Autocrine human growth hormone inhibits placental transforming growth factor-beta gene transcription to prevent apoptosis and allow cell cycle progression of human mammary carcinoma cells.

Multiple cellular effects of human growth hormone (hGH) are mediated by an indirect mechanism requiring transcriptional activation of genes encoding protein effector molecules such as insulin-like growth factor-1. Such protein effector molecules then act directly to mediate the cellular functions of hGH. We report here that autocrine hGH production by mammary carcinoma cells specifically results in the transcriptional repression of the p53-regulated placental transforming growth factor-beta (PTGF-beta) gene. Transcriptional repression of the PTGF-beta gene does not require the p53-binding sites in the PTGF-beta promoter, and autocrine hGH also desensitized the response of the PTGF-beta promoter to p53 overexpression. Transcriptional repression of the PTGF-beta gene is accompanied by consequent decreases in its protein product, Smad-mediated transcription, and its cellular effects that include cell cycle arrest and apoptosis. PTGF-beta specifically inhibited the autocrine hGH-stimulated expression of cyclin D1 required for autocrine hGH-stimulated mammary carcinoma cell cycle progression. Thus, one mechanism by which autocrine hGH promotes an increase in mammary carcinoma cell number is by transcriptional repression of protein effector molecules that promote cell cycle arrest and apoptosis. Such transcriptional repression of negative regulatory factors, such as PTGF-beta, may also be requisite for direct stimulation of mammary carcinoma cell mitogenesis by hGH.

The human growth hormone (hGH) 1 gene is expressed in epithelial cells of the normal human mammary gland. 2 In-creased epithelial expression of the hGH gene is associated with the acquisition of pathological proliferation, and the highest level of hGH gene expression is observed in metastatic mammary carcinoma cells. 2 hGH receptor gene expression per mammary epithelial cell remains constant throughout the process of neoplastic progression (1), and therefore changes in the local concentration of ligand are likely to be pivotal to determine the effects of hGH on the behavior of the mammary epithelial cell. We have recently generated a model system to study the role of autocrine-produced hGH in mammary carcinoma by stable transfection of either the hGH gene or a translation-deficient hGH gene into mammary carcinoma (MCF-7) cells (2). The autocrine production of hGH by mammary carcinoma cells results in a hyperproliferative state with marked synergism observed between trophic agents such as IGF-1 (2). The increase in mammary carcinoma cell number as a consequence of autocrine production of hGH is a result of both increased mitogenesis and decreased apoptosis (3). Autocrine hGH production also results in enhancement of the rate of mammary carcinoma cell spreading on a collagen substrate (4), suggesting that it may affect cell motility and dissemination of the carcinoma. All of the studied effects of autocrine hGH on mammary carcinoma cell behavior are mediated via the hGH receptor (3). Thus, autocrine production of hGH by mammary carcinoma cells may direct mammary carcinoma cell behavior to impact on the final clinical prognosis. Systematic analysis of the relevant mechanistic features by which autocrine hGH exerts its cellular effects is therefore required.
One major mechanism by which GH affects cellular and somatic function is by regulating the level of specific mRNA species (5). Some of these GH-regulated genes code for trophic factors such as IGF-1 (6), which act in an intermediary role to execute the cellular effects of GH. Indeed, GH has been demonstrated to regulate the level of a number of trophic factors in specific tissues including hepatocyte growth factor in liver (7), epidermal growth factor in kidney (8), basic fibroblast growth factor in chondrocytes (9), interleukin-6 in osteoblasts (10), bone morphogenetic proteins 2 and 4 in fibroblasts (11), interleukin-1␣ and interleukin-1␤ in thymus (12), and preadipocyte factor-1 in adipocytes (13) and islet ␤-cells (14). It is therefore likely that many of the effects of autocrine hGH on mammary carcinoma cell function are also mediated by genetic regulation of specific trophic factors. Here we have used a cDNA microarray to identify autocrine hGH-regulated genes encoding polypeptide effector molecules that will act in an intermediary manner to mediate the effects of hGH on mammary carcinoma cell function.
We observed that autocrine hGH decreased transcription of the PTGF-␤ gene with consequent decreases in its protein product and accompanying cellular effects, which include cell cycle arrest and apoptosis. PTGF-␤ specifically inhibited the autocrine hGH-stimulated expression of cyclin D1 to prevent mammary carcinoma cell cycle progression. Thus, one mechanism by which autocrine hGH promotes mammary carcinoma cell survival is by transcriptional repression of protein effector molecules that promote cell cycle arrest and apoptosis. Such a mechanism is analogous and complementary to the ability of hGH to transcriptionally activate protein effector molecules, such as IGF-1, which stimulate processes resulting in increased cell number (15).

EXPERIMENTAL PROCEDURES
Materials-Effectene TM transfection reagent and the One-step RT-PCR kit were obtained from Qiagen GmbH (Hilden, Germany). Fetal bovine serum was purchased from HyClone Laboratories (Logan, UT), and N- [1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium salt transfection reagent was from Roche Diagnostics. All other tissue culture materials were obtained from Invitrogen. ECL detection reagents, Hybond TM -N nylon membranes, and the Oligolabeling kit were purchased from Amersham Biosciences. TRI-REAGENT was obtained from Molecular Research Center Inc. (Cincinnati, OH), and BrdUrd staining kit was from Zymed Laboratories Inc. (San Francisco, CA). The anti-␤-actin, anti-PARP, and anti-cyclin D1 antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA); anti-␤-catenin, anti-p27 Kip1 , and anti-p21 Waf/Cip antibodies were from Transduction Laboratories (Lexington, KY). A polyclonal antibody against PTGF-␤ was generated as described previously (16). The Atlas Pure Total RNA Labeling system, Atlas Human Cancer 1.2 Array, and ExpressHyb solution were obtained from CLONTECH Laboratories (Palo Alto, CA). Hoechst 33528, denatured salmon testis DNA, and 5Ј-bromo-2Ј-deoxyuridine were purchased from Sigma.
The Smad luciferase reporter plasmid containing a 4ϫ SBE promoter element was a generous gift of Robert A. Weinberg (Massachusetts Institute of Technology, Cambridge, MA). The cyclin D1 promoterluciferase construct (Ϫ1745CD1LUC) was a kind gift of Dr. Richard G. Pestell (AECOM, New York) (17).
