(Received for publication, June 5, 1996, and in revised form, October 15, 1996)
From the Departments of Biochemistry and
§ Anesthesiology and Critical Care Medicine, Gifu
University School of Medicine, Tsukasamachi-40, Gifu 500, Japan
In response to dibutyryl cyclic AMP (dbcAMP) and
all-trans retinoic acid, human promyelocytic leukemic HL60
cells differentiate into granulocyte-like cells. In cell lysate and
in vitro reconstitution system, phospholipase D (PLD)
activity in response to guanosine 5-O-(3-thiotriphosphate)
(GTP
S) was up-regulated by dbcAMP or all-trans retinoic
acid treatment. In the present study, the mechanism(s) for increased
PLD activity during differentiation was examined. Western blot analysis
revealed that the contents of ADP-ribosylation factor, Rac2, and
Cdc42Hs but not RhoA and Rac1 in the cytosolic fraction were elevated
during differentiation. However, the cytosolic fraction from
undifferentiated cells was almost equally potent as the cytosolic
fraction from differentiated cells in the ability to stimulate membrane
PLD activity. It was shown that the GTP
S-dependent PLD
activity in membranes from differentiated cells was much higher than
that in membranes from undifferentiated cells, suggesting that the
increased PLD activity during differentiation was due to alterations in
some membrane component(s). Clostridium botulinum ADP-ribosyltransferase C3 and C. difficile toxin B, which
are known as inhibitors of RhoA and Rho family proteins, respectively, effectively suppressed PLD activity in membranes from differentiated cells. In fact, the amount of membrane-associated RhoA was increased during differentiation. Furthermore, the extent of
GTP
S-dependent PLD activity partially purified from
membranes from differentiated cells was greater than that from
membranes from undifferentiated cells in the presence of recombinant
ADP-ribosylation factor 1. The PLD (hPLD1) mRNA level was observed
to be up-regulated during differentiation, as inferred by reverse
transcription-polymerase chain reaction. Our results suggest the
possibility that the increased Rho proteins in membranes and the
changed level of PLD itself may be, at least in part, responsible for
the increase in GTP
S-dependent PLD activity during
granulocytic differentiation of HL60 cells.
Increasing evidence has indicated that phospholipase D
(PLD)1 plays an important role in signal
transduction in many types of cells (1). PLD hydrolyzes membrane
phospholipids, especially phosphatidylcholine (PC), and produces
phosphatidic acid, which can be further metabolized to diacylglycerol
by phosphatidic acid phosphohydrolase (2). PLD is activated by many
extracellular signal molecules, and several factors are involved in its
activation. In reconstitution experiments, activation of
membrane-associated PLD by nonhydrolyzable guanine nucleotide,
guanosine 5-O-(3-thiotriphosphate) (GTP
S), was observed
only when cytosol was present in the reaction mixture (3). Similar
findings were obtained in permeabilized cells in which the loss of
cytosolic components resulted in the reduction of
GTP
S-dependent PLD activity (4). These results imply
that cytosolic factors for PLD activation are presumed to be
GTP-binding proteins. Indeed, two small GTP-binding proteins have been
identified as regulatory factors for PLD activity (1). ADP-ribosylation
factor (Arf) from the brain cytosol acts as a cytosolic factor to
activate PLD in human promyelocytic leukemia HL60 membranes (5, 6). In
addition to Arf, evidence for the involvement of Rho family proteins in
PLD activation comes from the demonstration that Rho-specific GDP
dissociation inhibitor inhibits activation of PLD by GTP
S in
neutrophil membranes (7), rat liver plasma membranes (8), and HL60
membranes (9). Our recent study (10, 11) has demonstrated that RhoA and
protein kinase C exerted a synergistic stimulation of
membrane-associated PLD activity in HL60 cells.
