(Received for publication, November 11, 1996, and in revised form, February 10, 1997)
From the Department of Pediatrics and The Research
Institute, The Hospital for Sick Children,
Toronto, Ontario, Canada M5G 1X and the Departments of
¶ Pharmaceutical Sciences and
Molecular Pharmacology, St.
Jude Children's Research Hospital, Memphis, Tennessee 38105
The P-glycoprotein (Pgp) reversing
agent, reserpine, induces MDR1 mRNA and PGP protein in
human colon carcinoma cells (Schuetz, E. G., Beck, W. T., and Schuetz,
J. D. (1996) Mol. Pharmacol. 49, 311-318) and in H35 rat
hepatoma cells. Reserpine's interference with cellular dopamine
utilization suggested that dopamine and dopaminergics might be
important physiological regulators of PGP expression. Initial studies
demonstrated that the H35 cells express the D2 dopamine receptor. Pgp
protein and pgp2/mdr1b mRNA was increased (maximum of
10- and 8-fold, respectively) by the potent D2 dopamine receptor
agonists bromocriptine, R()-propylnorapomorphine hydrochloride, and quinpirole, and Pgp protein induction was blocked by
D2 receptor antagonists spiperone and clozapine. D2 receptor agonist
induction of pgp2/mdr1b mRNA was paralleled by
transcriptional activation of the pgp2/mdr1b promoter but
blocked by pretreatment with the D2 dopamine receptor antagonists,
spiperone, eticlopride, and clozapine. Co-transfection of a D2 dopamine
receptor expression vector enhanced bromocriptine's transcriptional
activation of the pgp2/mdr1b promoter. The G-protein,
G
i2, is required for bromocriptine transcriptional
activation because the G-protein inhibitor, pertussis toxin, suppressed
bromocriptine's activation of pgp2/mdr1b transcription and
co-transfection of a dominant negative G
i2 abrogated
bromocriptine activation of pgp2/mdr1b. Gi
proteins can transduce signals by activation of mitogen-activated protein kinases (MAPKs), and because Raf-1 is a known activator of
MDR1, we tested for Raf-1 involvement. Co-transfection of a dominant negative Raf-1 failed to block bromocriptine induction of
pgp2/mdr1b, and bromocriptine treatment caused no
phosphorylation of the MAP kinase kinase substrates p42 and p44,
demonstrating that the MAP kinase pathway was not involved. These are
the first studies demonstrating transcriptional activation of an
MDR gene by dopamine receptor agonists and that this
activation occurs by a signal transduction pathway requiring the D2
dopamine receptor coupled to a functional G-protein.
The multidrug resistance (MDR)1 gene family encodes a small family of plasma membrane ATP-dependent efflux transporters, referred to as P-glycoproteins (PGPs) (1). The MDR genes are part of a small gene family that is composed of three members in rodents and two in humans (2-7) for which cDNAs have been isolated and characterized. Full-length cDNAs for mouse mdr1 (3, 8), mouse mdr3 (3), and human MDR1 (4) but not mouse mdr2 (6) or human MDR2 (MDR3 (9)) can confer the multidrug-resistant phenotype when transfected and overexpressed in drug-sensitive cells. High levels of expression of the multidrug resistance gene (MDR1) commonly occur in human cancers derived from normal tissues that express PGP, such as carcinomas of the liver, colon, kidney, and pancreas and may contribute to the drug resistance of these cancers. The PGPs are involved in the transport of a variety of substances such as peptides (10), endogenous steroids (11), and xenobiotics (12) and may, under certain physiological conditions, function as a chloride ion channel (13). Since we and others (14-16) have shown that PGP expression and transcription can be regulated by substances it transports (e.g. steroids), it seemed possible that agents that interfered with the pump, but had no known cytotoxic effect (e.g. reversing agents), might provide insight into endogenous physiological pathways regulating PGP gene expression.
