Regulation of gene expression by PRA-910, a novel progesterone receptor modulator, in T47D cells

Steroids 68 (2003) 995–1003

Regulation of gene expression by PRA-910, a novel progesterone receptor modulator, in T47D cells Jeffrey D. Bray, Zhiming Zhang, Richard C. Winneker, C. Richard Lyttle∗ Women’s Health Research Institute, Wyeth Research, 500 Arcola Road, Collegeville, PA 19426, USA

Abstract Progestins play an important role in women’s health and are used in oral contraception, hormone therapy, and treatment of reproductive disorders. The effects of progestins upon gene expression in breast epithelium are poorly understood. In an attempt to characterize the molecular mechanism of progestin action, we used a gene expression profiling approach to examine the action of a novel progestin in the T47D cell model, a human breast cancer cell line. PRA-910 is a novel, nonsteroidal progesterone receptor modulator (PRM) with species-specific activities identified in a screen for selective PRMs. To understand the mechanism of action for PRA-910 in T47D cells, we compared its gene regulation to progesterone (P4) and RU486 through Affymetrix U95A GeneChip® analysis and TaqMan RT–PCR. PRA-910, P4, and RU486 regulated 50, 108, and 16 genes by threefold or greater versus vehicle, respectively, with 18 genes having similar regulation for P4 and PRA-910. These data confirm and extend previous findings for T47D cells. We also obtained time course, concentration–response, cyclohexamide sensitivity, and PR-specificity data for two progestin-regulated genes, ATP1A1 and CLDN8. Our data demonstrate that PRA-910 has a unique gene regulation profile distinct from both P4 and RU486. Further investigation of the underlying mechanism for these differences is ongoing. © 2003 Elsevier Inc. All rights reserved. Keywords: Progesterone; Progesterone receptor; Progesterone receptor modulator; Affymetrix GeneChip® ; PRA-910; Steroid

1. Introduction Progesterone (P4) acts through the progesterone receptor (PR) and is centrally involved in breast development and differentiation. In humans and mice, there are two isoforms of PR designated PR-A and PR-B encoded by a single gene from two promoters, with PR-B having an additional 164 amino acids at its N-terminus [1]. The PR knockout (PRKO) female mouse has a wide range of reproductive phenotypes including mammary defects in pregnancy-induced ductal side-branching and lobular alveolar development [2]. Selective ablations of PR-A or PR-B isoforms have identified the contributions of each isoform in mouse mammary gland development. Proliferation and differentiation of mammary epithelium in the PRAKO female are normal, suggesting that PR-B is sufficient for mammary gland development [3]. However, the PRBKO female has reduced ductal side-branching alveolobular differentiation [4,5]. These data suggest that PR-A and PR-B isoforms are functionally distinct in mediating P4 action in mammary gland development.

∗

Corresponding author. Tel.: +1-484-865-9393; fax: +1-484-865-9389. E-mail address: [email protected] (C.R. Lyttle).

0039-128X/$ – see front matter © 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0039-128X(03)00119-3

Progestins have therapeutic uses such as in oral contraception hormone therapy in combination with an estrogen, and as treatment for some gynecological disorders. Currently, all the clinically available PR modulators (PRMs) have steroidal structures. Development of nonsteroidal progestins with improved receptor selectivity, and a neutral or antiproliferative effect upon the breast is highly desirable. During the process of identifying PR-selective nonsteroidal PRMs, a novel 6-aryl benzoxazinone compound, designated PRA-910 (Fig. 1), with unique in vitro and in vivo activities was discovered. PRA-910 is a PR antagonist in CV-1 monkey kidney cells transfected with hPR-B [6]. In the rat, PRA-910 is a selective PR antagonist comparable to RU486 in blocking P4-induced uterine response and in reversing P4 inhibition of estrogen-induced complement C3 expression in the uterus [6–9]. However, PRA-910 is mainly a PR agonist in T47D cells [6,9]. It is also a PR agonist in the rabbit and primate. In summary, PRA-910 is a novel, nonsteroidal PRM with species-specific activity, differentiating it from most characterized PRMs. The species-specific activities of PRA-910 suggest differences in its mechanism of action compared to P4 or RU486. In order to understand the mechanism of PRA-910 action better, we first searched for genes that are differentially regulated by this PRM using GeneChip® transcription profiling

996

J.D. Bray et al. / Steroids 68 (2003) 995–1003

or PRA-910. To determine if P4 and/or PRA-910 was directly regulating gene expression, cells were pre-treated for 30 min with 10 ␮g/ml cyclohexamide and a progestin was added for 6 h. Gene expression was measured for F3as a cyclohexamide-insensitive control and for S100P as a cyclohexamide-sensitive control [10].