HL60 cells can be differentiated into a mature granulocyte-like
phenotype by compounds such as dibutyryl cyclic AMP (dbcAMP) (12),
dimethyl sulfoxide (13), and retinoic acid (14). Differentiated HL60
cells possess many of the functional characteristics of granulocytes (15). Several studies have indicated the possibility that
receptor-mediated PLD activation may play essential roles in the
secretory response and O 2 or
H2O2 generation in neutrophils and
differentiated HL60 cells (16-25). Differentiated (but not
undifferentiated) HL60 cells exhibit formylmethionylleucylphenylalamine
(fMLP)- and ATP-induced PLD activation (17). Moreover, it was shown
that PLD activities stimulated by 4
-phorbol 12-myristate 13-acetate
in intact cells and by GTP
S in electropermeabilized cells increased
during differentiation induced by dbcAMP (17).
To gain further insight into the mechanism underlying the enhancement
of PLD activity during differentiation, we have studied GTPS-dependent PLD activation in the lysate and the
reconstitution system (the membrane and cytosolic fractions) prepared
from undifferentiated and differentiated HL60 cells.
GTP
S-dependent membrane PLD activity is higher in
differentiated cells than in undifferentiated cells, although
comparable amounts of Arf and Rho family proteins were present in the
cytosolic fractions from both undifferentiated and differentiated HL60
cells. The results obtained in the present study suggest the
possibility that the translocation of RhoA from the cytosol to membrane
and the changed level of PLD protein may be, at least in part,
responsible for the increase in GTP
S-dependent PLD
activity during differentiation.
RPMI 1640 medium, penicillin, streptomycin, and
recombinant reverse transcriptase were obtained from Life Technologies,
Inc. [3H]Oleic acid and
[palmitoyl-3H]dipalmitoyl phosphatidylcholine
(DPPC) were from DuPont NEN. dbcAMP, all-trans retinoic acid
(ATRA), nitroblue tetrazolium (NBT), NAD, UDP-glucose, brefeldin A,
phosphatidylinositol 4,5-bisphosphate, egg PC, and
phosphatidylethanolamine were from Sigma. GTPS was from Boehringer Mannheim. Silica Gel 60 (LK6D) plates were from Whatman. Heparin-Sepharose was from Pharmacia Biotech Inc. Protein concentrator equipment (Centricon 10) was from Amicon. Protein determination reagents were from Bio-Rad. Antibodies against Rho family
small GTP-binding proteins (RhoA, Rac1, Rac2, and Cdc42Hs) were from
Santa Cruz Biotechnology. ECL was from Amersham. Taq DNA
polymerase and random hexamer primers were from Takara Shuzo. Escherichia coli bearing Arf1 plasmid (26) and monoclonal
antibody against Arf were generous gifts from Dr. Joel Moss (National
Institutes of Health). E. coli bearing Clostridium
botulinum ADP-ribosyltransferase C3 (C3 toxin) plasmid (27) was
generously provided by Dr. Alan Hall (University College London,
London, United Kingdom). C. difficile toxin B (Toxin B) (28)
was kindly provided by Dr. David M. Lyerly (TechLab, Inc., VPI
Corporate Research Center, VA). Other reagents were of the highest
quality available.
The human promyelocytic leukemic HL60 cell line was kindly supplied by Dr. T. Okazaki (Osaka Dental University, Japan). HL60 cells were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 µg/ml streptomycin in a humidified atmosphere containing 5% CO2 at 37 °C. For differentiation, cells were cultured in serum-free RPMI 1640 medium supplemented with 5 µg/ml insulin and 5 µg/ml transferrin for 24 h. Differentiation was initiated by the addition of 0.5 mM dbcAMP or 1 µM ATRA. For the differentiation marker, NBT reduction activity was measured (12). For assay of PLD activity utilizing the endogenous substrate, cells were labeled with [3H]oleic acid (0.5 µCi/ml) for 12-15 h. Under this condition, 60-65% of the total radioactivity was incorporated into the PC fraction.
Preparation of Lysates, the Membrane, and Cytosolic Fractions from HL60 CellsHL60 cells were washed twice with buffer A (25 mM HEPES, pH 7.4, 100 mM KCl, 3 mM NaCl, 5 mM MgCl2, 1 mM EGTA, 5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, and 10 µg/ml leupeptin) and resuspended in buffer A. Cells were then disrupted by N2 cavitation (600 p.s.i. at 4 °C for 30 min). After unbroken cells and nuclei were removed by centrifugation at 900 × g for 5 min, the resulting supernatant was used as the HL60 cell lysate for experiments. The membrane and cytosolic fractions were further separated by centrifugation at 100,000 × g for 60 min. Membranes were washed once and resuspended in buffer A. Cytosol proteins were concentrated using Centricon 10.