Although a number of the transcription factors that regulate the multidrug resistance genes have been identified (17, 18) very little is known about the molecular signals activating PGP expression in response to putative substrates, ligands, or modulators. In one example, Fojo et al. (19) demonstrated that PGP reversing agents, such as verapamil and cyclosporin A, increase MDR1 mRNA expression in a human colon carcinoma cell line. We and others (20) have found that a variety of agents, including the MDR1 reversing agent reserpine, increase MDR1 gene expression in these same cells. In similar studies we have also found that reserpine induces the amount of pgp2/mdr1b mRNA in the H35 rat hepatoma cell line by transcriptional activation of the pgp2/mdr1b gene2; however, the mechanism by which the pgp2/mdr1b gene is activated by reserpine is unknown.
Because reserpine up-regulates the synthesis of dopamine (21-23), inhibits the dopamine transporter (24, 25), and increases dopamine receptor RNA (18, 26), we hypothesized that dopamine or dopaminergics might serve as endogenous physiological regulators of MDR gene expression. Using H35 hepatoma cells (27) we have defined the initial components of the D2 dopamine receptor signal transduction cascade leading to transcriptional activation of pgp2/mdr1b. We used specific D2 and D1 dopamine receptor agonists and antagonists, as well as D1 and D2 receptor expression vectors, to define the role of the classical D2 dopamine receptor in pgp2/mdr1b gene activation. In total, these studies reveal a novel D2 dopamine receptor-mediated transcriptional activation pathway for the pgp2/mdr1b gene in the H35 rat hepatoma cells that is coupled to G-proteins.
Reuber H35 rat hepatoma cells (American Type
Culture Collection, Rockville, MD) were maintained in a minimal
essential medium containing 10% fetal calf serum supplemented with
penicillin, streptomycin, and glutamine at 37 °C in 5%
CO2. All drugs used used at a final concentration of 10 µM, except where stated otherwise. Pertussis toxin was
used at a final concentration of 100 ng/ml of media. Drug-containing
medium was changed every 24 h with freshly supplemented medium.
(R)-()-Propylnorapomorphine hydrochloride (NPA),
quinpirole, SCH23390, spiperone, SKF38393, clozapine, bromocriptine, eticlopride, and pertussis toxin were obtained from Research
Biochemicals International (Natick, MA). All other drugs and chemicals
were obtained from Sigma.
Total RNA was isolated from cells
pooled from one 100-mm tissue culture dish using the phenol-chloroform
method (28). Northern blot analysis was performed as described
previously (29) on 20 µg of total RNA. The integrity of the RNA and
evenness of loading after transfer to a positively charged membrane
(Magna NT, MSI Separations, Westborough, MA) was confirmed by
comparison of the 18 and 28 S ribosomal bands which were apparent with
ethidium bromide staining. Blots were probed with a specific
pgp2/mdr1b oligonucleotide (14) labeled with
[-32P]ATP using the 5
-DNA terminus labeling kit (Life
Technologies, Inc.). These same blots were re-probed with a cDNA
probe for cyclophilin (kindly provided by Dr. J. Sutcliffe (30)).
First strand cDNA was prepared by reverse transcription of 8 µg of total RNA using 200 ng of random primers (Pharmacia Biotech Inc.) and 200 units of Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc.). The amount of first strand cDNA used in polymerase chain reaction (PCR) amplification was increased stepwise from 12.5 to 200.0 ng. PCR reactions were performed in 100-µl final volumes using rat pgp2/mdr1b gene-specific primers (Center for Biotechnology, St. Jude Children's Research Hospital). The pgp2/mdr1b sense primer corresponded to bp 3533-3562; pgp2/mdr1b antisense primer corresponded to bp 3835-3864 of the cDNA sequence (31, 32). Aliquots of the PCR reaction were then separated on a 1.0% NuSieve, 0.5% agarose gel, demonstrating a 332-bp amplification product. The gel was transferred to nylon membrane and probed with an internal pgp2/mdr1b oligonucleotide. Amplification of a 202-bp fragment of the glyceraldehyde-3-phosphate (GAPDH) cDNA (using published oligonucleotide sequences (33)) was chosen as an internal control for normalization because its level in cells in tissue culture has been shown to be independent of culture confluency and xenobiotic treatment (33, 34). Quantitative comparisons were made over the linear range of amplification for each treatment group after each blot was probed with a GAPDH or pgp2/mdr1b oligonucleotide specific to internal sequences of the amplimer and densitometric measurement of band intensity.
