| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
Neuroendocrine Unit, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Anne Klibanski, M.D., Neuroendocrine Unit, Bulfinch 457B, Massachusetts General Hospital, Boston, Massachusetts 02114. E-mail: aklibanski{at}partners.org.
Abstract
Activin is a member of the TGFß family of cytokines involved in the control of cell proliferation. We have previously shown that the majority of clinically nonfunctioning pituitary tumors do not respond to activin-induced growth suppression. Human pituitary tumors specifically express alternatively spliced activin type I receptor Alk4 mRNAs, producing C-terminus truncated isoforms designated Alk4-2, 4-3, and 4-4. However, it is not known whether these truncated activin receptors suppress activin effects on cell proliferation in human pituitary cells. Therefore, we investigated activin signaling in a human pituitary tumor cell line, HP75, derived from a clinically nonfunctioning pituitary tumor. HP75 cells express activin A mRNA and secrete activin A, as measured by ELISA and a functional bioassay. TGFß administration decreases the proliferation of HP75 cells, suggesting that the signaling pathway shared by TGFß and activin is functional in this cell line. However, activin neither inhibits cell proliferation nor stimulates reporter gene expression in HP75 cells, indicating that activin signaling is specifically blocked at the receptor level. HP75 cells express all truncated activin type I receptor Alk4 isoforms, as determined by RT-PCR. Because truncated Alk4 receptor isoforms inhibit activin signaling by competing with the wild-type receptor for binding to activin type II receptors, we hypothesized that overexpression of wild-type activin type I receptor will restore activin signaling. In HP75 cells, cotransfection of the wild-type activin type I receptor Alk4-1 expression vector increases activin-responsive reporter activity. Furthermore, transfection with wild-type activin receptor type I results in activin-mediated suppression of cell proliferation. These data indicate that truncated Alk4 isoforms interfere with activin signaling pathways and thereby may contribute to uncontrolled cell growth. Overexpression of the wild-type Alk4-1 receptor restores responsiveness to activin in human pituitary tumor-derived cells.
ACTIVIN IS A member of the TGFß superfamily of cytokines that regulates cell proliferation and differentiation (1, 2). Activin induces growth suppression in a number of normal cell types and in human cancer cell lines (1), including rat and mouse pituitary tumor cell lines (3, 4). Activin has been shown to have antiproliferative effects in only a small subset of human pituitary tumors (5), and activation of cell cycle inhibitors by functional TGFß signaling results in G1 growth arrest and inhibition of DNA synthesis (6, 7). We have shown that in these activin-responsive human pituitary tumors, treatment with activin in primary culture results in up-regulation of p21WAF1/cip1, a cyclin-dependent kinase inhibitor. However, the majority of pituitary tumors are not responsive to activin in vitro, and p21WAF1/cip1 expression remains at steady state levels in these tumors. We hypothesized that defects in activin receptor function in pituitary adenomas may be an underlying cellular mechanism modulating uncontrolled pituitary cell growth.
Activin signals through a heteromeric complex of specific type I and type II transmembrane Ser/Thr kinase receptors (8). Activin binds to the type II receptor, which, in turn, binds the type I activin receptor and activates it by trans-phosphorylation. The activated type I receptor signals through the Smad protein cascade to control gene expression and cellular response. The intracytoplasmic kinase subdomains of activin receptors are critical to signal transduction, and alterations in this region have been shown to inactivate the signaling pathway (7, 9, 10, 11). Human pituitary tumors specifically express alternatively spliced activin type I receptor Alk4 mRNAs, designated Alk4-2, 4-3, and 4-4, encoding truncated receptors that lack specific carboxyl kinase subdomains of the wild-type receptor Alk4-1 (12). Using the erythroleukemia K562 cell line, we have previously shown that these truncated Alk4 isoforms act as suppressors of the activin signaling pathway by forming inactive complexes with type II receptors (7). However, it is not known whether truncated receptors act as dominant negative isoforms to suppress activin signaling and the effect of activin on human pituitary tumor proliferation.
We therefore investigated activin effects on cell proliferation and the role of truncated activin receptor isoforms in activin signaling using a cell line derived from a human gonadotroph pituitary tumor (13). These cells have been shown to decrease cell proliferation in presence of TGFß (13), which suggests that the intracellular signaling pathway for TGFß is intact. We have found that tumor-specific Alk4 isoforms interfere with activin signaling and activin-induced growth suppression in these pituitary cells. Overexpression of Alk4-1 restores activin signaling and activin-induced growth arrest.
