved in
regulating cell growth and PC-PLC activity. Absence of this protein in
IIC9 cells results in: (a) a transformed phenotype, and
(b) a 10-fold increase in basal PC-PLC activity without
increasing PLD, PLA2, CK, sphingomyelinase, and PI-PLC
activities. These results indicate that Go selectively
regulates PC-PLC and is an important component involved in the
regulation of cell growth.
While some G proteins (e.g. Ras, Gi, and
Gs) are known to be involved in regulating cell growth (1,
42, 43), the data in this work are the first to demonstrate that
Go plays a role in mitogenesis. Ablation of
Go results in a transformed phenotype. This is indicated
by anchorage-independent growth (Table I and Fig. 3), formation of foci
in confluent cultures (Fig. 2), and increased level of thymidine
incorporation when cells are serum-deprived (Fig. 4). We should note
that the Go
is selectively ablated in this cells. Other
homologous G proteins, including Gi
whose coding sequence shows the greatest homology to Go
, are not
affected (Fig. 1). This is likely due to the fact that the antisense
construct included the 5
- and 3
-untranslated region of Go
which does not posses significant homology to other G proteins,
including Gi
(Basic Local Alignment Search Tool (BLAST)
data base, NCBI, Dec. 1996).
The above data suggest that the presence of Go in
quiescent cells is involved in suppressing cell growth and supports the notion that this protein acts as a tumor suppressor (44). Consistent with this hypothesis, Go
was selectively deficient in
two pituitary tumors (7315a, MtTW15) (45). In this regard, it is
interesting to note that the human Go
gene maps to the
chromosomal region (16q13) (Online Mendelian Inheritance in Man,
OMIMTM), associated with the loss of heterozygosity identified in a
variety of tumors (46-51). Furthermore, because dominant negative
variants of tumor suppressor genes often result in the overexpression
of their proteins, our data suggest that the increased levels of
Go
protein observed in many neuroendocrine tumors (52,
53) and in Merkel cell carcinoma (54) may result from a dominant
negative defect in this G protein, consistent with its role in
suppressing cell proliferation. Taken together, these data demonstrate
that Go
is an important component of a mechanism
involved in the regulation of cell growth. The suppression of
mitogenesis by Go
is a novel observation in that most
defects in G proteins resulting in cell transformation involve
expression of a constitutively active form (GTPase deficient) of G
proteins (4, 5).
In addition to a transformed phenotype, a phospholipase associated with
the regulation of cell growth (11, 15, 55-58), PC-PLC, is
constitutively active in the Go-ablated cells. The basal
levels of other phospholipases involved in signal transduction cascades, PLA2 (Fig. 6A), PLD (Fig.
5C), and PI-PLC (Fig. 6B) are not affected in the
ablated cells. Thrombin-induced levels of PLA2 and PLD are
also unaffected. Interestingly, induced levels of PI-PLC appears to be
blunted (Fig. 6B), which may be due to a protein kinase
C-mediated inhibition of this enzyme resulting from the elevated levels
of DAG as a result of constitutive PC-PLC activation. Importantly,
however, the suppression of PI hydrolysis cannot account for the
increase in DAG levels observed in the Go
-ablated
cells.
The DAG elevation is significant in that the mass amount of this lipid
is strictly regulated in wild type cells reflecting its importance in
initiating cell proliferation. It is noteworthy that the amount of DAG
present in serum-deprived Go-ablated cells is similar to
that observed in wild type cells stimulated with a maximal
concentration of mitogen (2 NIH units/ml thrombin or 10% FCS). It is
conceivable, therefore, that the elevated level of DAG in the ablated
cells is at least partly involved in generating the transformed
phenotype. Consistent with these data, comparable DAG levels were
observed in NIH 3T3 cells stably transfected with Bacillus
cereus PC-PLC which, interestingly, also displayed a transformed
phenotype (29).
In the accompanying article (66), we present data demonstrating that
Ras and ERK are constitutively active, as well as elevated expression
of cyclin D1 and constitutively active cyclin D1-CDK complexes, in the
Go-ablated cells. In this regard, it is interesting to
note that PC-PLC has been suggested to play a role in this pathway.