The MCF-7 cell line was obtained from the ATCC and stably transfected with an expression plasmid containing the wild-type hGH gene (pMT-hGH) (18) under the control of the metallothionein 1a promoter (designated MCF-hGH) (2). For control purposes the ATG start site in pMT-hGH was disabled via a mutation to TTG generated by standard techniques (pMT-MUT) (18), and MCF-7 cells stably transfected with this plasmid were designated MCF-MUT (2). MCF-MUT cells therefore transcribe the hGH gene but do not translate the mRNA into protein. A detailed description of the characterization of these cell lines has been published previously (2). Neither MCF-7 nor MCF-MUT cells produce detectable amounts of hGH protein under serum-free conditions, whereas MCF-hGH cells secrete ϳ100 pM hGH into 2 ml of media over a 24-h period under the culture conditions described here. MCF-7 and MCF-MUT cells behave identically to each other in terms of proliferation, transcriptional activation (2), and cell spreading (4).
Preparation of Total RNA-Total RNA was isolated from MCF-MUT and MCF-hGH cells using TRI-REAGENT according to the manufacturer's instructions and resuspended in diethyl pyrocarbonate-treated water. Quantification and purity of the RNA was assessed by A 260 /A 280 absorption, and RNA quality was assessed by agarose gel electrophoresis. RNA samples with ratios greater than 1.6 were stored at Ϫ80°C for further analysis.
Analysis of Differential Gene Expression by Use of Atlas Human Cancer 1.2 Array-Poly(A) ϩ RNA was isolated from total RNA using the Atlas Pure Total RNA labeling system. Three independently derived total RNA samples from the respective cell lines were pooled before labeling for hybridization to the Atlas Human Cancer 1.2 Array.
Poly(A) ϩ RNA was reverse-transcribed in the presence of [␣-32 P]dATP for generation of radiolabeled cDNA. Probes were purified and hybridized to the Atlas Human Cancer 1.2 Array according to the manufacturer's instructions (overnight at 68°C). After a series of high stringency washes (three 20-min washes in 2ϫ saline/sodium citrate (SSC), 1% SDS followed by two 20-min washes in 0.1ϫ SSC, 0.5% SDS) at 68°C, the membranes were exposed to x-ray film and subjected to autoradiography. The relative levels of gene expression were quantified by densitometric scanning by use of the GS-700 imaging densitometer from Bio-Rad according to the manufacturer's instructions. Genes were considered differentially expressed when they exhibited a 2-fold or greater increase or decrease in the presence of autocrine hGH (MCF-hGH cells) compared with the absence of autocrine hGH (MCF-MUT cells) in three independently performed experiments. The relative expression of housekeeping genes (glyceraldehyde-3-phosphate dehydrogenase, ␤-actin, and ribosomal protein S9) served to normalize gene expression levels and did not differ by more than 10% between MCF-MUT and MCF-hGH cells.
Probe Labeling for Northern Blot Analysis-The PTGF-␤ cDNA fragment was derived from the PTGF-␤ expression plasmid by digestion with the restriction enzyme EcoRI and purification of the fragment on an agarose gel. The DNA was then labeled with [␣-32 P]dCTP (3000 Ci/mmol) using the Oligolabeling kit. In brief, 50 ng of DNA was denatured by heating for 2-3 min at 95-100°C and was directly transferred to ice for 2 min. 10 l of reagent mix (containing dATP, dGTP, and dTTP and random hexanucleotides) and 50 Ci of [␣-32 P]dCTP were added to the denatured DNA, and the volume was adjusted to 49 l with distilled water. 1 l of FPLCpure TM Klenow fragment (5-10 units) was added, and the reaction mixture was incubated at 37°C for 60 min. The labeled PTGF-␤ cDNA fragment was denatured by heating at 95-100°C for 2 min and cooled immediately on ice. The labeled PTGF-␤ cDNA fragment was used directly as a hybridization probe.
RNA Gel Electrophoresis and Northern Blot Analysis-800 ng of poly(A) ϩ RNA (mRNA) was fractionated by 0.7% formaldehyde-agarose gel electrophoresis. The mRNA in the gel was transferred to a Hybond TM -N nylon membrane by vacuum blotter (Bio-Rad) at the pressure of 5 inches of Hg in 10ϫ SSC (for 90 min). The nylon membrane was rinsed with 2ϫ SSC and allowed to air-dry. The RNA was immobilized onto the membrane by UV light cross-linking. The membrane was pre-hybridized in ExpressHyb with 0.1 mg/ml heat-denatured salmon testes DNA at 68°C for 30 min. Approximately 50 g of the PTGF-␤ cDNA fragment labeled to a specific activity of 1-2 ϫ 10 9 dpm/g was added, and the membrane was incubated at 68°C overnight. The membrane was washed three times with pre-warmed wash solution 1 (2ϫ SSC, 1% SDS) at 68°C for 30 min, followed by one washing step with pre-warmed wash solution 2 (0.1ϫ SSC, 0.5% SDS) at 68°C for 30 min. The membrane was then washed in 2ϫ SSC at room temperature for 5 min, and the radioactivity was detected by autoradiography. For reprobing to detect ␤-actin as loading control, the membrane was stripped by boiling in 0.5% SDS for 10 min and rinsed once with wash solution 1. The ␤-actin DNA fragment was labeled and hybridized to the stripped membrane and subjected to autoradiography as described above.
Densitometric Analysis of Band Intensities-The intensity of the respective bands on the blots was quantitated using a Bio-Rad GS-700 imaging densitometer and analyzed with the Multianalyst (version 1.0.1) program (Bio-Rad).
Reverse Transcriptase-PCR-RT-PCR was performed in a final volume of 50 l containing 0.2 g of mRNA template, 0.6 M primers, 2 l of enzyme mix, 400 M of each dNTP, 1ϫ reaction buffer, and 1ϫ Q-Solution by use of the Qiagen One-step RT-PCR kit. Briefly, RNA template was reverse-transcribed into cDNA for 30 min at 50°C; Hotstart TaqDNA polymerase was activated by heating for 15 min at 95°C; the denatured cDNA templates were amplified by the following cycles: 94°C/30 s, 55°C/30 s, and 72°C/60 s. A final extension was performed for 10 min at 72°C. In order to compare the PCR products semiquantitatively, 15-40 cycles of PCR (annealing temperature 55°C) were performed to determine the linearity of the PCR amplification, and the amplified ␤-actin cDNA served as an internal control for cDNA quantity and quality. All RNA samples were treated with DNase I to avoid genomic DNA contaminations.