Materials and Methods
Cell line
Human pituitary HP75 cells, provided by Dr. Ricardo Lloyd (Mayo Clinic, Rochester, MN), were cultured in DMEM supplemented with 15% horse serum, 2.5% fetal bovine serum, and antibiotics at 37 C in humidified air with 5% CO2. This gonadotroph-derived cell line was obtained by transformation of clinically nonfunctioning tumor cells using a replication-defective recombinant human adenovirus expressing simian virus 40 large T antigen (13). HP75 cells retain differentiated functions, such as synthesis of FSH ß-, LH ß-, and 
-subunits and secretion of FSH in culture medium. Early passages of this cell line (passages 20–30) were used in our studies.
Tritiated thymidine incorporation
Quantification of DNA synthesis by [3H]thymidine incorporation as an indirect measurement of proliferation was performed as previously reported (5, 14). Ten thousand HP75 cells were aliquoted into six-well plates. After incubation for 48 h in DMEM with 10% fetal bovine serum, fresh medium with 10% fetal bovine serum, 10 and 50 ng/ml recombinant human activin A (National Hormone and Pituitary Program, Torrance, CA), and 0.1 and 0.2 ng/ml TGFß1, as a positive control, were added to corresponding wells for 24 h. We used these doses because previous studies have demonstrated the effectiveness of such treatment to inhibit cell proliferation in various cell lines and tumor cell primary cultures (5, 7, 13, 14). Then, 1 µCi/ml [3H]thymidine/well was added for 4 h. Tumor genomic DNA containing incorporated [3H]thymidine was precipitated onto glass-fiber scintillation filters using 10% trichloroacetic acid as previously described. Radioactivity was measured by scintillation counting of filters for 1 min in 10 ml Scintiverse BD liquid scintillation counting liquid (Fisher Scientific, Pittsburgh, PA).
ELISA of activin A in conditioned medium
Cells (106) were plated in 100-mm cell culture dishes with 10 ml culture medium. After 48 h, the culture medium was collected as conditioned medium (CM), and the concentration of activin A was measured using an activin A assay kit (Serotec, Oxford, UK) according to the manufacturer’s protocol. Serial dilutions of HP75 CM were included in each assay and were compared in parallel to activin A standards. To measure the linearity of activin production by this cell line, HP75 cells were plated in six-well plates at concentrations of 105, 2 x 105, and 2.5 x 105 cells/well. Conditioned medium was collected at 24, 48, and 120 h, and then activin A levels were measured by activin A assay kit.
RT-PCR
Total RNA was extracted from 5 x 105 HP75 cells, PDFS human folliculostellate cells (cells derived from a clinically nonfunctioning pituitary tumor) (15), or 100 mg normal pituitary tissue with TRIzol reagent (Life Technologies, Inc., Gaithersburg, MD), and RT-PCR was performed as previously described (5, 12, 15, 16). Control reactions were carried out in the absence of reverse transcriptase. All primers were described previously (12). The expected activin receptor PCR product sizes are: Alk4-1, 530 bp; Alk4-2, 453 bp; Alk4-3, 346 bp; and Alk4-4, 248 bp. All PCR reactions started with 5 min at 94 C, followed by 30 cycles (1 min at 94 C, 1 min at optimal annealing temperature, and 1 min and 15 sec at 72 C), with a final step of 10 min at 72 C as reported. A reaction with no template was included in each PCR set as a negative control. Products were resolved by electrophoresis on 1% agarose gels, stained with ethidium bromide, and visualized with UV light.
Purification of receptor complexes
Isolation of activin receptor type I/II complexes in HP75 cells was performed as previously described (7). Briefly, cells were grown in 60-mm tissue culture dishes and were cotransfected with a myc-tagged activin receptor type IIB construct and a FLAG-tagged Alk4-1 or Alk4-2 construct. After labeling with [35S]methionine (NEN Life Science Products, Boston, MA), cell lysates were prepared, and the activin receptor complexes were isolated by sequential immunoprecipitation using anti-FLAG and anti-myc antibodies. The precipitated receptors were resolved by SDS-PAGE.