RT-PCR: Dopamine ReceptorsTo demonstrate the presence of dopamine receptor mRNA in H35 cells, first strand cDNA reverse-transcribed from H35 rat hepatoma cell total RNA was used in the PCR assay (35). Oligonucleotide primer pairs used to amplify the D1 dopamine receptor and D2 dopamine receptor short form (D2S) and long form (D2L) were synthesized (Center for Biotechnology, St. Jude Children's Research Hospital) using the sequences published by Rao et al. (35). D1, D2S, and D2L dopamine receptor expression vectors (36) were used as specific, positive controls (kindly provided by Dr. S. Senogles, University of Tennessee, Memphis).
Immunoblot AnalysisCrude membranes were extracted from H35 rat hepatoma cells using a modified method of Lee et al. (37). Cells were scraped from the dishes in phosphate-buffered saline and were pelleted at 10,000 × g at 4 °C. The pellet was resuspended in membrane storage buffer (MSB; 100 mM potassium phosphate (pH 7.4), 1.0 mM EDTA, 20% glycerol, 1 mM dithiothreitol, 20 µM butylated hydroxytoluene, and 2 mM phenylmethylsulfonyl fluoride) and lysed for 35 s at 30% power with an Ultrasonic homogenizer (Cole Parmer Corp., Chicago, IL). The crude membranes were isolated by centrifugation at 10,000 × g for 5 min at 4 °C. This crude membrane pellet was resuspended in a small volume of MSB. Protein determinations were done using the method of Lowry et al. (38). Thirty-five µg of crude membrane proteins were resuspended in standard Laemmli sample preparation buffer (39) and were immediately loaded onto a 7.5% polyacrylamide gel and resolved overnight. Proteins were transferred to Protran® nitrocellulose filters (Schleicher & Schuell) as described previously (37, 40). Filters were incubated sequentially with primary polyclonal rabbit anti-mdr(ab-1) IgG (Oncogene Science, Uniondale, NY) and peroxidase-conjugated anti-rabbit IgG and developed using the Amersham enhanced chemiluminescence detection system. The relative amount of Pgp was determined by densitometric analysis.
VectorsA 519-bp fragment containing the promoter of the
pgp2/mdr1b (369 to +150 bp) gene was amplified by PCR and
fused to either a luciferase or chloramphenicol reporter as described
previously (14). Activated raf-1 and dominant negative raf expression
vectors were provided by Dr. John Cleveland (St. Jude Children's
Research Hospital, Memphis, TN). G
o (A205L) was provided
by Dr. Henry Bourne (University of California San Francisco, CA).
Wild-type G
i2 and dominant negative G
i2
S48C (41) were from Dr. Melvin Simon (California Institute of
Technology, Pasadena, CA).
H35 rat hepatoma cells were subcultured by trypsinization and plated at 3-4 × 105 cells per 60-mm tissue cultures dishes. When the cells had reached approximately 25-35% confluence, they were transfected for 18 h with 10 µg of plasmid DNA by the calcium-phosphate co-precipitation method (42). The H35 cells were then washed once with medium, and fresh medium with drug was added. After a 24-h treatment cells were harvested for either chloramphenicol acetyltransferase or luciferase assays.
H35 cells were co-transfected with
10 µg of a RSV promoter-driven -galactosidase expression plasmid
to normalize for transfection efficiency (43). The
-galactosidase
assay was performed using standard methods (43).
After washing once with phosphate-buffered saline, the cells were briefly incubated in a CAT harvest buffer (150 mM NaCl, 40 mM Tris (pH 7.4), 5 mM EDTA). The cells were scraped from the 60-mm tissue culture dishes, and cellular CAT activity was assayed as described previously (44, 45) with the exception that the H35 cell protein extracts (60 µg) were heat-inactivated for 15 min at 65 °C to destroy endogenous acetylase activity. CAT activity relative to the untreated control dishes was determined after subtraction of background activity obtained from mock transfected control dishes.