Luciferase Reporter Assay for PTGF-␤ and Cyclin D1 Promoter Constructs-MCF cells were cultured to 80% confluence in 6-well plates. Transient transfection was performed in serum-free RPMI with Effectene TM according to the manufacturer's instructions. 0.2 g of the respective luciferase construct and 0.2 g of pCMV␤ were transfected per well in serum-free RPMI medium for 12 h before the medium was changed to fresh serum-free RPMI with or without 50 nM hGH, 10% FBS, 10 nM hIGF-1, or 10 nM 17-␤-estradiol. After a further 24 h, cells were washed with PBS and scraped into assay buffer (250 mM Tris-HCl, pH 8.0, 1 mM dithiothreitol). The protein content of the samples was normalized, and luciferase assays were performed as described previously (19). Results were normalized to the level of ␤-galactosidase activity to control for transfection efficiency.
Western Blot Analysis-MCF-MUT and MCF-hGH cells were grown to 80% confluence in 6-well plates. Transient transfection was performed in serum-free RPMI with Effectene TM according to the manufacturer's instructions. 0.2 g of PTGF-␤ expression plasmid or the control vector were transfected per well in serum-free RPMI medium for 12 h before the medium was changed to fresh serum-free RPMI. After 48 h cells were lysed at 4°C in 150 l of lysis buffer (50 mM Tris-HCl, pH 7.4, 1% Triton X-100, 150 mM NaCl, 1 mM EGTA, 1 mM EDTA, 0.2 mM sodium orthovanadate, 0.5% Nonidet P-40, 0.2% phenylmethylsulfonyl fluoride) for 30 min with regular vortices. Cell lysates were then centrifuged at 14,000 ϫ g for 15 min; the resulting supernatants were collected, and protein concentration was determined by the Lowry method, using bovine serum albumin as a standard. 5ϫ SDS sample buffer (50 mM Tris-HCl, pH 6.6, 5% SDS, 5% ␤-mercaptoethanol, and bromphenol blue) was added to 20 g of total protein, boiled for 5 min, and centrifuged at 14,000 ϫ g for 5 min. The supernatants were collected and subjected to 7.5% SDS-PAGE. Proteins were transferred to nitrocellulose membrane using a standard semidry electroblotting apparatus in Laemmli buffer containing 10% methanol.
For the analysis of the levels of secreted PTGF-␤ protein, MCF-MUT and MCF-hGH cells were grown in serum-free medium. Medium was collected, and 400 l were precipitated with acetone at Ϫ20°C for 2 h, centrifuged at 16,000 ϫ g for 30 min at 4°C, and the supernatant discarded. The pellet was washed once with 70% methanol and centrifuged at 16,000 ϫ g for 15 min at 4°C. The pellet was boiled in 5ϫ SDS sample buffer for 5 min and centrifuged at 14000 ϫ g for 5 min. The supernatants were collected and subjected to 7.5% SDS-PAGE. Proteins were transferred to nitrocellulose membranes using a standard semidry electroblotting apparatus in Laemmli buffer containing 10% methanol.
5Ј-Bromo-2Ј-deoxyuridine Incorporation Assay-Mitogenesis was directly assayed by measuring the incorporation of BrdUrd (3). For incorporation of BrdUrd, subconfluent MCF-MUT and MCF-hGH cells were washed twice with PBS and seeded to glass coverslips in either serum-free RPMI medium or serum-free medium supplemented with 50 nM hGH for 24 h. Both cell lines were pulse-labeled with 20 mM BrdUrd for 30 min, washed twice with PBS, and fixed in cold 70% ethanol for 30 min. BrdUrd detection was performed by use of the BrdUrd staining kit according to the manufacturer's instructions. A total population of over 3 times 300 cells was analyzed in several arbitrarily chosen microscopic fields to determine the BrdUrd labeling index (percentage of cells synthesizing DNA).
Cells were cultured for 24 h in serum-free medium and then transfected with 800 nM oligonucleotide for 8 h in N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium salt (10 l/ml medium). Cells were either lysed and processed as described above for Western blot analysis or processed for determination of nuclear BrdUrd incorporation (as described above).
Measurement of Apoptosis-Apoptotic cell death was measured by fluorescent microscopic analysis of cell DNA staining patterns with Hoechst 33258 (21). To prepare PGTF-␤ conditioned medium for determination of apoptosis, subconfluent MCF-7 cells were transfected with the PGTF-␤ expression plasmid or for control with the empty vector in serum-free media for 12 h and serum-starved for 24 h. The media were collected from three transfection experiments and spun down for 5 min at 600 ϫ g, and the supernatant was collected.
MCF-MUT and MCF-hGH cells were trypsinized with 0.5% trypsin and washed twice with serum-free RPMI medium. The cells were then seeded to glass cover slips in 6-well plates and incubated in serum-free RPMI medium and transfected with either a control vector or an expression vector containing PTGF-␤ cDNA. After a culture period of 24 h in serum-free RPMI medium, the cells were fixed for 20 min in 4% paraformaldehyde in PBS, pH 7.4, at room temperature. Alternatively, cells were cultured in the presence of conditioned media. The cells were then rinsed twice in PBS and stained with the karyophilic dye Hoechst 33258 (20 g/ml) for 5 min at room temperature. Following washing with PBS, nuclear morphology was examined under a UV-visible fluorescence microscope (Zeiss Axioplan). Apoptotic cells were distinguished from viable cells by their nuclear morphology characterized by nuclear condensation and fragmentation as well as the higher intensity of the blue fluorescence of the nuclei. For statistical analysis, three times 300 cells were counted in eight random microscopic fields at ϫ400 magnification.
Statistics-All experiments were repeated at least three to five times. All numerical data are expressed as mean Ϯ S.E., and the data were analyzed using Instat 3.0 from GraphPad Software Inc.