Partial Purification of Membrane-associated PLD ActivityThe PLD fraction was prepared from membranes by the
method described by Brown et al. (5). Isolated membranes
were resuspended in buffer B (20 mM Na-HEPES, pH 7.5, 1 mM EDTA, 1 mM dithiothreitol, and 10 µg/ml
leupeptin) containing 1% cholate. After incubation at 4 °C for 60 min, the suspension was centrifuged at 100,000 × g for
60 min to obtain the membrane extract. Aliquots were loaded onto a
heparin-Sepharose column equilibrated with buffer C (buffer B
containing 0.7% n-octyl -glucopyranoside). Proteins
adsorbed onto the column including PLD activity were eluted with 1 M NaCl in buffer C and concentrated.
PLD activity was measured by
production of [3H]phosphatidylbutanol (PBut), a specific
product via its transphosphatidylation activity.
[3H]Oleic acid-labeled HL60 cell lysates (200 µg of
protein/assay) or membranes (20 µg of protein/assay) were incubated
in buffer A containing 1 mM MgATP and CaCl2 to
give a final free Ca2+ concentration of 1 µM
(total, 0.l ml) and stimulated with 10 µM GTPS at
37 °C for 30 min in the presence of butanol (0.3%, v/v).
Alternatively, PLD activity was measured in unlabeled membranes and
cytosol and extracted PLD fractions using substrate phospholipid vesicles prepared according to Brown et al. (5) with minor modifications. Mixed-lipid vesicles
(phosphatidylethanolamine/phosphatidylinositol 4,5-bisphosphate/egg PC,
10:1.5:1 molar ratio) containing
[palmitoyl-3H]DPPC (3 µCi/ml) were added to
membranes and cytosolic and extracted PLD fractions in a reaction
mixture containing 50 mM Na-HEPES (pH 7.5), 80 mM KCl, 1 mM dithiothreitol, 3 mM
MgCl2, 3 mM EGTA, and 2 mM
CaCl2 to give a final free Ca2+ concentration
of 300 nM (total, 0.1 ml) and stimulated with 10 µM GTP
S at 37 °C for 30 min in the presence of
butanol (0.3%, v/v). To measure oleate-dependent PLD
activity using exogenous substrate, the assays were carried out with
egg PC and [palmitoyl-3H]DPPC essentially
according to the method described by Massenburg et al. (29).
Reactions were terminated by the addition of chloroform/methanol (1:2,
v/v). Lipids were extracted and separated on Silica Gel 60 TLC plates
in a solvent system of the upper phase of ethyl acetate/2,2,4-trimethylpentane/acetic acid/water (13:2:3:10, v/v) as
described previously (30). The plates were exposed to iodine vapor, and
[3H]PBut was identified by comigration with PBut
standard. The spots scraped off the plates were mixed with
scintillation mixture, and the radioactivity was counted in a liquid
scintillation counter (Beckman LS-6500).
The membrane and cytosolic fractions were isolated from HL60 cells as described above. Membranes were washed once in buffer D (20 mM Tris-HCl, pH 7.4, 10 mM EGTA, 2 mM EDTA, 5 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and 10 µg/ml leupeptin) and then resuspended in buffer D containing 1% Triton X-100. After incubation at 4 °C for 60 min, the suspension was centrifuged at 100,000 × g for 60 min to obtain the membrane extract. Proteins of the membrane extracts and cytosolic fractions were separated by SDS-polyacrylamide gel electrophoresis on a 13% polyacrylamide gel (31) and then electrophoretically transferred onto nitrocellulose membrane (32). Blocking was performed in Tris-buffered saline containing 5% skimmed-milk powder and 0.05% Tween 20. Western blot analysis using specific antibodies was performed as described previously (33). The intensity of the bands was quantified by a densitometer (Densitograph; ATTO, Japan).