Luciferase AssayH35 cells were washed twice in phosphate-buffered saline, incubated for 15 min in Reporter Lysis buffer (Promega, Madison, WI), and scraped from the culture dishes. Lysate protein concentrations were determined using the method of Lowry et al. (38). Luciferase activity in 20 µg of cell protein extract was measured according to the manufacturer's instructions (Luciferase Assay Kit, Promega, Madison, WI) using an Optocomp 1 Luminometer (MGM Instruments, Hamden, CT) with a counting window of 10 s.
MTT AssayThe assay was performed essentially as described (46). H35 cells were subcultured by trypsinization and plated at various densities in 96-well microtiter plates. Fresh medium was added before drug treatment. After a 24-h treatment medium was aspirated, and cells were washed with phosphate-buffered saline, and the MTT reagent was added to a final concentration of 2 mg/ml. Following a 3-h incubation period at 37 °C the plates were spun at 500 × g for 5 min, the MTT reagent aspirated, dimethyl sulfoxide was added, and the plates were read using a Thermomax microplate reader at the test wavelength of 590 nm and the reference 650 nm. The assay was read within the linear range with an r2 = 0.94 when comparing cell number versus the absorbance ratio.
p42/p44 PhosphorylationH35 rat hepatoma cells were plated into 2 ml of complete medium at a density of 2.5 × 105 cells/well in six-well plates (Corning-Costar, Cambridge, MA). After 2 days of incubation at 37 °C in a humidified atmosphere, the medium was removed from the adhered cells, and 2 ml of serum-free medium was added to each well. After 2 days of starvation, the quiescent cells were stimulated by direct addition of bromocriptine (10 µM) or fetal bovine serum (10%). Cells were incubated at 37 °C in 5% CO2 for the duration outlined by the time course assay; stimulation was terminated by removal of the medium. The cells were then washed with 2 ml of ice-cold phosphate-buffered saline prior to lysis for Western blot analysis.
Cells were lysed with Laemmli sample preparation buffer and were briefly sonicated. The cell lysates were heated at 95 °C for 5 min, cooled on ice, and then were centrifuged at 14,000 × g for 5 min prior to gel electrophoresis. Cell lysate proteins were loaded onto a 10% SDS-polyacrylamide minigel with resolution at 200 V for approximately 45 min followed by electrotransblotting onto polyvinylidene difluoride membrane (0.2-micron pore size, Bio-Rad) at 100 V for 1.5 h at 4 °C.
Immunoblotting was performed at room temperature. The membrane was blocked for 1 h in 5% non-fat dry milk (Bio-Rad) and incubated overnight with the phospho-specific MAPK antibody. Rabbit polyclonal phospho-specific MAPK antibody (New England Biolabs, Beverly, MA) was raised against a synthetic phosphotyrosine peptide comprised of residues 196-209 (DHTGFLTEY(P)VATRWC) of the human p44MAPK. This antibody recognizes only p42 and p44 that is catalytically active due to phosphorylation at tyrosine 204. Goat anti-rabbit IgG conjugated with alkaline phosphatase was used as a secondary antibody. Alkaline phosphatase signal was detected using the PhototopeR Chemiluminescent Western Detection System (New England Biolabs, Beverly, MA) with Kodak XAR-2 film (Eastman Kodak).