Identification of PTGF-␤ as a Gene Negatively Regulated by the Autocrine Production of hGH in Human Mammary
Carcinoma Cells-To identify genes regulated by autocrine production of hGH in mammary carcinoma cells, we screened a high density cDNA array with labeled cDNA derived from either MCF-7 cells stably transfected with the hGH gene but with the start codon mutated to TTG (MCF-MUT) or from MCF-7 cells stably transfected with the hGH gene (MCF-hGH). We have detailed previously (2, 4) the extensive characterization of these two cell lines. We used the CLONTECH Atlas Human Cancer 1.2 array and were particularly examining for factors, which inhibited proliferation or promoted cell death. One gene that we observed to be consistently decreased in MCF-hGH cells in comparison to MCF-MUT cells was macrophage inhib- itory cytokine (MIC-1) (Fig. 1) (22). MIC-1 has since been demonstrated to be homologous to placental transforming growth factor-␤ (23). Because the array used only a single spot for a specific gene, we will not present the data for other genes on the array that are potentially regulated by the autocrine production of hGH in mammary carcinoma cells. Other potentially regulated genes on the array will need to be verified by RT-PCR or Northern blot analysis before publication and are not in the scope of the present study. The relative expression of housekeeping genes (glyceraldehyde-3-phosphate dehydrogenase, ␤-actin, and ribosomal protein S9) served to normalize gene expression levels and did not usually differ by more than 10% between MCF-MUT and MCF-hGH cells. We therefore proceeded to characterize the observed decrease in PTGF-␤ mRNA in mammary carcinoma cells as a consequence of autocrine production of hGH.
Semi-quantitative RT-PCR and Northern Blot Analysis of the Effect of Autocrine hGH, Exogenous hGH, and FBS on the Level of PTGF-␤ mRNA in Human Mammary Carcinoma Cells-To verify the autocrine hGH-dependent down-regulation of PTGF-␤ mRNA observed by cDNA array screening in mammary carcinoma cells, we first examined the level of PTGF-␤ mRNA in MCF-MUT and MCF-hGH cells by semi-quantitative RT-PCR. The validation that the conditions of RT-PCR used here yielded semi-quantitative estimates of mRNA is described under "Experimental Procedures" and has been utililized previously (24). One amplified fragment of the predicted size (368 bp) appropriate for PTGF-␤ mRNA was detected in MCF-MUT and MCF-hGH cells ( Fig. 2A). Autocrine production of hGH by MCF-hGH cells resulted in a decreased level of PTGF-␤ mRNA in comparison to MCF-MUT cells. The level of ␤-actin mRNA did not differ between the two cell lines and was used as a control for RNA quality ( Fig. 2A).
To verify further that autocrine hGH production in mammary carcinoma cells resulted in a decreased level of PTGF-␤ mRNA, we resorted to Northern blot analysis. PTGF-␤ mRNA of the appropriate size (1.  (Fig. 2A).
Effect of Autocrine hGH, Exogenous hGH, and FBS on Luciferase Activity from a Reporter Plasmid Containing the Promoter Region of the PTGF-␤ Gene-To determine whether the autocrine hGH stimulated decrease in PTGF-␤ mRNA was due to down-regulation of PTGF-␤ gene transcription, we utilized a luciferase reporter plasmid containing the promoter region (923 bp) of the PTGF-␤ gene (16). Autocrine hGH production by MCF-hGH cells decreased luciferase expression 3-4-fold to that observed in MCF-MUT cells (Fig. 3A). Exogenous hGH (50 nM) did not affect PTGF-␤ gene transcription by MCF-MUT cells nor did it further repress transcription of the PTGF-␤ gene in MCF-hGH cells. 10% FBS resulted in a slight enhancement of PTGF-␤ gene transcription in MCF-MUT cells, and this enhancement of PTGF-␤ gene transcription by 10% FBS was similarly abrogated by autocrine production of hGH in MCF-hGH cells.
To exclude definitively the possibility that the differences in PTGF-␤ gene transcription observed between MCF-MUT and MCF-hGH cell lines were due to clonal selection artifact, we examined the response of PTGF-␤ gene transcription in parental MCF-7 cells to transient transfection of the hGH gene. For control purposes the ATG start site in pMT-hGH was disabled via a mutation to TTG, and this construct served as the control vector (as for MCF-MUT cells; pMT-MUT). As observed in Exogenous hGH (either 50 or 100 nM) did not affect PTGF-␤ gene transcription in MCF-7 cells as was observed for exogenous hGH applied to MCF-MUT cells. Thus PTGF-␤ is one gene selectively regulated by the autocrine production of hGH and is not co-regulated by exogenous hGH.
Interaction of Autocrine Production of hGH by Mammary Carcinoma Cells with hIGF-1 and 17-␤-Estradiol (E 2 ) on the Level of PTGF-␤ Gene Transcription-We have demonstrated previously (2) that autocrine hGH production by mammary carcinoma cells synergizes with exogenous administration of hIGF-1 but not E 2 to increase cell number. We therefore first examined the effect of 10 nM exogenously added hIGF-1 on PTGF-␤ gene transcription in MCF-MUT and MCF-hGH cells. Treatment of MCF-MUT cells with hIGF-1 increased luciferase activity from the reporter construct containing the promoter region of the PTGF-␤ gene, indicating that hIGF-1 increased transcription of the PTGF-␤ gene (Fig. 4A). However, hIGF-1 did not significantly release the inhibition of PTGF-␤ gene transcription due to autocrine production of hGH in MCF-hGH cells. Thus, autocrine hGH exerted the dominant effect on PTGF-␤ gene transcription. Treatment of both MCF-MUT and MCF-hGH cells with E 2 did not consistently alter transcription of the PTGF-␤ gene although there was a slight tendency to increase reporter activity in some experiments (Fig. 4B).
Autocrine hGH Inhibition of PTGF-␤ Gene Transcription Is p53-independent-PTGF-␤ has been identified as a p53 target gene (16). We therefore examined whether autocrine hGH inhibition of PTGF-␤ gene transcription involved a p53-dependent component. p53 has been demonstrated previously to bind to two p53-binding sites present in the promoter region of the PTGF-␤ gene utilized here (16). These are located at Ϫ898 to Ϫ879 and Ϫ43 to Ϫ24 upstream from the putative translation initiation site (16). Deletion of both p53-binding sites in the PTGF-␤ gene promoter dramatically reduced transcription of the PTGF-␤ gene (observe the relative comparative values between luciferase activity generated from the wild-type promoter and the p53-binding site deleted promoter in Fig. 3A and Fig. 5A. Despite the dramatic reduction in reporter activity upon deletion of the p53-binding sites in the PTGF-␤ gene promoter, the autocrine production of hGH by MCF-hGH cells maintained the equivalent inhibition of PTGF-␤ gene transcription (Fig. 5A). Autocrine production of hGH by MCF-hGH cells also abrogated the enhanced response of the intact PTGF-␤ promoter to overexpression of wild-type p53 (Fig. 5A). Furthermore, the inhibitory effect of autocrine production of hGH by MCF-hGH cells on the PTGF-␤ promoter was maintained when several PTGF-␤ gene promoter inactive (p53-mu143A and p53-mu273H) or active (p53-mu281G) mutants of p53 (16) were utilized (Fig. 5B). Thus autocrine hGH inhibition of PTGF-␤ gene transcription in mammary carcinoma cells does not involve p53.