Transfection of 3T3 with bacterial PC-PLC reversed the inhibition of
cell growth mediated by a dominant negative Ras, but not dominant
negative Raf, constructs (55, 59). Consistent with this notion,
transfection of cells with bacterial PC-PLC results in a transformed
phenotype (60). Although Ras also activates choline kinase that
confuses the identification of a PC-PLC (30, 61), there is direct
evidence for a Ras-mediated activation of PC-PLC (57). Homogenates of
Ras-transformed cells contain a higher PC-PLC activity than their
non-transformed counterparts, that appears to be accompanied by an
increase in membrane-associated PC-PLC protein (57). Together, these
data support the notion that the increased PC-PLC activity observed in
the Go
-ablated cells reflects the constitutive
activation of the Ras/Raf/ERK signaling pathway.
The increase in PC-PLC is particularly interesting in that there has
been some controversy regarding the existence and regulation of this
enzyme (7). As mentioned, it is indeed possible that the "apparent"
PC-PLC activity observed in some studies was due to the combined
increase in PC-PLD, PAPH, and/or CK (30, 36, 37). The data in this
work, however, provide definitive evidence that a PC-PLC is present and
constitutively active in the Go-ablated cells.
Phosphorylcholine and DAG, two products for PC-PLC mediated hydrolysis,
are elevated in the ablated cells (Fig. 5, A and
B). The increase in phosphorylcholine is not due to an
increase in PLD (Fig. 5C) and/or CK (Fig. 5D)
activities or sphingomyelin hydrolysis (data not shown). DAG levels are
not due to any other phospholipid hydrolysis as products (head groups)
of other phospholipases are not elevated in the ablated cells (Fig.
6B and data not shown).
Three simplified working models for the mechanism by which
Go modulates PC-PLC. In one scheme (referred
to as Scheme I), the activation model, PC-PLC is activated by
dimers dissociated from Go in response to agonist
stimulation. A second model (referred to as Scheme II),
the Go-bound inhibitor model, PC-PLC is bound to an
inhibitor which is in a complex with the Go heterotrimer in
quiescent cells. Receptor activation leads to a dissociation of this
complex which relieves the inhibition of PC-PLC. In both Schemes I and
II, the interaction of the G protein with the enzyme may be direct or
may involve an auxiliary protein. This is particularly true for Scheme
I as Ras has been implicated in mediating the
dimer-induced
activation of PLD. The third model (referred to as Scheme
III), the dual G protein model, PC-PLC is regulated by
two G proteins, Go, and another as yet unidentified G
protein, G?. The GTP-bound form of Go
suppresses PC-PLC activity while this form of G?
stimulates activity. In the basal state, a certain proportion of the
G
subunits of each G protein is present in the free GTP-bound state
(62), but the predominant influence is suppression of PC-PLC by
Go
. Receptor stimulation results in an increase in
GTP-G?
binding to PC-PLC to stimulate its activity.
These models can be used to explain the increase in PC-PLC activity
observed in the Go-ablated cells. In Scheme I, ablation of Go
results in an increased level of
dimers
resulting in an increase in PC-PLC activity. In Scheme II, ablation of
Go
results in an inability of the inhibitor to complex
and inhibit PC-PLC. In Scheme III, the ablation of Go
removes the inhibitory G protein, resulting in the presence of the
stimulatory G protein, G?
, only. We should note that
Schemes II and III are consistent with the observation that
Gi
constitutively suppresses PI-PLC. Watkins et
al. (63) demonstrated that ablation of Gi2
resulted in enhanced basal PI-PLC in a
dimer-independent manner. The data
in Fig. 7, however, lend strong support to Scheme I. This is supported
in the accompanying paper (66) demonstrating that the specific
growth-associated activities that are constitutively elevated in Goa1
(Ras/ERK and cyclin D1) are reduced to basal levels in the presence of
Gt
. We cannot, however, completely rule out the
possibility that Gt
functionally complements
Go
in Schemes II or III. Experiments to discriminate
among these three models, and to completely define the role of this G
protein in regulating PC-PLC and cell growth, are in progress.
We thank Drs. Jeremy Nathans and Yanshu Wang for critically reading this manuscript and helpful suggestions.