We recently demonstrated that the PGP reversing agent, reserpine,
can up-regulate human MDR1 gene expression in a human colon carcinoma cell line (20) and in primary cultures of human
hepatocytes.2 Similarly, we have found that reserpine can
up-regulate expression of the rat pgp2/mdr1b gene in rat H35
Reuber hepatoma cells.2 Because reserpine can alter the
expression of the dopamine receptor in some tissues (18, 26), we
speculated that the pgp2/mdr1b gene could be up-regulated by
reserpine by signaling through the dopamine receptor. We first
determined if the H35 cells express the dopamine receptor. Both
functional studies (activation of Na+/K+-ATPase
(47)) and PCR analysis (35, 47) have previously demonstrated that the
liver expresses the D2 dopamine receptor. We used PCR primers (35) that
spanned the region where alternate splicing creates either a long (D2L)
or short form (D2S) of the D2 dopamine receptor to generate a cDNA
from H35 cells. The D2L and D2S (28 amino acids shorter than D2L)
dopamine receptor isoforms can readily be distinguished on agarose gels
(Fig. 1). The specificity of the D2 dopamine receptor
oligonucleotides for the D2 receptor was demonstrated by testing them
against templates of cloned authentic D2L, D2S, or D1 dopamine
receptors (Fig. 1). Since the D2L and D2S share common sequences we
found that amplification readily occurred using the D2L and D2S
dopamine receptor templates as anticipated, whereas no amplification
was observed using the unrelated D1 dopamine receptor template. When
these same primers were incubated with the H35-derived cDNA, we
found amplification of both D2S and D2L dopamine receptor isoforms,
with the D2L isoform mRNA amplified to a greater extent. We cannot
with certainty state how much of the corresponding proteins are made
because of the lack of suitable reagents to detect the D2L and D2S
isoforms in these cells.
Next, we treated H35 cells for 24 h with the potent D2 receptor
agonist bromocriptine. Bromocriptine treatment resulted in a
dose-dependent increase in Pgp expression (2-fold by 0.1 µM drug and up to 10-fold by 100 µM drug)
(Fig. 2). Bromocriptine also up-regulated
pgp2/mdr1b mRNA (up to 10-fold at 10 µM
drug) (Fig. 3A), whereas dopamine was less
effective than bromocriptine as an inducer of pgp2/mdr1b
mRNA (Fig. 3A). The latter finding can in all likelihood
be attributed to the rapid oxidation and cellular metabolism of
dopamine in culture (48, 49). We confirmed and extended the Northern
blot result by performing RT-PCR with pgp2/mdr1b-specific
primers on first strand cDNA prepared from RNA isolated from H35
cells exposed to varying concentrations of bromocriptine (Fig.
3B). pgp2/mdr1b mRNA was
dose-dependently increased, to a maximum of 15- and 50-fold
above control at 10 and 50 µM bromocriptine,
respectively. While bromocriptine has been reported to have some
effects on cell viability (50, 51), we found that acute bromocriptine
exposure had no effect on either cell cycle pattern or viability as
assessed by the MTT assay. A 24-h treatment of H35 cells with 10 µM bromocriptine produced no significant difference in
the tetrazolium dye signal compared with the control cells
(bromocriptine = 0.277 ± 0.072 (n = 12), and
control = 0.264 ± 0.04 (n = 12)). Thus, the
increase in pgp2/mdr1b expression by bromocriptine is not
secondary to an acute cytotoxic insult.
To further confirm a role for the D2 dopamine receptor in
pgp2/mdr1b gene expression, we determined whether endogenous
Pgp expression could be altered by a series of known agonists and antagonists specific for the D2 receptor. Treatment of H35 cells with
the D2 dopamine receptor agonists, NPA and quinpirole, increased the
expression of Pgp (Fig. 4). Agonist induction of Pgp
expression was antagonized by pretreatment with the D2 dopamine
receptor antagonists spiperone and clozapine, whereas the antagonists
themselves had little effect on Pgp expression.
To assess whether bromocriptine transcriptionally activated the
pgp2/mdr1b gene, H35 rat hepatoma cells were transiently
transfected with the Pgp2LUC construct containing the
pgp2/mdr1b promoter (bp 369 to +150) and treated with
bromocriptine (Fig. 5). There was significant induction
of Pgp2LUC by bromocriptine (up to 12-fold); maximal transcriptional
activation of Pgp2LUC occurred between 10 and 50 µM
bromocriptine (Fig. 5A) with an estimated EC50
of approximately 0.5 µM. Moreover, the transcriptional
activation of the pgp2/mdr1b promoter was specific because
neither the vector control (pGL2-Basic)2 nor RSV-LUC (Fig.