CN

Me

Me F

O

2.3. T47D alkaline phosphatase assay N H

O

Fig. 1. Structure of PRA-910.

in comparison with P4 and RU486 in T47D cells. Furthermore, we showed steroid receptor selectivity, time course and concentration–response data for several candidate genes including ATP1A1 and CLDN8. Our data demonstrate that PRA-910 differentially regulates gene expression in T47D cells compared to P4 and RU486, indicating it has a distinct mechanism of action.

T47D alkaline phosphatase assay was run as described [7,11,12]. Briefly, 50,000 cells/well were plated in 96-well plates in phenol red-free DMEM/F12 containing 5% charcoal stripped FBS overnight. The cells were treated with compounds for 18 h. The medium was removed and 50 ␮l of 0.1 M Tris–Cl, pH 9.8 containing 0.2% Triton X-100 was added to each well with 15 min shaking. Then, 150 ␮l of 0.1 M Tris–Cl, pH 9.8 containing 4 mM p-nitrophenyl phosphate was added. Optical density measurements were taken at 5 min intervals for 30 min at 405 nm wavelength using SpectraMAX PLUS (Molecular Devices, Sunnyvale, CA) and the data analyzed to obtain ED50 or IC50 using JMP software (SAS Institute, Inc., Cary, NC).

2. Experimental

2.4. Affymetrix GeneChip® experiments

2.1. Reagents

T47D cells were plated at 80% confluence in phenol red-free DMEM/F12 containing 5% charcoal stripped FBS for 24 h and treated with 3 nM P4, 1 nM RU486, or 60 nM PRA-910 or EtOH alone (vehicle) for 18 h. These concentrations were the approximate ED50 for each PRMs as determined experimentally by the T47D alkaline phosphatase assay. Total RNA was prepared using TRIzol (Life Technologies, Inc., Rockville, MD) and cleaned using RNeasy columns according to the manufacturers’ protocols (Qiagen, Valencia, CA). Double-stranded cDNA was synthesized from 10 ␮g total RNA using the SuperScript System (Invitrogen, Carlsbad, CA). Briefly, the RNA was mixed with 100 pmol oligonucleotide GGCCATGGAATTGTAATACGACTCACTATAGGGAGGCGG (dT)24 in 20 ␮l water, annealed at 70 ◦ C for 10 min, and quick-chilled. Buffer, dithiothreitol, and dNTP mix were then added and incubated at 37 ◦ C for 2 min. SuperScript II reverse transcriptase was added, and the 50 ◦ C incubation was continued for 60 min. Second-strand synthesis was performed by adding reaction buffer, dNTPs (200 ␮M), DNA ligase (10 U), DNA polymerase (40 U), ribonuclease H (2 U), and water (to a final volume of 150 ␮l), and the reaction was incubated for 2 h at 16 ◦ C. This was followed by addition of 10 U T4 DNA polymerase and incubation at 16 ◦ C for 5 min. The cDNA was purified by phenol/chloroform extraction, precipitated, and transcribed in vitro using T7 RNA polymerase. Biotinylated cRNA was generated using the BioArray HighYield RNA Transcription Kit (Enzo Diagnostics, Inc., Farmingdale, NY). The cRNA was purified by RNeasy column (QIAGEN, Chatsworth, CA) and fragmented by incubation in 40 mM Tris (pH 8.1), 100 mM