Cell transfection and luciferase assay
Mink lung L17 cells were transiently transfected using Lipofectamine (Life Technologies, Inc.) as previously described (15) with Alk41-Flag-pCI, containing the FLAG-tagged, full-length activin receptor IB cDNA cloned into the mammalian expression vector pCI-Neo (Promega Corp., Madison, WI) and p3TPlux reporter, in which a luciferase gene is under the control of an activin-responsive element-containing promoter. As an internal control, pRSV-lacZ was included in each experiment. After 4-h incubation at 37 C, the culture medium was replaced with fresh medium for 20 h. The medium was then changed to contain 50% or 100% HP75 CM, 25 ng human recombinant activin A, or 50% HeLa CM. We cultured L17 transfected cells in medium plus vehicle alone as a control. After an additional 24-h incubation, cells were harvested, luciferase activity was measured, and results were normalized to ß-galactosidase activity. HP75 cells were similarly transiently transfected with Alk41-Flag-pCI or Flag-pCI vectors and p3TPlux. To quantitate luciferase activity, the internal control pRSV-lacZ was included in each experiment. Cells were harvested after 24-h incubation in fresh medium with or without 50 ng activin A. The activity of luciferase and ß-galactosidase were measured.
Bromdeoxiuridine (BrdU) incorporation assay
To study the effect of overexpression of wild-type activin receptor IB on cell proliferation in HP75 cells, a cDNA encoding FLAG-tagged Alk4-1 was inserted upstream of the encephalomyocarditis virus internal ribosomal entry site (IRES) of the bicistrionic expression vector IRES2-enhanced green fluorescence protein (EGFP) containing the enhanced variant of GFP (CLONTECH Laboratories, Inc., Palo Alto, CA), to generate Alk41-IRES2-EGFP. A plasmid containing the ß-galactosidase gene (lacZ-IRES2-EGFP) was generated as a control construct. HP75 cells were seeded (2 x 105 cells/well) in chamber slides in complete medium. After overnight incubation, 2 µg Alk41- or lacZ- IRES2-EGFP were transfected with 10 µl Lipofectamine for 4 h. After an additional 20-h incubation in 1% FCS/DMEM, cells were cultured with fresh medium with or without 25 ng/ml activin A for 24 h. To assess DNA synthesis in proliferating HP75 cells, 10 µmol/ml BrdU were added in each well for 1 h. Cells were then washed, fixed with 3% paraformaldehyde/2% glucose in PBS for 10 min, permeated with 2% Triton X-100, and blocked overnight in 3% BSA. Immunostaining for BrdU was performed as previously described. Briefly, anti-BrdU mouse antibody (BD Biosciences, Franklin Lakes, NJ), was diluted in PBS at a 1:10 concentration in the presence of 100 U/ml ribonuclease-free deoxyribonuclease (Roche, Indianapolis, IN) and 5 mM MgSO4. Cells were incubated at room temperature for 1 h, followed by three washes in PBS with 1% Triton X-100. Rhodamine-conjugated goat antimouse antibody (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) was used as a secondary antibody, diluted 1:200 in PBS with 1% Triton X-100, for 15 min at room temperature. After three more washes, 4',6-Diamidino-2-phenylindole dihydrochloride (1:500 in PBS) was used to stain nuclei, the slides were washed twice in PBS, and then coverslips were mounted with VectaShield (Vector Laboratories, Inc., Burlingame, CA). Using a Nikon microscope (Melville, NY) with UV light, 100 cells expressing EGFP were counted for each treatment, and rhodamine-positive cells were compared with rhodamine-negative 4',6-Diamidino-2-phenylindole dihydrochloride-stained cells.
Statistical analysis
Each experiment was repeated three times, and statistical significance was determined by t test.
Results
Effects of activin A on HP75 cell proliferation
We examined the activin-mediated antiproliferation effect in HP75 cells by [3H]thymidine incorporation. As shown in Fig. 1
, HP75 cells did not respond to activin treatment, although exposure to TGFß significantly decreased cell proliferation by 64% (Fig. 1), consistent with previously reported results (13).