5A) was transcriptionally activated by bromocriptine. Similar bromocriptine-mediated activation of the identical
pgp2/mdr1b promoter when it was fused to a CAT reporter
(Pgp2CAT (14) (Fig. 5B) ruled out the possibility that
transcriptional activation of the pgp2/mdr1b promoter was
due to selective stabilization of the luciferase gene product by
bromocriptine. The decreased magnitude of bromocriptine induction for
the CAT reporter construct is most likely due to the non-signal
sequence-dependent export of CAT into the media (52), a
finding we have previously noted (53).
To determine the ligand specificity of the transcriptional activation of the pgp2/mdr1b gene, we transiently transfected H35 cells with only Pgp2LUC and treated the transfectants with ligands for the following receptors: dopamine, adrenergic, serotonin, and Sigma receptor agonists (Table I). Addition of the D1 receptor agonist, SKF38393 at doses from 0.1 to 50 µM, or addition of agonists for other receptors (adrenergic, serotinergic, and Sigma) did not transcriptionally activate the pgp2/mdr1b promoter thus demonstrating that only D2 dopamine receptor ligands transcriptionally activate the pgp2/mdr1b promoter.
|
We next evaluated whether pharmacological antagonists of the D2
dopamine receptor could block the transcriptional activation of the
pgp2/mdr1b promoter. H35 cells were transiently transfected with Pgp2LUC. A 1-h pretreatment with spiperone almost completely blocked bromocriptine activation of the pgp2/mdr1b promoter
(Fig. 6A), while pretreatment with D2
dopamine receptor antagonists of lower affinity (clozapine,
eticlopride) were less potent inhibitors of bromocriptine activation of
pgp2/mdr1b transcription, consistent with the tighter
binding of spiperone to the D2 dopamine receptor. Inhibition of the
pgp2/mdr1b promoter by the D2 dopamine receptor antagonists
appeared to be specific because no effect was seen when H35 cells were
preincubated with SCH23390, a D1 dopamine receptor antagonist prior to
bromocriptine addition (Fig. 6B). This finding complements
the studies shown in Table I by demonstrating that selective D2
dopamine receptor antagonists block bromocriptine activation of the
pgp2/mdr1b promoter.
Although the H35 cells express D2 dopamine receptor isoforms (Fig. 1),
we reasoned that we could enhance bromocriptine transcriptional activation of pgp2/mdr1b by co-transfection of the
expression vectors for the D2 dopamine receptor (Fig.
7). A 3.5-fold increase in luciferase activity was seen
in response to co-transfection of the long form of the D2 dopamine
receptor (D2L) with the Pgp2LUC construct. Addition of bromocriptine
further increased pgp2/mdr1b promoter activity to almost
9-fold above vector control transfectants. The D1 dopamine receptor
expression vector had no effect on transcriptional activity when
compared with the empty vector (pcDNA3) (Fig. 7). These
data show that only addition of the D2 receptor causes an increase in
both the basal and bromocriptine-inducible activity of the
pgp2/mdr1b promoter.
The D2 dopamine receptor upon binding its ligand activates a
transmembrane signaling pathway coupled to
Gi/Go proteins before converging on other
cellular effector molecules (36, 54). To examine the coupling of D2
dopamine receptor activation to a Gi/Go protein
and its role in bromocriptine activation of the pgp2/mdr1b
promoter, we transiently transfected H35 cells with Pgp2LUC. Prior to
bromocriptine treatment, we applied pertussis toxin to interfere with
the coupling between the endogenous D2 dopamine receptor and the
heteromeric G-proteins (36). Cells were then treated with varying
concentrations of bromocriptine (Fig. 8A).
Pertussis toxin treatment did not alter basal pgp2/mdr1b or
thymidine kinase promoter activity.2 In contrast, pertussis
toxin suppressed bromocriptine induction of the pgp2/mdr1b
promoter at all doses of bromocriptine. These studies indicate that a
majority of the bromocriptine-elicited activation of the
pgp2/mdr1b promoter requires coupling with
Gi/Go.