RNA Extraction and Reverse Transcription-Polymerase Chain Reaction (RT-PCR)Total RNA was isolated from HL60 cells by acid
guanidine thiocyanate method (34). 2 µg of RNA was
reverse-transcribed by using random hexamer mixed primers. Temperatures
used for PCR were: denaturation, 94 °C for 30 s; annealing,
58 °C for 1 min; and extension, 72 °C for 1 min. Primers for
hPLD1 were 5-(2549)-TGTCGTGATACCACTTCTGCCA-3
(sense) and
5
-(3080)-AGCATTTCGAGCTGCTGTTGAA-3
(antisense) (35). For
normalization, glyceraldehyde-3-phosphate dehydrogenase was amplified
simultaneously using the sense primer 5
-(27)-ACGGATTTGGTCGTATTGGG-3
and the antisense primer 5
-(257)-TGATTTTGGAGGGATCTCGC-3
(36). Amplification cycles were determined for individual primer sets to
maintain an exponential rate of product amplification. Amplified DNA
fragments were subjected to electrophoresis on a 1.5% agarose gel and
visualized by ethidium bromide staining. The intensity of the bands was
quantified by a densitometer (Densitograph; ATTO, Japan).
The temporal correlation between cell
differentiation and PLD activity was investigated by examining changes
in GTPS-dependent PLD activity in HL60 cell lysates at
various stages during differentiation. Differentiation was initiated by
the addition of 0.5 mM dbcAMP and assessed by NBT reduction
test. Less than 5% of cells expressed NBT reduction activity in the
undifferentiated state. With dbcAMP treatment, NBT-positive cells
increased in a time-dependent manner and reached the
maximal level at 72 h (Fig. 1A). As
shown in Fig. 1B, PLD activation induced by 10 µM GTP
S in the cell lysates was up-regulated after
dbcAMP treatment and reached an almost maximal level at 72 h (an
approximately 4-fold increase compared with that of undifferentiated
cells). The temporal profile of NBT reducing activity showed a good
correlation with that of GTP
S-dependent PLD activation.
In addition, ATRA, which is also known to induce granulocytic
differentiation of HL60 cells, caused an almost identical increase in
NBT-positive cells (Fig. 1C). GTP
S-induced PLD activity in HL60 cells was also increased by ATRA treatment (an approximately 1.8-fold increase compared with that of undifferentiated cells) but the
magnitude was lower than that observed in dbcAMP-treated cells (Fig.
1D).
GTP
The reconstitution
systems from both undifferentiated and differentiated HL60 cells were
designed to further investigate the mechanisms for the increase in
GTPS-dependent PLD activity during differentiation.
Recent studies with the cell-free system (3) and cytosol-depleted
permeabilized cells (4) have indicated that some cytosolic factors are
required for activation of PLD by GTP
S. We have examined PLD
activation by GTP
S in the membranes from undifferentiated and
differentiated HL60 cells labeled with [3H]oleic acid. In
the absence of the cytosolic fraction, GTP
S caused a marginal PLD
activation in undifferentiated membranes (Fig.
2A). However, when the cytosolic fraction
from undifferentiated HL60 cells was present in the reaction mixture,
the PLD activity was enhanced nearly 4-fold by the addition of 10 µM GTP
S (Fig. 2A). In contrast, as shown in
Fig. 2B, GTP
S and the cytosolic fraction exerted much
greater effects on the membrane PLD activation in differentiated HL60
cells (an approximately 3-fold increase compared with that of
undifferentiated cells). When the GTP
S-dependent PLD
activity in the cytosolic as well as membrane fractions was determined
using the exogenous substrate
(phosphatidylethanolamine/phosphatidylinositol 4,5-bisphosphate/egg PC, 10:1.5:1), the differentiation induced by
dbcAMP caused a small but significant enhancement in cytosolic PLD
activity (Fig. 3). The GTP
S-induced PLD activation in
both undifferentiated and differentiated membranes was dependent on the
cytosolic fraction from undifferentiated and differentiated cells,
respectively. Furthermore, the maximal PLD activation in membranes from
differentiated cells was approximately 3-fold higher than that in
undifferentiated membranes (Fig. 3B). Thus, similar results
were obtained in both endogenous and exogenous substrate systems.