Progesterone (P4), dexamethasone (DEX), 5-dihydrotestosterone (DHT), 17␣-ethinyl estradiol (E2 ) were purchased from Sigma Chemical Co. (St. Louis, MO). Mifepristone (RU486) was purchased from Shanghai Organic Chemical Institute (China). Onapristone (ONA, ZK98299), a PR antagonist, and 2-hydroxyflutamide (2-FLUT), an androgen receptor antagonist, were gifts from Ligand Pharmaceuticals (San Diego, CA). Tissue culture media DMEM/F12 was obtained from GibcoBRL (Grand Island, NY). Fetal bovine serum (FBS) was obtained from Hyclone (Logan, UT). PRA-910 was prepared by the Medicinal Chemistry Department at Wyeth [6]. All other chemicals were purchased from Sigma. 2.2. Cell culture and treatments The human breast carcinoma cell line T47D was obtained from American Type Culture Collection (Rockville, MD). T47D cells were maintained in DMEM/F12 supplemented with 10% FBS at 37 ◦ C with 5% CO2 . Cells were split twice weekly at a 1:3 to 1:5 ratio. For experiments, cells were plated at 80% confluence in phenol red-free DMEM/F12 containing 5% charcoal stripped FBS for 24 h, then were treated with 3 nM P4, 1 nM RU486, or 60 nM PRA-910 dissolved in 100% EtOH, or EtOH alone (vehicle) for 18 h, except when otherwise indicated. In the PR-selectivity experiments cells were treated with 10 nM of P4, DEX, DHT, or E2 , and 60 nM of PRA-910, and 100 nM 2-FLUT. ONA was used at 10-fold excess in the co-treatments with P4

J.D. Bray et al. / Steroids 68 (2003) 995–1003

potassium acetate, and 30 mM magnesium acetate buffer at 94 ◦ C for 35 min. Fragmented cRNA were hybridized to a human U95A GeneChip® (Affymetrix, Santa Clara) at 45 ◦ C for 18 h as recommended by the manufacturer. The hybridized chips were washed and stained using Affymetrix Fluidics Station 400 and EukGE-WS1 Standard Format as recommended by the manufacturer. The staining was performed using streptavidin–phycoerythrin conjugate (SAPE; Molecular Probes, Eugene, OR), followed by biotinylated antibody against streptavidin (Vector Laboratories, Burlingame, CA), and then SAPE. The chips were scanned using a Hewlett-Packard GeneArray Scanner and analyzed using Affymetrix MAS 4.0 software. Hybridization intensities were normalized using a spike-in standard curve normalization of Eukaryotic Hybridization Controls. The 5 /3 ratio for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and for ␤-actin ranged from 0.8 to 1.1. Two independent, time-separated experiments were performed and analyzed. 2.5. TaqManTM quantitative RT–PCR assays Cells were grown and plated as above. Total RNA was prepared as above, subjected to DNase I treatment using the RNase-Free DNase I set (Qiagen), and cleaned using RNeasy columns according to the manufacturers’ protocols (Qiagen). Selected regulated genes identified via GeneChip® were verified by real-time quantitative RT–PCR using an ABI PRISM 7900 Sequence Detection System according to the manufacturer’s protocol (PE Applied Biosystems, Foster City, CA). Briefly, TaqMan primers and probes were designed using Primer Express® software (Applied Biosystems) and custom-made using oligonucleotide synthesizer from Applied Biosystems followed by HPLC purification (Genetics Institute, Cambridge, MA). The genes were F3 (forward: 5 -CCTTTTGCACATAACATGCTTTAGATTA-3 ; reverse: 5 -GACATTTTCCCATTTGTTTTTGC-3 ; probe: 5 -6-FAM-CCGCACTTAAGGATTAACCAGGTCGTCC-TAMRA-3 ), S100P (forward: 5 -GGATGCCGTGGATAAATTGC-3 ; reverse: 5 CGAACACGATGAACTCACTGAAG-3 ; probe: 5 -6FAM-ACGCCAATGGAGATGCCCAGGTG-TAMRA-3 ), ATP1A1 (forward: 5 -CACACAGCCTTCTTCGTCAGTATC-3 ; reverse: 5 -CGAATTCCTCCTGGTCTTACAGA3 ; probe: 5 -6-FAM-TGGTGGTGCAGTGGGCCGACTTTAMRA-3 ), CLDN8 (forward: 5 -CCATCGCACAACCCAAAAA-3 ; reverse: 5 -CACAACTACACATACTGACTTCTGGAGTA-3 ; probe: 5-6-FAM-ACACCGGAAAGAAGTCACCGAGCGT-TAMRA-3 ). All primers were used at 0.5 ␮M and probes at 0.2 ␮M concentrations in the PCR reactions. Quantified RNA (2 ␮g) was converted to cDNA using TaqMan Reverse Transcription Reagents (PE Applied Biosystems). Gene-specific primers and probes were used with the TaqMan Universal PCR Master Mix (PE Applied Biosystems) to amplify the equivalent of 50 ng of RNA generated from the cDNA. Reactions were incu-