To define the Gi protein involved in bromocriptine signal
transduction, we co-transfected increasing amounts of dominant negative Gi2 S48C along with the pgp2/mdr1b promoter
(Fig. 8B). At low amounts of dominant negative
G
i2, bromocriptine induction of pgp2/mdr1b
transcription was unaffected. As the concentration of co-transfected
G
i2 increased, bromocriptine failed to transcriptionally activate the pgp2/mdr1b promoter. The effect of
G
i2 was specific because G
i2 had no
effect on the thymidine kinase promoter.2 To control for
the possibility that G
i2 might produce nonspecific effects we also co-transfected a G-protein not known to couple with D2
receptors, G
o (36, 54). Co-transfection of wild-type G
o had no effect on bromocriptine transcriptional
activation of pgp2/mdr1b (Fig. 8C). Combined with
the pertussis toxin findings, these data show that
pgp2/mdr1b transcriptional activation by bromocriptine
requires functional Gi
2.
Some Gi/Go-coupled receptors, such as the
thrombin receptor, are known to stimulate the MAP kinase pathway in a
pertussis toxin-sensitive manner (55, 56). Because the Raf-1 MAP kinase pathway has been proposed as a control point in the regulation of
MDR1 transcription (57, 58) and, furthermore, because the induction of pgp2/mdr1b by D2 dopamine receptor agonists is
specifically abrogated by a Gi2 dominant negative, we
reasoned that the Raf-1 MAP kinase pathway might be involved in the
downstream signal transduction cascade for bromocriptine activation of
pgp2/mdr1b. To assess a potential role of Raf-1 MAP kinase,
a plasmid expressing a dominant negative Raf (59) was co-transfected in
varying amounts into H35 cells to determine if bromocriptine's
activation of pgp2/mdr1b could be blocked (Fig.
9A). Consistent with the previous findings reported for the human MDR1 promoter (58) the dominant
negative Raf suppressed the pgp2/mdr1b promoter with maximal
suppression being over 80%. However, the dominant negative Raf had no
effect on pgp2/mdr1b transcriptional activation by
bromocriptine (Fig. 9B). To further confirm that the MAP
kinase pathway was functional and not perturbed by bromocriptine, we
directly assessed whether the MAP kinase kinase substrates p42 and p44
were phosphorylated in response to bromocriptine (Fig.
10). We demonstrated that the MAP kinase pathway was
active in the H35 cells by serum starving the cells for 48 h and
then stimulating them with fresh serum containing medium. The
phosphorylation of p42 and p44 was assessed at varying times afterward
(Fig. 10). The time course of p42 and p44 phosphorylation showed that
p42 and p44 are phosphorylated within 15 min of serum replacement and
that the phosphorylated p42 and p44 rapidly decreases thereafter. In
contrast, addition of bromocriptine to H35 cells produced no detectable
change in the phosphorylation of p42 and p44. Thus, these findings
indicted that while bromocriptine activation of the
pgp2/mdr1b promoter involves activation of the dopamine D2
receptor coupled to Gi, the Raf-1 MAP kinase pathway is not
the downstream effector leading to transcriptional activation of
pgp2/mdr1b.
We and others (20, 60) have previously shown that the PGP reversing agent reserpine can increase MDR1/PGP expression in vitro in rat and human cells and can activate transcription of the pgp2/mdr1b promoter.2 Because reserpine, a dopamine reuptake inhibitor, can affect expression of the dopamine receptor (18, 26) and since dopamine receptors are expressed in the liver (35, 61) and the H35 hepatoma cells (Fig. 1), we hypothesized that reserpine might induce pgp2/mdr1b by altering the amount of an endogenous substrate (dopamine) that serves as a natural intracellular controller of pgp2/mdr1b gene expression in H35 cells. In the present study, we have shown that a D2 dopamine receptor ligand, bromocriptine, can increase Pgp and pgp2/mdr1b mRNA expression in H35 rat hepatoma cells and that this correlates with increased transcriptional activity of the pgp2/mdr1b promoter. The specific involvement of the D2 dopamine receptor in bromocriptine transcriptional activation of the pgp2/mdr1b promoter was strongly indicated because (a) transcriptional activation was specific for D2 dopamine receptor agonists, (b) agonist activation of pgp2/mdr1b transcription could be blocked by D2 dopamine receptor antagonists. and (c) pgp2/mdr1b promoter activation by bromocriptine was enhanced only by the D2 dopamine receptor.