|
Derived from a human gonadotroph tumor, HP75 cells express mRNA for inhibin - and ßA-subunits, but not ßB-subunit, as shown by RT-PCR (Fig. 2A). The positive controls used in this study were normal human pituitary tissue, which has been shown to express mRNA for all subunits of inhibin (16), and human folliculostellate PDFS cells, which express mRNA for inhibin - and ßA-subunits (15). Also, activin A dimers of ßA-subunits were detected in HP75 CM by ELISA. As shown in Fig. 2B, activin production was linear over time at different cell concentrations. HP75 CM was used to treat L17 cells transfected with the wild-type activin receptor IB and 3TPLux reporter constructs. As shown in Fig. 2C, the degree of 3TPLux reporter activation by HP75 CM was similar to that of L17 cells treated with 25 ng/ml recombinant activin A. In contrast, HeLa CM failed to induce transcription from 3TPLux in this system. Activin-responsive reporter activity did not increase when L17 cells transfected with Flag-pCI control vector were treated with HP75 CM or recombinant activin A.
|
To investigate the activin-specific signaling pathway in HP75 cells, we examined the mRNA expression of wild-type activin receptor type IB and its alternate splice variants by RT-PCR. As shown in Fig. 3A, HP75 cells express mRNA for all isoforms of Alk4, including the wild-type-1 and the tumor-specific truncated forms -2, -3, and -4. Expression of these receptors was also found in human pituitary tumor-derived folliculostellate PDFS cells, which were used as a positive control.
|
, arrow C) regardless of activin treatment. The full-length Alk4-1 found in the receptor complexes appeared on SDS-PAGE as a diffuse band (Fig. 3B, arrows A and A*), consistent with our previous findings that it contains both phosphorylated (A*) and nonphosphorylated (A) forms of the receptors (7). This phosphorylated Alk4-1 receptor was also found in the complexes from cells not treated with activin, indicating that endogenous activin is sufficient to induce phosphorylation of type I receptors. However, no corresponding slowly migrating bands were observed in complexes containing Alk4-2 despite the presence of endogenous activin (Fig. 3B, arrow B), indicating that Alk4-2 was not phosphorylated, although it formed complexes with the type II receptors. Restoration of activin signaling in HP75 cells
In HP75 cells transfected with 3TPlux reporter construct, no increase in luciferase activity was observed in response to activin A (Fig. 4A). To test whether overexpression of wild-type Alk4 can overcome the blocking effect of the truncated receptors, we transiently transfected Alk41 expression construct plus 3TPlux into HP75 cells. As shown in Fig. 4A, cotransfection of Alk41-Flag-pCI significantly increased luciferase activity. Addition of exogenous activin A further enhanced luciferase expression by approximately 5-fold compared with that in control cells transfected with the vector alone.
|
), indicating that the overexpression of the wild-type activin receptor restores the antiproliferative effect of activin in this human pituitary tumor cell line. Discussion
The intracellular signaling pathway shared by TGFß and activin is functional in HP75 cells, as shown by decreased cellular proliferation in response to TGFß peptide exposure. However, activin A fails to decrease cell proliferation in these cells, consistent with dysfunctional activin receptor signaling. We demonstrate that this gonadotroph tumor-derived HP75 cell line expresses tumor-specific, alternatively spliced activin receptors Alk4-2, -3, and -4, and therefore activin signaling and activin-induced growth arrest in this cell line are abolished by these dominant negative receptor isoforms. Overexpression of wild-type activin receptor Alk4-1 restores the responsiveness to activin in these human gonadotroph-derived tumor cells.
In the human pituitary, activin has been demonstrated to be a specific cell differentiation factor and to regulate hormone secretion (17, 18). Previously, we have shown that activin has an antiproliferative effect in a subgroup of clinically nonfunctioning human pituitary tumors (5). However, the majority of pituitary tumors did not decrease cell proliferation after activin exposure in vitro. Alternatively spliced activin receptor isoforms, Alk4-2, 4-3, and 4-4, are expressed in human pituitary tumors (12). Furthermore, the majority of clinically nonfunctioning and GH- and PRL-secreting adenomas express one or more alternatively spliced Alk4 receptors, usually at much higher levels than the full length wild-type receptor (7, 12). In contrast, normal pituitary tissue does not express detectable levels of Alk4-2 or Alk4-3. In a human erythroleukemia cell line model, these tumor-specific, truncated receptors function as dominant negative competitors, interfering with activin signaling mediated by the wild-type receptor (7). Our study is the first to demonstrate that truncated activin receptors function as inhibitors of activin signaling in human pituitary tumor cells. Human pituitary HP75 cells synthesize and secrete functional activin A and express all alternatively spliced Alk4 receptor isoforms, as is characteristic of human pituitary tumors. Specific activin A receptor signaling transduction is abolished in this cell line. However, overexpression of the wild-type activin type IB receptor restores the responsiveness to activin in these cells, as determined by both reporter gene expression and cell proliferation assessment. These data strongly suggest a role for truncated receptors in interfering with activin-mediated negative regulation of cell proliferation, thereby potentially contributing to human pituitary tumor growth.