997

bated at 50 ◦ C for 2 min followed by 10 min at 95 ◦ C then 40 cycles of PCR as follows: 95 ◦ C for 15 s then 60 ◦ C for 1 min in an ABI 7900. The data were analyzed using Sequence Detector version 2.0 software (PE Applied Biosystems) and were normalized to GAPDH using the PE Applied Biosystems primer set. Statistical analyses were performed to obtain a mean ± S.E. on triplicate PCR reactions. 2.6. Data analyses and statistics For the Affymetrix GeneChip® experiments, GeneExpress 2000 Software System Fold Change Analysis (Gaithersburg, MD) tool was used to identify genes expressed at least threefold greater in any treatment compared to vehicle treated for the appropriate time point. For each gene fragment, the tool calculates the ratio of the geometric means of the expression intensities in the treated samples and vehicle control samples as the fold change. Confidence limits and P-values were calculated using a two-sided Welch modified two-sample t-test on the difference of the means of the logs of the intensities. GeneLogic Normalization, a global scaling method that takes into account that small and large expression intensity values are distributed differently, was used in the data analysis. Candidate target gene selection was based on three criteria: the target expression intensity greater than twice the chip sensitivity, show a threefold induction versus vehicle for any treatment and time point, and a P < 0.05 for any treatment’s fold change. For TaqMan statistical analyses, the data were normalized to GAPDH and a ratio of gene/GAPDH was obtained for triplicate PCR reactions. These data were analyzed by one-way ANOVA with Huber Weighting to down-weight the effects of outliers and Least Square Differences for all data except ED50 using JMP software (SAS Institute, Inc.). For obtaining ED50 data, Square root-transformed data were used for ANOVA and nonlinear concentration–response curve fitting in TaqMan analysis and T47D cell alkaline phosphatase assay with Huber Weighting using JMP software (SAS Institute, Inc.). Values of ED50 and IC50 were calculated from the re-transformed values.

3. Results 3.1. Comparison of gene regulation by P4, PRA-910, and RU486 in T47D cells Fig. 1 shows the result of two, time-separated Affymetrix GeneChip® transcriptional profiling experiments represented by a Venn diagram. There were 108 genes regulated by threefold or greater by P4 with twice as many increasing by 18 h. PRA-910 regulates 50 genes with an equal number being up- or downregulated. In contrast, RU486 had little effect upon gene expression and three times as many were downregulated than upregulated. PRA-910 shares

998

J.D. Bray et al. / Steroids 68 (2003) 995–1003 Table 1 Fold-change differences of ATP1A1 and CLDN8 expression by 3 nM P4, 60 nM PRA-910, and 1 nM RU486 compared to vehicle in T47D cells at 18 h

P4 (108)

Fold induction

85 (57↑; 28↓) ATP1A1 CLDN8

RU486

10.4 6.5

5.0 2.8

1.0 0.7

(3↑; 3↓)

TaqManTM quantitative RT–PCR experiments confirmed the GeneChip® results (data not shown). ATP1A1 and CLDN8 are differentially upregulated by PRA-910 compared to P4 and RU486.

1 (1↓)

27

PRA-910

6

16 (12↑; 4↓)

(11↑; 16↓)

P4

6 (1↑; 5↓)

3 (3↓)

3.3. Time course studies PRA-910

RU486

(50)

(16)

Fig. 2. Distribution of genes regulated by P4, PRA-910, and RU486. Numbers in the overlapping regions indicate regulation of genes shared by all three PRMs (bold italic), by two of the three PRMs (italic); most genes are differentially regulated by only one PRM (bold) from two separate experiments. The numbers before the arrows in parentheses represent the genes upregulated (↑) or downregulated (↓) by threefold or greater at 18 h, respectively (P < 0.05).