Effect of Autocrine hGH on the Level of PTGF-␤ Protein in Mammary Carcinoma Cells-PTGF-␤ is a secreted protein and is secreted to the media in both a premature 40-kDa form and . Luciferase assays were performed as described under "Experimental Procedures." Results are presented as the relative luciferase activity normalized to constitutive ␤-galactosidase expression and are given as means Ϯ S.E. of triplicate determinations. *, p value of Ͻ0.05; **, p value of Ͻ0.01; ***, p value of Ͻ0.001 based on ANOVA, followed by a Bonferroni multiple comparison test. a mature 15-kDa form (22). To determine whether the autocrine hGH inhibition of PTGF-␤ gene transcription also resulted in decreased PTGF-␤ protein, we examined the level of PTGF-␤ in media collected from MCF-MUT and MCF-hGH cells by Western blot analysis. Both the premature PTGF-␤ at 40 kDa and the mature PTGF-␤ at 15 kDa could be detected by Western blot analysis in media from MCF-MUT and MCF-hGH cells (Fig. 6A). However, MCF-hGH cells secreted significantly less of both the premature and mature forms of PTGF-␤ to the media. Thus autocrine hGH production by mammary carcinoma cells also significantly decreases PTGF-␤ production.
Effect of Autocrine hGH on Smad-mediated Transcriptional Activation in Mammary Carcinoma Cells-It has been demonstrated previously (16,25) that PTGF-␤ functions through the TGF-␤ receptor to activate Smad-mediated transcription. Because autocrine production of hGH by MCF-hGH cells decreased the level of PTGF-␤, it could be expected that MCF-hGH cells would also exhibit a decrease in Smad-mediated transcription in comparison to MCF-MUT cells. We therefore used an artificial luciferase reporter construct containing four tandem repeats of the Smad3/4-binding consensus sequence (SBE, GTCTAGAC) (26) to examine the level of Smad-mediated transcription in MCF-MUT compared with MCF-hGH cells. Smad-mediated transcriptional activity was present in MCF-MUT cells cultured in serum-free media, and autocrine production of hGH by MCF-hGH cells significantly reduced the level of Smad-mediated transcription (Fig. 6B). (2) that autocrine production of hGH by mammary carcinoma cells increased cell proliferation. The D family of cyclins is pivotal to initiate progression through the G 1 phase of the cell cycle (27). Cyclin D1 is the predominant member of this family expressed in mammary gland (28) and is primarily regulated at the tran-

FIG. 5. Autocrine hGH inhibition of PTGF-␤ gene transcription is p53-independent.
A, MCF-MUT and MCF-hGH cells in serum-free media or serum-free media supplemented with 50 nM hGH or 10% FBS or both were transiently transfected with a luciferase reporter construct in which the PTGF-␤ promoter was lacking the p53-binding sites (PTGF-␤-without p53BS) and 1 g of pCMV␤. B, MCF-MUT and MCF-hGH cells in serum-free media were transiently transfected with the luciferase reporter construct containing the PTGF-␤ promoter (PTGF-␤-w/p53BS) and pCMV␤ and expression plasmids for p53-wildtype or several PTGF-␤ gene promoter inactive (p53-mu143A and p53-mu273H) or active (p53-mu281G) mutants of p53. Luciferase assays were performed as described under "Experimental Procedures." Results are presented as the relative luciferase activity normalized to constitutive ␤-galactosidase expression and are given as means Ϯ S.E. of triplicate determinations. *, p value of Ͻ0.05; **, p value of Ͻ0.01; and ***, p value of Ͻ0.001 based on ANOVA, followed by a Bonferroni multiple comparison test. Cells were cultured to confluency and transiently transfected with the 4ϫ SBE reporter plasmid, and luciferase assays were performed as described under "Experimental Procedures." Results are presented as the relative luciferase activity normalized to constitutive ␤-galactosidase expression and are given as means Ϯ S.E. of triplicate determinations, unpaired t test, p Ͻ 0.01 scriptional level by changes in promoter activity (53). We therefore first examined the effect of autocrine production of hGH by mammary carcinoma cells on the activity of the cyclin D1 promoter (17). Autocrine production of hGH in MCF-hGH cells resulted in a tripling of the promoter activity compared with MCF-MUT cells indicative of autocrine hGH-stimulated transcription of the cyclin D1 gene (Fig. 7A). Forced expression of PTGF-␤ in MCF-hGH cells completely prevented the autocrine hGH-stimulated increase in cyclin D1 promoter activity (Fig.  7A). We subsequently examined the level of cyclin D1 protein in MCF-MUT compared with MCF-hGH cells. The level of cyclin D1 protein in MCF-hGH cells was also dramatically increased in MCF-hGH compared with MCF-MUT cells. Forced expression of PTGF-␤ in MCF-hGH cells completely prevented the autocrine hGH-stimulated increase in the level of cyclin D1 protein (Fig. 7B). PTGF-␤ treatment of mammary carcinoma cells has been demonstrated previously (25) to result in G 1 arrest associated with an increase in p21 Waf1/Cip1 levels. We consequently also examined p21 Waf1/Cip1 levels in MCF-MUT and MCF-hGH cells Ϯ forced expression of PTGF-␤. Interestingly, MCF-hGH cells exhibited a marked increase in p21 Waf1/Cip1 protein in comparison to MCF-MUT cells (Fig. 7B). Equal loading of the cell extracts was verified by reprobing the stripped membrane for ␤-actin (Fig. 7B). Forced expression of PTGF-␤ in MCF-MUT cells resulted in an increase in p21 Waf1/Cip1 protein as could be expected (25), but no effect of forced expression of PTGF-␤ on the level of p21 Waf1/Cip1 protein in MCF-hGH cells was observed (Fig. 7B). We next examined the protein level of the cdk inhibitor p27 Kip1 . Decreased expression of p27 Kip1 is associated with G 1 /S phase transition (29). The level of p27 Kip1 was decreased in MCF-hGH cells compared with MCF-MUT cells concordant with increased cell cycle progression of MCF-hGH cells in comparison to MCF-MUT cells (see below). Forced expression of PTGF-␤ did not significantly alter the expression level of p27 Kip1 in MCF-hGH cells (Fig. 7A). To determine the effects of forced expression of PTGF-␤ on pro- gression of the cell cycle in MCF-MUT and MCF-hGH cells, we examined the nuclear incorporation of 5Ј-bromo-2Ј-deoxyuridine in these cells. MCF-hGH cells exhibited a significantly higher percentage of nuclear BrdUrd incorporation compared with MCF-MUT cells (Fig. 7C). Forced expression of PTGF-␤ in either cell line completely prevented nuclear BrdUrd incorporation (Fig. 7C). Forced expression of PTGF-␤ is likely to result in pharmacological levels of PTGF-␤ that would not necessarily represent its physiological role. We therefore overexpressed PTGF-␤ cDNA in MCF-7 cells, collected the conditioned media, diluted the conditioned media until PTGF-␤ was present at the same concentration as MCF-MUT cells (as estimated by Western blot analysis), and then used this physiological concentration of PTGF-␤ to examine the effects of PTGF-␤ on MCF-hGH cell cycle progression. As observed in Fig. 7D, use of this concentration of PTGF-␤ dramatically reduced cell cycle progression in MCF-hGH cells. Therefore, autocrine hGH-stimulated cell cycle progression in mammary carcinoma cells cannot proceed in the presence of an amount of PTGF-␤ sufficient to result in cell cycle arrest.