Because recent studies have demonstrated the implication of Arf and Rho
family proteins as cytosolic factors in the regulation of PLD activity
in HL60 cells (5, 6, 9), changes in the contents of Arf and Rho family
proteins in the cytosolic fractions were examined by Western blot
analysis. As shown in Fig. 4, the contents of Arf, Rac2,
and Cdc42Hs in the cytosolic fractions from differentiated cells
(Cyt-D) were approximately 3.5-fold higher than those in
undifferentiated cells (Cyt-UD). In contrast, there were no
significant increases in the levels of RhoA and Rac1 (Fig. 4).
In the next experiments, the effects of the cytosolic fraction from
undifferentiated or differentiated cells were examined for
GTPS-dependent PLD activity in undifferentiated
membranes (Mem-UD). Two cytosolic fractions
(Cyt-D and Cyt-UD) enhanced PLD activity in a
concentration-dependent manner (Fig. 5).
However, relatively modest differences were observed between the
undifferentiated and differentiated cytosolic fractions. The PLD
activation by GTP
S in membranes isolated from undifferentiated or
differentiated cells was compared in the presence of undifferentiated
cytosol (Cyt-UD). The activation of
GTP
S-dependent PLD in membranes (Mem-D) from
differentiated cells was markedly distinct from that in membranes (Mem-UD) from undifferentiated cells (Fig.
6). These results suggested that the increased PLD
activity during differentiation was caused primarily by changes of
membrane component(s), e.g. activating factor(s) and/or PLD
itself, but not of cytosolic factors.
Changes in GTP
PLD activity in differentiated membranes was stimulated
by GTPS even in the absence of cytosolic fractions (Fig.
7A), suggesting the existence of
membrane-associated small GTP-binding protein(s). Therefore, the effect
of C3 toxin, which inhibits RhoA (27), on GTP
S-induced PLD
activation was examined. The GTP
S-induced PLD activation in
differentiated HL60 membranes was suppressed in a
concentration-dependent manner by C3 toxin in the presence of
10 µM NAD. Nearly maximal inhibition was obtained at 1 µg/ml (data not shown). Under the same condition, RhoA in
differentiated HL60 membranes was ADP-ribosylated by 1 µg/ml C3
toxin, as assessed by the incorporation of [32P]NAD (data
not shown). As shown in Fig. 7B, pretreatment of
differentiated HL60 membranes with 1 µg/ml C3 toxin caused an
approximately 65% inhibition of GTP
S-induced PLD activation. As
demonstrated by Siddiqi et al. (9), in addition to RhoA,
Cdc42Hs and Rac were able to activate PLD in HL60 membranes. Thus, to
further examine the roles of Cdc42Hs and Rac in PLD activation of
differentiated HL60 membranes, we used Toxin B, which was recently
shown to glucosylate Rho family proteins and inhibit their interaction
with effectors (28). As shown in Fig. 7B, GTP
S-induced
PLD activation in differentiated HL60 membranes was almost abolished by
Toxin B treatment. These results suggest that RhoA may play a major
role in GTP
S-induced PLD activation in differentiated HL60
membranes. Then, the contents of Rho family proteins in membranes of
both undifferentiated and differentiated cells were examined by Western
blot analysis. As shown in Fig. 8, A and
B, the amounts of RhoA greatly increased in membranes during
differentiation. There was a smaller but appreciable increase in
Cdc42Hs. The levels of Rac1 and Rac2 in membranes were unchanged during
differentiation. Furthermore, the RhoA contents of HL60 membranes also
increased during ATRA-induced differentiation (Fig. 8, C and
D). In contrast, brefeldin A, which is often used as an
inhibitor of Arf protein (37), did not affect PLD activation by GTP
S
(data not shown). In addition, the level of Arf1 in membranes as
assessed by Western blot analysis was not altered during
differentiation induced by dbcAMP and ATRA (data not shown).
Effects of Arf1 and Oleate on Membrane-associated PLD Activity from Undifferentiated and Differentiated HL60 Cells
At least two types
of PLD have been described in several tissues: the small GTP-binding
proteins-dependent type and the
oleate-dependent type (29, 38-40). The effects of Arf1 and
sodium oleate were examined for PLD activity in undifferentiated and
differentiated membranes. The Arf1-stimulated PLD activity in
differentiated membranes (Mem-D) was found to be 3.5-fold
higher than that in undifferentiated membranes (Mem-UD, Fig.