regulation of 16 genes with P4 and 3 with RU486, while P4 and RU486 share regulation of three genes (Fig. 2, Italics). P4, PRA-910, and RU486 uniquely regulated 85, 27, and 3 genes, respectively, further indicating the unique profile for each PRM (Fig. 2, bold). Only one gene shares common regulation by all three PRMs, with downregulation of Stromal Derived Factor, SDF1 (Fig. 2, bold italics). PRA-910 differentially regulates T47D cell gene expression compared to P4 or RU486. 3.2. Differential regulation of ATP1A1 and CLDN8 by PRA-910 and P4 Genes chosen for confirmation and further study were upregulated by both P4 and PRA-910 in T47D cells in our profiling studies and were known to be expressed in differentiated epithelium. Preference was also given to novel or poorly described PR-regulated genes. Two of these genes, ATP1A1 and CLDN8, were described herein. ATP1A1 is the catalytic subunit of the Na+ ,K+ -ATPase and was previously described as a progestin-regulated gene [13]. CLDN8 is a novel PR-regulated gene and encodes a tight junction structural protein, Cldn8 [14,15]. Table 1 shows the fold change differences as reported by the GeneExpress 2000 Fold Change Analysis Tool. Both ATP1A1 and CLDN8 were induced by P4 twice as much as by PRA-910, while RU486 has no effect on the expression of either gene. Independent

Fig. 3 shows the results of time course studies for ATP1A1 and CLDN8. Following P4 treatment, ATP1A1 mRNA levels increased significantly by 3 h, peaked by 12 h, and dropped by 24 h (Fig. 3A). In contrast, PRA-910 treatment led to a slower and less robust response. The increase in ATP1A1 mRNA levels only became significant at 12 h and maximal induction was seen between 18 and 24 h (Fig. 3A). For CLDN8, significant induction was observed by 3 h for P4 and 6 h for PRA-910 (Fig. 3B). P4 elicited maximal response by 18 h before falling at 24 h, while PRA-910 led to a fairly stable induction from 6 to 24 h. No significant regulation of ATP1A1 or CLDN8 was detected from 36 to 60 h for either treatment (data not shown). These time course data show that, while P4-induced ATP1A1 and CLDN8 expression in a comparable manner, PRA-910 was more potent inducing expression of CLDN8 compared to ATP1A1. 3.4. Concentration–response studies We performed concentration–response experiments to investigate gene-specific differences in PRA-910 potency compared to P4. As demonstrated by a representative concentration–response curve, P4 was clearly more potent and more efficacious than PRA-910 upon ATP1A1 expression (Fig. 4A). However, PRA-910 was more potent in the regulation of CLDN8 expression than ATP1A1, while P4 showed similar potency for CLDN8 and ATP1A1 (Fig. 4B), corresponding to the time course data. Also, CLDN8 was more efficacious than ATP1A1 when compared to P4. Table 2 summarizes weighted ED50 s for P4 and PRA-910, and the P4/PRA-910 ED50 ratio for ATP1A1, CLDN8, and two additional PRM regulated genes, MAN1A1, and SEC14L2. MAN1A1 encodes a Type II endoplasmic reticulum glycosylation enzyme [16] and SEC14L2 encodes a predicted human protein with homology to the yeast phosphatidylinositol/phosphatidylcholine-transfer protein, Sec14p, which is involved in intracellular transport of

J.D. Bray et al. / Steroids 68 (2003) 995–1003

999

0.90 P4 *

0.70

PRA-910

0.60 *

0.50 0.40 *

0.30

*

ATP1A1/GAPDH Gene Expression

*

Vehicle

0.80

**

**

**

0.20 0.10 0.00 3h

6h

12h

(A)

18h

24h

Time

CLDN8/GAPDH Gene Expression

0.90 *

Vehicle

0.80

P4

0.70

*

*

PRA-910

0.60 *

0.50

* *

0.40

*

0.30

**

**

0.20 0.10 0.00 3h

6h

(B)

12h

18h

24h

Time

Fig. 3. Time course of ATP1A1 and CLDN8 gene expression by P4 and PRA-910 in T47D cells. (A) ATP1A1, (B) CLDN8. T47D cells were treated with vehicle (white bars), 3 nM P4 (black bars) or 60 nM PRA-910 (gray bars). The amount of each gene transcript relative to the GAPDH transcript is shown as mean ± S.E. of three separate experiments. (∗) Significantly different from matched time point vehicle control (P < 0.0001), (∗∗) significantly different from matched time point vehicle control (P < 0.05).