Cyclin D1 Is Required for Autocrine hGH-stimulated Mammary Carcinoma Cell Cycle Progression-Autocrine hGH production in mammary carcinoma cells resulted in increased cyclin D1 and cell cycle progression, both of which were inhibited by forced expression of PTGF-␤. To determine whether cyclin D1 was required for mammary carcinoma cell cycle progression stimulated by autocrine hGH, we reduced cellular cyclin D1 by use of antisense oligonucleotides (30) as observed in Fig. 8. Transfection of antisense oligonucleotides to cyclin D1 reduced the level of cyclin D1 in MCF-hGH cells to that observed in MCF-MUT cells, whereas sense control oligonucleotides did not affect the level of cyclin D1 (Fig. 8A). Similarly, transfection of cyclin D1 antisense oligonucleotides in MCF-hGH also reduced the percentage of cells with nuclear BrdUrd incorporation to that observed in MCF-MUT cells (Fig. 8B). Thus, an autocrine hGH-stimulated increase in cyclin D1 is required for autocrine hGH-stimulated cell cycle progression. Therefore, PTGF-␤ inhibition of autocrine hGH-stimulated cyclin D1 expression in mammary carcinoma cells is sufficient to prevent autocrine hGH-stimulated cell cycle progression.
Effect of PTGF-␤ on MCF-MUT and MCF-hGH Cell Survival-We have demonstrated previously (3,24) that autocrine production of hGH in MCF-hGH cells affords dramatic protection from apoptotic cell death in comparison to MCF-MUT cells. In contrast, exogenous hGH only marginally reduces the level of apoptosis of MCF-MUT cells in serum-free media (3). It can therefore be suggested that down-regulation of PTGF-␤ by autocrine hGH in mammary carcinoma cells is responsible for decreased apoptotic cell death in MCF-hGH cells compared with MCF-MUT cells. We first examined the effect of transient overexpression of PTGF-␤ on the formation of the 60-kDa ␤-catenin cleavage product which is associated with caspase activation and subsequent apoptosis (31,32). A 60-kDa cleavage product for ␤-catenin could be detected in both MCF-MUT and MCF-hGH cells (Fig. 9A). Forced expression of PTGF-␤ in both MCF-MUT and MCF-hGH resulted in a dramatic increase in the level of the 60-kDa ␤-catenin cleavage product. We also examined the effect of forced expression of PTGF-␤ on the formation of a caspase 3-dependent 85-kDa cleavage product from PARP (33). Western blot analysis for PARP expression and cleavage demonstrated a dramatically increased expression of PARP in MCF-hGH compared with MCF-MUT cells (Fig. 9A). Forced expression of PTGF-␤ in both MCF-MUT and MCF-hGH cells did not result in a detectable cleavage product. This is concordant with the observation that PARP cleavage occurs as a result of caspase 3 activation (32). MCF-7 cells are caspase 3-deficient (34) and are subject to caspase 3-independent apoptosis (35). We were also not able to detect caspase 3 activity nor PTGF-␤-stimulated caspase 3 activity in MCF-MUT and MCF-hGH cells with a specific caspase 3 activity assay (data not shown). Concordant with the increased level of the 60-kDa ␤-catenin cleavage product upon forced expression of PTGF-␤, forced expression of PTGF-␤ also dramatically increased apoptotic cell death in both MCF-MUT and MCF-hGH cells, with the apoptotic rate of MCF-hGH cells transfected with PTGF-␤ cDNA similar to that observed in MCF-MUT cells in serum-free media. PTGF-␤ conditioned media containing PTGF-␤ at the level of MCF-MUT cells derived as described above also dramatically increased the percentage of apoptotic cells in both MCF-MUT and MCF-hGH cell lines, with the level of apoptosis in MCF-hGH cells once again approximating that in MCF-MUT cells (Fig. 9C). Thus, PTGF-␤ promotes apoptotic cell death in mammary carcinoma cells, and down-regulation of PTGF-␤ gene transcription by hGH is one mechanism by which autocrine hGH prevents apoptotic cell death. DISCUSSION We have demonstrated here that autocrine production of hGH by mammary carcinoma cells results in a specific decrease of PTGF-␤ gene transcription. Thus, autocrine hGH function in the mammary epithelial cell, such as mitogenesis and cell survival, may be partly mediated by, or first require, repression of the inhibitory effects of PTGF-␤. Furthermore, because PTGF-␤ is a secreted protein, it will also act in a paracrine fashion to affect the function of neighboring cells. In carcinoma of the mammary gland, such paracrine interactions usually occur between carcinoma cells of epithelial origin and neighboring stromal fibroblasts and are pivotal to the development of carcinoma (36). It is noteworthy to mention that the hGH receptor is predominantly expressed on the normal or neoplastic epithelial cells of the human mammary gland, and therefore autocrine hGH production by mammary epithelial cells will, via the paracrine action of PTGF-␤, affect stromal cell function.