9A). On the other hand, sodium oleate (4 mM) failed to activate PLD in membranes from either
undifferentiated or differentiated cells (Fig. 9B). The oleate-dependent PLD activity in rat brain membranes was
observed under the same assay condition (data not shown).
To separate PLD from RhoA, membranes were treated with 1% cholate and
subjected to heparin-Sepharose column chromatography. RhoA was
recovered in the flow-through fraction (Fig.
10A), in which Arf1-dependent
PLD activity was not detected (Fig. 10B). In contrast,
Arf1-dependent PLD activity was obtained in 1 M
NaCl extract, in which RhoA protein was undetectable by Western blot analysis (Fig. 10, A and B). The effects of
recombinant Arf1 on this partially purified PLD from undifferentiated
and differentiated membranes were examined. As shown in Fig.
10C, GTPS in the presence of recombinant Arf1, but
neither one alone, enhanced PLD activity. The Arf1-mediated PLD
activity from differentiated membranes (PLD-D) was found to
be nearly 3-fold higher than that from undifferentiated membranes
(PLD-UD, Fig. 10C).
Changes in hPLD1 mRNA Level during HL60 Cell Differentiation
Recently, a gene encoding
Arf-dependent PLD (hPLD1) has been cloned from the HeLa
cell cDNA library (35). To investigate the possible changes
responsible for the increase in Arf-induced PLD activation during
differentiation, hPLD1 mRNA levels were examined by RT-PCR. As
shown in Fig. 11, A and B,
treatment of HL60 cells with dbcAMP resulted in the elevated expression
of hPLD1 mRNA. The hPLD1 messages nearly reached maximal levels at 48 h. Increases in the levels of hPLD1 mRNA were also observed in ATRA-treated HL60 cells (Fig. 11, C and D). In
contrast with dbcAMP treatment, a relatively modest elevation was
observed in hPLD1 by RT-PCR. Because the antibody for hPLD1 is not yet
available, we are unable to determine the level of hPLD1 protein.
However, the observed increase in hPLD1 mRNA levels leads us to
presume that its protein level would be increased in differentiated
HL60 cells induced by dbcAMP or ATRA.
The possible involvement of PLD in secretory response and
O 2 or H2O2 generation has
been indicated in neutrophils and differentiated HL60 cells (16-24).
Because Arf, an activator of PLD, is known to play an important role in
vesicular trafficking (44), Arf-mediated PLD activation can be
considered to take part in the vesicular transport process (42, 45). A
recent report (46) showed evidence that PLD plays important roles in
cytoskeletal organization. These findings also suggest that Arf- and
Rho-mediated PLD activity may be involved in the regulation of
bactericidal function and motile responses in neutrophils. HL60 cells
could be a useful model to verify these hypotheses because the cells
are endowed with neutrophil-like functions during differentiation. It
is well documented that HL60 cells are differentiated into
granulocyte-like cells by treatment with dbcAMP (12, 17, 19, 47, 48). Xie et al. (17) have demonstrated that dbcAMP-differentiated but not undifferentiated HL60 cells exhibited receptor-mediated PLD
activation. HL60 cell differentiation is accompanied by the expression
of cell surface receptors for fMLP (15). Moreover, PLD activities
stimulated by 4
-phorbol 12-myristate 13-acetate in intact cells and
by GTP
S in electropermeabilized cells were increased during HL60
differentiation (17). We have also observed that PLD activity in
response to fMLP or 4
-phorbol 12-myristate 13-acetate was enhanced
in dbcAMP-treated HL60 cells (data not shown). The increase in PLD
activity during differentiation correlated well with that of NBT
reduction activity (Fig. 1). These findings have prompted us to
investigate the mechanism underlying increased PLD activity during
differentiation. GTP
S-dependent PLD activity increased
in a time-dependent manner during differentiation induced by dbcAMP in HL60 cells (Fig. 1), suggesting a lack of either activating factor(s) for PLD or of PLD itself in undifferentiated cells. In fact, the protein levels of Arf, Rac2, and Cdc42Hs in the
HL60 cytosolic fraction were elevated during differentiation (Fig. 4).