The signal transmitted by the D2 dopamine receptor, in the H35 cells,
required a functional Gi as demonstrated by (a)
the dramatic suppression of bromocriptine activation of the
pgp2/mdr1b promoter by pertussis toxin, and (b)
the specific abrogation of bromocriptine transcriptional activation by
the dominant negative Gi2. These findings support the
idea that a D2 dopamine receptor initiated transmembrane signal
transduction pathway being mediated by the G
i2. While D2
receptor activation would lead heterotrimeric G-proteins to dissociate
and activate downstream signaling pathways, either via
G
i GTP or G-protein
subunits, the cis-elements mediating transcriptional activation of pgp2/mdr1b are
unknown. While G
i can lead to AP-1 activation, and the
pgp2/mdr1b promoter contains an AP-1 site (14),
G
i activation of AP-1 requires the MAPK pathway that our
findings show is not involved in bromocriptine activation of
pgp2/mdr1b. It is also possible that bromocriptine mediates
its effect through transcription factors that are themselves directly
regulated by dopaminergic compounds. Clearly our future studies with
deletion constructs of the pgp2/mdr1b promoter will delineate the important cis-elements and additional intracellular signals required for pgp2/mdr1b transcriptional activation
by bromocriptine.
Since the type and amount of dopamine receptor varies from tissue to
tissue (35, 61, 62), the specific dopamine receptor isoform expressed
may be an important factor controlling Pgp expression. In normal rat
liver, Giros et al. (61) detected the long form of the D2
dopamine receptor by Northern blot analysis. Rao et al. (35)
similarly found in rats that the long form of the D2 dopamine receptor
was detectable in normal liver by RT-PCR assay. Similarly, we found
that both the long and short forms of the D2 dopamine receptor were
detectable, but we have no explanation for why the alternatively
spliced form would be detected in H35 cells and not in normal liver.
Nevertheless, both the short and long forms of the D2 dopamine receptor
couple via the guanine nucleotide-binding protein,
Gi/Go. Although the D2 dopamine receptor isoforms appear to utilize the same G-protein, it is clear that in
different tissues the second messenger pathways significantly differ.
For instance, in the pituitary, the D2 dopamine receptor couples via a
G-protein to produce a decrease in cAMP by inhibition of adenylate
cyclase (36, 63). In contrast, in isolated lactotrophs, D2 dopamine
receptor activation results in activation of K+ channels or
Ca2+ currents (63, 64). Other studies have suggested that
activation of the D2 dopamine receptor leads to induction of
phosphoinositide hydrolysis (65) or potentiation of arachidonic acid
release (66). In the H35 cells, transcriptional activation of
pgp2/mdr1b by bromocriptine required coupling to
Gi2; however, the downstream effector pathway is unknown
in these cells. It is unlikely that the MAP kinase pathway is involved
because the dominant negative Raf-1 did not abrogate
bromocriptine-induced pgp2/mdr1b gene activation, although
it did suppress basal pgp2/mdr1b promoter activity which is
consistent with previous reports of dominant negative Raf-1 effects on
the human MDR1 promoter (57, 58). This finding was further
supported by the fact that bromocriptine did not alter the
phosphorylation of p42 and p44 in H35 cells. Thus, our data indicate
that bromocriptine-induced transcriptional activation of
pgp2/mdr1b does not involve Raf-1.
To our knowledge, this is the first time that a dopaminergic pathway for the transcriptional activation of the pgp2/mdr1b promoter and transcriptional regulation of pgp2/mdr1b expression has been described. Further studies are necessary to delineate the downstream signaling pathways involved in the dopaminergic regulation of pgp2/mdr1b.
We gratefully acknowledge the comments of our colleagues Drs. Robert J. Rooney and Linda Harris. The excellent technical assistance of Amber Troutman is gratefully acknowledged.