For example, fibroadenoma of the mammary gland, where we observe increased expression of the hGH gene in epithelial cells, 2 exhibits an intense stromal reaction. It is possible that loss of paracrine PTGF-␤ secretion by the mammary epithelial cell in response to autocrine hGH production may contribute to changes in stromal architecture. TGF-␤ secreted by mammary carcinoma cells has been demonstrated to affect stromal architecture and tumor progression (36). Use of in vivo models will allow us to define the effect of autocrine production of hGH by mammary epithelial or mammary carcinoma cells on the surrounding stroma and delineate the relative contribution to neoplastic progression.
PTGF-␤ is a recently identified secretory protein that shares ϳ25% sequence identity with TGF-␤ family members (23) and possesses several characteristics of the TGF-␤ superfamily including a signal peptide, a consensus RXXR(A/S) cleavage signal for processing to the mature form, and seven conserved cysteine residues in the carboxyl terminus (mature form) (22,23,(37)(38)(39). PTGF-␤ has also been demonstrated to utilize either the type I TGF-␤ receptor or type II TGF-␤ receptor to mediate cell cycle arrest (16). Activation of type I TGF-␤ receptor or type II TGF-␤ receptor results in a complex series of downstream signaling events resulting in phosphorylation of Smad proteins that translocate to the nucleus, associate with transcriptional co-activators, and transactivate TGF-␤-regulated genes (40). We also observed here that autocrine production of hGH by mammary carcinoma cells results in decreased Smad-mediated gene transcription in accord with the decreased production of PTGF-␤. Other mechanisms may also exist for the observed autocrine hGH-stimulated decrease in Smad-mediated transcription. By use of cDNA microarray technology we have reported recently (24) that the ski oncogene is ϳ4-fold up-regulated in response to autocrine production of hGH. It has been demonstrated recently that ski associates with both Smad2 and Smad3 resulting in repression of TGF-␤-responsive promoters via the Smad-binding element (SBE) used here (41). The TGF-␤ pathway usually functions to suppress cellular proliferation and cellular transformation. It has been proposed that the repression of TGF-␤-inducible genes (which function as negative regulators of cell cycle function in mammary epithelial cells) may be pivotal to the cellular transforming ability of ski (41). Thus, autocrine hGH production by mammary carcinoma cells acts in concert to both induce negative regulators and suppress positive regulators of the TGF-␤ pathway. That autocrine production of hGH by mammary carcinoma cells results in such coordinated repression of the TGF-␤ axis is suggestive that many functions of autocrine hGH in mammary epithelial cells may be achieved by simple antagonism of this pathway.
PTGF-␤ was identified as a p53-regulated gene (16,25). We have demonstrated here that autocrine production of hGH by mammary carcinoma cells decreased transcription of the PTGF-␤ gene in a p53-independent manner. Thus, we observe that autocrine production of hGH results in a similar percentage decrease in PTGF-␤ gene transcription in the absence of the two putative p53-binding sites in the PTGF-␤ promoter, and the inhibitory effect of autocrine production of hGH by MCF-hGH cells on the PTGF-␤ promoter was maintained when several PTGF-␤ gene promoter inactive (p53-(1-143) and p53-(1-273)) or active (p53-(1-281)) mutants of p53 (16,42) were utilized. Thus autocrine production of hGH by mammary epithelial cells will antagonize the cellular response to p53 and therefore promote inappropriate cell survival potentially leading to neoplastic transformation. p53 is an important mediator of the cellular response to DNA damage and activates genes responsible for both cell cycle arrest and apoptosis (43). How- FIG. 9. Effect of forced expression of PTGF-␤ on autocrine hGH-stimulated cell survival in mammary carcinoma cells. A, Western blot analysis to detect the effect of forced expression of PTGF-␤ on the generation of a 60-kDa ␤-catenin cleavage product and PARP expression in MCF-MUT and MCF-hGH cells. MCF-MUT and MCF-hGH cells were transfected in serum-free media with either control vector or an expression vector containing PTGF-␤ cDNA. Cell extracts were prepared and subjected to SDS-PAGE, and Western blot analysis was performed as described under "Experimental Procedures." B, the effect of forced expression of PTGF-␤ on apoptotic cell death in MCF-MUT and MCF-hGH cells in serum-free media. MCF-MUT and MCF-hGH cells were transfected in serum-free media with either control vector or an expression vector containing PTGF-␤ cDNA and processed for determination of apoptotic cell death as described under "Experimental Procedures." C, the effect of PTGF-␤-conditioned media on apoptotic cell death in MCF-MUT and MCF-hGH cells in serum-free media. PTGF-␤ cDNA was transfected in MCF-7 cells, and conditioned media were collected and diluted until PTGF-␤ was present at the same concentration as from MCF-MUT cells. Control or PTGF-␤-conditioned media were applied to MCF-MUT and MCF-hGH cells for 24 h and processed for determination of apoptotic cell death as described under "Experimental Procedures." Results for (B and C) represent means Ϯ S.E. of triplicate determinations of the percentage of apoptotic cells based on ANOVA, followed by a Bonferroni multiple comparison test. ever, inhibition of p53-regulated genes by autocrine production of hGH is not a general cellular response as autocrine hGH has been demonstrated previously by us to up-regulate two p53 regulated genes, namely GADD45 (24) and p21 waf1/cip1 . 3 Furthermore, IGF-BP3 is both a p53- (44) and hGH-regulated (45) gene. Interestingly, there exists a STAT-binding site in the promoter region of the PTGF-␤ gene (25), and this may constitute one mechanism for the observed effect of autocrine hGH on PTGF-␤ gene transcription. STAT molecules can either stimulate or repress transcription depending on the promoter context (46,47), and hGH has been demonstrated to utilize STATs for many of its transcriptional effects (48). Further analysis of the PTGF-␤ promoter should allow for the precise definition of the regulatory elements utilized by autocrine hGH to suppress PTGF-␤ gene transcription.