Table 2 Weighted P4 and PRA-910 ED50 s, and P4/PRA-910 ED50 ratios for expression of ATP1A1, CLDN8, MAN1A1, and SEC14L2 in T47D cells Gene

P4 ED50 (nM)

ATP1A CLDN8 MAN1A1 SEC14L2

1.1 3.6 2.0 67.6

± ± ± ±

0.2 2.4 1.2 28.3

PRA-910 ED50 (nM) 219.0 11.4 101.6 14.7

± ± ± ±

30.9 6.0 31.0 2.7

P4/PRA-910 ED50 ratio 0.005a 0.316 0.020a 4.60a

Values represent the weighted mean ± S.E. of four separate experiments. a Weighted P4 ED 50 significantly different from weighted PRA-910 ED50 (P < 0.05).

phospholipids [17]. Each of these four genes exhibits a different P4/PRA-910 ED50 ratio. Interestingly, the potency of PRA-910 contributes more to these ratio differences than P4 potency. There is nearly a 1000-fold difference in these ratios. For example, P4 was 200 times more potent than PRA-910 in the induction of ATP1A1, while PRA-910 was five times more potent than P4 in the regulation of SEC14L2 expression. 3.5. Steroid receptor specificity We examined the potential regulation by other steroids on the expression of ATP1A1 and CLDN8 in T47D cells.

1000

J.D. Bray et al. / Steroids 68 (2003) 995–1003

1.6 CLDN8/GAPDH Gene Expression

P4 mean

(A)

1.4

PRA-910 mean

1.2 1 0.8 0.6 0.4 0.2 0 0.001

0.01

0.1

1

10

100

1000

10000 100000

1000

10000 100000

Concentration (nM) 1.8 ATP1A1/GAPDH Gene Expression

P4 me an

1.6

PRA-910 mean

1.4 1.2 1 0.8 0.6 0.4 0.2

(B)

0 0.001

0.01

0.1

1

10

100

Concentration (nM) Fig. 4. Concentration curve results for ATP1A1 and CLDN8 gene expression regulation by P4 and PRA-910 in T47D cells. (A) ATP1A1, (B) CLDN8. T47D cells were treated with 0.03–3000 nM P4 (squares) or PRA-910 (diamonds) at 0.5 log concentrations for 18 h. Shown is a representative concentration–response curve of mean values.

Progestin-induced expression for these genes was cyclohexamide insensitive at 6 h, suggesting direct regulation through PR (data not shown). P4 and PRA-910-induced ATP1A1 expression by 17- and 4-fold, respectively, and a 10-fold excess of the PR antagonist onapristone suppressed this induction (Fig. 5A). None of the other steroids tested had an effect on ATP1A1 gene expression. P4 and PRA-910-induced CLDN8 levels by 10- and 7-fold, respectively, and the induction was suppressed by onapristone (Fig. 5B). CLDN8 also shows a sixfold induction by DHT (Fig. 5B), but is not suppressed by 2-FLUT, an androgen receptor antagonist (data not shown). The differ-

ences in steroid selective profiles upon these genes suggest that promoter context may contribute to PR-dependent expression.

4. Discussion This is the first study to directly examine the effects of a nonsteroidal PRM upon global and specific gene expression. We are ultimately interested in understanding the mixed agonist/antagonist profile and species-specific activity differences of PRA-910. In this study we focused on

J.D. Bray et al. / Steroids 68 (2003) 995–1003

1001

(A)

(B)

Fig. 5. Regulation of ATP1A1 and CLDN8 gene expression by P4, E2, DEX, DHT, ONA, and PRA-910 in T47D cells. (A) ATP1A1, (B) CLDN8. T47D cells were treated with vehicle (VEH), 10 nM P4, E2, DEX, or DHT, 60 nM PRA-910, or 100 nM ONA for 18 h, then RNA was extracted TaqMan quantitative RT–PCR. ONA was added at 10-fold concentration to antagonize P4 and PRA-910. The amount of each gene transcript relative to the GAPDH transcript is shown as mean ± S.E. of three separate experiments. (∗) Significantly different from vehicle control (P < 0.0001), (∗∗) significantly different from vehicle control (P < 0.05).