However, the levels of cytosolic small GTP-binding proteins for the
membrane-associated PLD activation were thought to be sufficient in
undifferentiated cells because the membrane-associated PLD activity was
equally activated by the cytosolic fractions from either
undifferentiated or differentiated cells (Fig. 5).
It was noted that the PLD activity in differentiated membrane was
higher than that in undifferentiated membrane (Fig. 7A), suggesting that activating factor(s) for PLD activity and/or PLD itself
in membrane were increased during differentiation. Indeed, C3 toxin and
Toxin B effectively inhibited GTPS-induced [3H]PBut
formation (Fig. 7B). These results indicate that RhoA may play a major role in GTP
S-induced PLD activation in differentiated HL60 membranes. This notion was supported by Western blot analysis showing the increases in RhoA and Cdc42Hs in differentiated membranes (Fig. 8). The enhancement of GTP
S-dependent PLD activity
during differentiation can be explained, at least in part, by an
increase in RhoA content in the membrane. RhoA and Arf exert a
synergistic effect on PLD activation in several cell types, including
HL60 cells (9, 41, 43). On the other hand, our previous study (10) has
shown that Rho-specific GDP dissociation inhibitor, which removes Rho
proteins from membranes, suppressed synergistic PLD activation by
GTP
S and protein kinase C in HL60 membranes and has also shown that
this suppressed PLD activation was restored by the addition of
recombinant RhoA. Thus, RhoA plays a key role in Arf- and protein
kinase C-induced PLD activity in HL60 cells.
On the other hand, the increase in RhoA content in membranes could be explained by its translocation from cytosol or newly synthesized RhoA to membrane probably after posttranslational modification (geranyl geranylation). The pretreatment of cells with mevastatin to inhibit hydroxymethylglutaryl-CoA reductase involved in the formation of isoprenoids (49) did not prevent an increase in membrane-associated RhoA during differentiation induced by dbcAMP (data not shown). Thus, the former hypothesis is more likely. Cytosolic Rho family proteins form complexes with the Rho-specific GDP dissociation inhibitor in the GDP-bound form (50). Translocation to membrane, which is believed to be an activation process, requires dissociation from the Rho-specific GDP dissociation inhibitor and an exchange of bound GDP for GTP (48). Aepfelbacher et al. (51) have demonstrated that monocytic differentiation in U-937 cells was associated with an increase in membrane-associated Cdc42Hs. However, its detailed mechanism is not clearly understood. Further investigation of the increases in membrane-associated RhoA during differentiation will provide a clue toward the better understanding of the mechanism underlying the membrane translocation of RhoA.
There is an alternate possibility for the differentiation-dependent increase in the membrane PLD activity. Arf-mediated PLD activity in differentiated membranes was shown to be higher than that in undifferentiated membranes (Figs. 9 and 10), suggesting the increased expression of membrane-associated PLD during differentiation. This notion was supported by RT-PCR with primers designed for hPLD1. The hPLD1 mRNA levels were increased by dbcAMP or ATRA treatment of HL60 cells (Fig. 11). However, at present we do not know whether the hPLD1 protein level was increased because its antibody is not yet available.
In summary, GTPS-dependent membrane PLD activity was
elevated during HL60 cell differentiation induced by dbcAMP. Similar results were obtained in experiments with differentiated HL60 cells
induced by ATRA, thus supporting the notion that the increase in RhoA
content in membranes and changes in PLD expression are associated with
granulocytic differentiation. However, additional studies are required
to disclose the precise mechanism for the increased GTP
S-induced PLD
activation during the differentiation of HL60 cells.
We gratefully acknowledge Dr. Joel Moss for the kind gifts of E. coli bearing Arf1 plasmid and monoclonal antibody against Arf, Dr. Alan Hall for E. coli bearing C3 transferase plasmid, and Dr. David M. Lyerly for C. difficile Toxin B.
After submission of this manuscript, a paper was published describing the lysophosphatidic acid-induced translocation of RhoA in Rat1 fibroblasts and its possible implication in PLD activation (Malcolm, K. C., Elliott, C. M., and Exton, J. H. (1996) J. Biol. Chem. 271, 13135-13139).