We have reported previously that autocrine production of hGH by mammary carcinoma cells results in increased entry to S-phase (3) and increased cell number (2). We observe here that forced expression of PTGF-␤ completely prevents S-phase entry by either MCF-MUT or MCF-hGH cells. Thus, decreased expression of PTGF-␤ would be required for autocrine hGH to promote cell cycle progression of mammary epithelial cells. However, exogenous application of hGH to mammary carcinoma cells (MCF-7) still results in equivalent entry to S-phase (3) but without a decrease in PTGF-␤ expression (this study). It may be that high expression of PTGF-␤ is incompatible with neoplastic proliferation, and therefore MCF-7 cells already have diminished PTGF-␤ expression below a critical threshold required for survival and cell cycle progression. This would therefore allow proliferation in response to a mitogen such as exogenously applied hGH without further decreases in the level of PTGF-␤. This is concordant with the fact that forced expression of PTGF-␤ alone results in cell cycle arrest and apoptosis of mammary carcinoma cells (this study and Ref. 25), including mammary carcinoma cells with autocrine production of hGH. It is interesting to note that autocrine hGH production by mammary carcinoma cells results in dramatically increased p21 Waf1/Cip1 expression. Increased p21 Waf1/Cip1 expression has been associated previously with inhibition of cell cycle progression stimulated by either PTGF-␤ (16) or TGF-␤ (49), and we also observe here that forced expression of PTGF-␤ in MCF-MUT cells results in increased p21 Waf1/Cip1 . This observation is concordant with the published role of p21 Waf1/Cip1 as the major mediator of p53 induced G 1 arrest (50). However, other investigators (51) have also demonstrated recently that autocrine GH results in specific up-regulation of p21 Waf1/Cip1 in Ba/F3 cells associated with cell proliferation. We also report here decreased p27 Kip1 expression in MCF-hGH compared with MCF-MUT cells. Interestingly, Raf-mediated cell proliferation is associated with elevated p21 Waf1/Cip1 , which specifically binds to and activates Cdk4-cyclin D complexes and also with decreased p27 Kip1 expression (52). Autocrine hGH production by mammary carcinoma cells results in increased raf gene expression (24), and this may be the mechanism for the observed changes in p21 Waf1/Cip1 and p27 Kip1 expression. In any case, cell cycle arrest as a consequence of forced expression of PTGF-␤ in MCF-hGH cells is not due to increased expression of p21 Waf1/Cip1 . We did observe that autocrine hGH stimulation of mammary carcinoma cells resulted in a dramatic increase in the level of cyclin D1 concordant with the observed increased entry of MCF-hGH cells to S-phase compared with MCF-MUT cells. The level of cyclin D1 is elevated in a high percentage of carcinomas of the mammary gland (53), and cyclin D1 overexpression in transgenic mice results in formation of mammary carcinoma (54). Increased expression of cyclin D1 has also been demonstrated to be sufficient for G 1 progression in mammary carcinoma cells (28). Forced expression of PTGF-␤ prevented the autocrine hGH-stimulated increase in cyclin D1 observed in MCF-hGH cells. The mechanism by which PTGF-␤ prevents the autocrine hGH-stimulated increase in cyclin D1 remains to be determined. The level of cyclin D1 is predominantly regulated at the transcriptional level by rapid changes in the activity of the cyclin D1 promoter which is under complex control by multiple signaling pathways (55). We observed here that autocrine hGH production in mammary carcinoma cells results in increased transcription of the cyclin D1 gene which is specifically repressed by PTGF-␤. Although p44/42 MAP kinase has been demonstrated to be required for cyclin D1 gene expression (56), we have observed that both autocrine hGH and PTGF-␤ resulted in increased activation of p44/42 MAP kinase in mammary carcinoma cells, 4 and therefore PTGF-␤ inhibition of cyclin D1 expression is not p44/42 MAP kinase-dependent. The transcription of the cyclin D1 gene is also regulated by STAT5 (57), and STAT5 is required for GH-stimulated mitogenesis of islet ␤-cells (58). We have observed that PTGF-␤ decreases STAT5-mediated transcription in mammary carcinoma cells, 3 and whether this constitutes the mechanism of the observed PTGF-␤ decrease in cyclin D1 is under investigation. In any case PTGF-␤ prevents autocrine hGH-stimulated mammary carcinoma cell cycle progression by inhibition of autocrine hGH-stimulated cyclin D1 expression.
In addition to blocking cell cycle progression of mammary carcinoma cells in response to autocrine production of hGH, we observed that forced expression of PTGF-␤ resulted in apoptotic cell death. Overexpression of PTGF-␤ has also been demonstrated by other investigators to result in apoptotic cell death of mammary carcinoma cells (25). In this regard it is interesting that the decreased transcription of the PTGF-␤ gene is only observed with autocrine-produced and not exogenously added hGH. Similarly, protection from apoptotic cell death is also only provided by autocrine-produced and not exogenously added hGH (3). Thus the suppression of PTGF-␤ gene transcription by autocrine hGH may be one mechanism by which autocrine hGH differentially affects mammary carcinoma cell behavior in contrast to exogenously added or "endocrine" hGH. Presumably the decreased production of PTGF-␤ will also result in the diminished transcription of pro-apoptotic genes and release of transcriptional repression of genes required for cell survival. Cell survival and cell death genes regulated by PTGF-␤ in mammary carcinoma cells remain to be identified. hGH may also directly regulate genes required for cell survival, and we have demonstrated recently that autocrine hGH increases transcription of the CHOP gene to result in survival of mammary carcinoma cells in a p38 MAP kinasedependent manner (24). What needs to be determined is the mechanism and sequential order by which these genes are regulated by autocrine production of hGH. Detailed and sequential promoter analyses, identification of the relevant transcription factor response elements combined with the relevant dissection of upstream signaling pathways, should allow for the identification of primary versus secondary or tertiary events in the effects of autocrine hGH on mammary carcinoma cell behavior and in particular apoptotic cell death.
In conclusion, we have demonstrated that autocrine production of hGH by mammary carcinoma cells results in transcriptional repression of the PTGF-␤ gene with consequent de-creases in its protein product and accompanying cellular effects that include cell cycle arrest and apoptosis. Thus, one mechanism by which autocrine hGH promotes mammary carcinoma cell survival is by transcriptional repression of protein effector molecules that promote cell cycle arrest and apoptosis. Such a mechanism is analogous and complementary to the ability of hGH to activate transcriptionally the protein effector molecules that stimulate cell cycle progression and cell survival (15). It remains to be determined what effects of autocrine hGH on mammary carcinoma cell behavior are mediated directly by autocrine hGH or indirectly via utilization of effector molecules, which themselves directly affect cellular function.