the human breast cancer cell line T47D and the global gene regulation profile associated with PRA-910 compared to P4 and RU486. Additionally, even where genes are regulated in common, PRA-910 demonstrated a unique profile of gene regulation compared to P4. Here, we have eliminated species and tissue differences and still observe differences in P4- and PRA-910-induced gene expression in

T47D cells suggesting a distinct mechanism of action for PRA-910. Regulation of gene expression by PRA-910 is different from P4 or RU486 in T47D cells as detected by GeneChip® analysis. Whereas P4 is mainly a transcriptional activator, PRA-910 repressed 50% of genes it regulated. Detailed analyses of several marker genes, including ATP1A1 and

1002

J.D. Bray et al. / Steroids 68 (2003) 995–1003

CLDN8, show that ED50 values for PRA-910 and the P4/PRA-910 ED50 ratios are different for these genes. Changes of PRA-910 ED50 values contribute most significantly to these ratio differences, while ED50 values for P4 vary little in comparison. If the difference between P4 and PRA-910 were due solely to potency, one would expect to see similar ratios for any gene examined. Finally, steroid selectivity profiles and time-dependent activation differs for individual genes. Further studies are underway to elucidate the potential mechanism of action for this novel compound. Many studies suggest that PR-A and PR-B are not functionally redundant, and that each performs a precise role in P4-dependent responses [18–24]. To address this issue, we have begun experimentation in T47D breast cancer cell lines engineered to express either PR-A or PR-B to determine the role of each isoform in PRA-910 gene regulation. P4 is known to regulate different subsets of genes, including genes known to be involved in mammary gland development and/or breast cancer in these T47D cells engineered to express either PR-A or PR-B [10]. Preliminary results suggest that PR isoform selectivity may contribute to PRA-910’s unique effects in T47D cells (data not shown). PRA-910 appears to be a selective progesterone receptor modulator (SPRM); a compound with mixed PR agonist and antagonist activities depending upon the cell type or species [25]. Currently, the mechanisms of action of SPRMs are unknown, but may be due to cell-specific levels and ratios of PR isoforms and/or its coregulators. The presence and ratios of specific coregulators will likely play a role in PRA-910 mechanism of action. The antiprogestin RU486 demonstrates a differential ability to activate PR-dependent transcription though different ratios of coregulators in cultured cells [26]. Corepressor levels have been demonstrated as critical components to the tissue selective effects of selective estrogen receptor modulators [27,28]. Presently it is unknown if PRA-910 recruits different regulators in T47D cells compared to P4 or RU486. The observed gene selective regulation involves promoter differences. The differences in steroid activation profiles for ATP1A1 and CLDN8 indicate that promoter differences may have a significant effect upon expression of individual genes. Promoter differences may contribute to the differing potencies of a progestin upon expression of an individual gene. The use of endogenous PR-regulated promoters may also be more useful than consensus PREs to determine the contribution of the promoter to the mechanism of PRA-910 action as compared to P4. While these studies use T47D breast cancer cells, there is a correlation between a cancer cell line, the primary tumor, and the normal tissue of origin based on gene expression patterns [29]. P4 induces a differentiation of T47D cells and is associated with a secretory phenotype typical of fully differentiated breast epithelial cells [12]. Also, progestins induce differentiation in T47D cells based on immunofluo-

rescence and identification of progestin-regulated genes, including ATP1A1 [13]. Our gene expression results are consistent with these reports and also demonstrate more genes upregulated than down in T47D cells by progestins [10,13]. Based on these findings, we find that T47D cells are a useful model system for examining PRM-regulated differential gene expression, and have confirmed and extended previous gene expression studies.

Acknowledgements The authors wish to thank Jeff Cohen for his technical expertise and tissue culture assistance, and Yihe Wang of the WHRI DNA Core for his technical expertise with the Affymetrix GeneChip® sample processing. We also acknowledge Dr. Maryann Whitley of Wyeth Expression Profiling Informatics for her comments and help with the GeneChip® data analyses.