In several cancer cell lines, activin has been shown to have suppressive effects and induce apoptosis (1, 19). Activin has a key role in controlling prostate cancer growth by opposing androgen-induced cell proliferation, and prostate tumor progression has been suggested to be caused by a decrease in the inhibitory effect of locally produced activin by down-regulated mRNA expression of either activin or activin receptor type IB (20). Activin is an important mediator of programmed cell death in the liver (21). Similar to hepatic growth factor, activin promotes the migration of hepatic neoplastic cells through its specific receptors on metastatic cells, suggesting a role in organ-specific metastasis in the liver (22). Recently, by studying the mutational patterns of TGFß superfamily receptors in pancreatic cancer through loss of heterozygosity in genomic DNA, it has been revealed that mutations of the activin type IB receptor are involved in multifactorial tumorigenesis in pancreatic cells (23). In addition, Smad4 (Dpc4/Madh4) is a critical mediator in the TGFß superfamily pathway, and it is phosphorylated by several pathway-specific Smad proteins (8). In pancreatic cells, the function of Smad4 is lost in the most advanced stages of neoplasia, and inactivating mutations in the TGFß superfamily/Smad system offer a selective advantage for proliferation (24, 25, 26). In this stepwise process that inactivates the tumor suppressor signaling pathway of Smad, activin receptor inactivation plays a determinant role in more advanced pancreatic cancers (23). This is the first direct evidence to support a role for the Alk4 gene as a tumor suppressor gene in human tumors and suggests a potential therapeutic role for stimulation of activin-induced proliferation control in the early neoplastic stages. Activin A signaling also inhibits tumor cell growth and induces apoptosis in the early stages of ovarian cancer (27) and in breast cancer (28) and has a local neurotrophic role in human benign pheochromocytomas (29). However, activin expression is lost in malignant pheochromocytomas (30). It has been recently demonstrated that thyroid papillary and follicular malignant cells are resistant to the antiproliferative effect of activin and TGFß in vitro. In these thyroid cancers a significant decrease in the expression of wild-type activin receptor Alk4 mRNA was observed compared with that in normal thyroid cells (31). Other factors, such as inhibins and follistatin, may play important roles in interfering with activin-induced growth arrest in these human tumors, and this requires further investigation.
Our data support the hypothesis that activin signaling, in particular, the wild-type activin type IB receptor and its truncated isoforms, may play a role in clonal cell proliferation in human pituitary tumors. Further studies are needed to clarify the mechanism of activin receptor isoform-specific activation of neoplastic cell growth.
Acknowledgments
Recombinant human activin A was kindly provided by Dr. A. F. Parlow (National Hormone and Pituitary Program, Torrance, CA).
Footnotes
This work was supported in part by NIH Grants R01-DK-40947 and F32-CA-88519-02, and the Jarislowsky Foundation.
Abbreviations: BrdU, Bromodeoxyuridine; CM, conditioned medium; EGFP, enhanced green fluorescence protein; PDFS, cells derived from a clinically nonfunctioning pituitary tumor.
Received April 3, 2002.
Accepted July 18, 2002.
References
This article has been cited by other articles:
 |
|
 |
|
  S. Mazerbourg, K. Sangkuhl, C.-W. Luo, S. Sudo, C. Klein, and A. J. W. Hsueh Identification of Receptors and Signaling Pathways for Orphan Bone Morphogenetic Protein/Growth Differentiation Factor Ligands Based on Genomic Analyses J. Biol. Chem., September 16, 2005; 280(37): 32122 - 32132. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
), then treated with 50% or 100% HP75 CM, 25 ng human recombinant activin A, or 50% HeLa CM for 24 h. Parental L17 cells transfected with Flag-pCI vector and 3TPLux were treated in parallel as negative controls (
). Luciferase activity was expressed as a percentage of the control value (mean ± SEM). *, P < 0.05 compared with control by t test. In addition, activin levels in conditioned medium were confirmed by ELISA.