References [1] Kastner P, Krust A, Turcotte B, Stropp U, Tora L, Gronemeyer H, et al. Two distinct estrogen-regulated promoters generate transcripts encoding two functionally different human progesterone receptor forms A and B. EMBO J 1990;9:1603–14. [2] Lydon JP, DeMayo FJ, Funk CR, Mani SK, Hughes AR, Montgomery Jr CA, et al. Mice lacking progesterone receptors exhibit pleiotropic reproductive abnormalities. Genes Dev 1995;9:2266–78. [3] Mulac-Jericevic B, Mullinax RA, DeMayo FJ, Lydon JP, Conneely OM. Subgroup of reproductive functions of progesterone mediated by progesterone receptor-B isoform. Science 2000;289:1751–4. [4] Conneely OM, Jericevic BM. Progesterone regulation of reproductive function through functionally distinct progesterone receptor isoforms. Rev Endocr Metab Disord 2002;3:201–9. [5] Conneely OM, Mulac-Jericevic B, DeMayo F, Lydon JP, O’Malley BW. Reproductive functions of progesterone receptors. Recent Prog Horm Res 2002;57:339–55. [6] Zhang P, Terefenko EA, Fensome A, Wrobel J, Winneker R, Lundeen S, et al. 6-Aryl-1,4-dihydro-benzo[d][1,3]oxazin-2-ones: a novel class of potent, selective, and orally active nonsteroidal progesterone receptor antagonists. J Med Chem 2002;45:4379–82. [7] Zhang Z, Lundeen SG, Zhu Y, Carver JM, Winneker RC. In vitro characterization of trimegestone: a potent and selective progestin. Steroids 2000;65:637–43. [8] Lundeen SG, Zhang Z, Zhu Y, Carver JM, Winneker RC. Rat uterine complement C3 expression as a model for progesterone receptor modulators: characterization of the new progestin trimegestone. J Steroid Biochem Mol Biol 2001;78:137–43. [9] Zhang Z, Zhu Y, Rudnick K, Lundeen S, Slayden O, Zhang P, et al. Identification and characterization of a novel nonsteroidal, species-specific progesterone receptor modulator PRA-910. In: Poster presented at the 84th Endocrine Society Meeting, 2002. [10] Richer JK, Jacobsen BM, Manning NG, Abel MG, Wolf DM, Horwitz KB. Differential gene regulation by the two progesterone receptor isoforms in human breast cancer cells. J Biol Chem 2002;277:5209– 18. [11] Di Lorenzo D, Albertini A, Zava D. Progestin regulation of alkaline phosphatase in the human breast cancer cell line T47D. Cancer Res 1991;51:4470–5.

J.D. Bray et al. / Steroids 68 (2003) 995–1003 [12] Di Lorenzo D, Gianni M, Savoldi GF, Ferrari F, Albertini A, Garattini E. Progesterone induced expression of alkaline phosphatase is associated with a secretory phenotype in T47D breast cancer cells. Biochem Biophys Res Commun 1993;192:1077–82. [13] Kester HA, van der Leede BM, van der Saag PT, van der Burg B. Novel progesterone target genes identified by an improved differential display technique suggest that progestin-induced growth inhibition of breast cancer cells coincides with enhancement of differentiation. J Biol Chem 1997;272:16637–43. [14] Morita K, Furuse M, Fujimoto K, Tsukita S. Claudin multigene family encoding four-transmembrane domain protein components of tight junction strands. Proc Natl Acad Sci USA 1999;96: 511–6. [15] Tsukida S, Furuse M. Pores in the wall: claudins constitute tight junction strands containing aqueous pores. J Cell Biol 2000;149: 13–6. [16] Bause E, Bieberich E, Rolfs A, Volker C, Schmidt B. Molecular cloning and primary structure of Man9-mannosidase from human kidney. Eur J Biochem 1993;217:535–40. [17] Cockcroft S. Phosphatidylinositol transfer proteins: a requirement in signal transduction and vesicle traffic. BioEssays 1998;20:423–32. [18] Takimoto GS, Tasset DM, Eppert AC, Horwitz KB. Hormone-induced progesterone receptor phosphorylation consists of sequential DNA-independent and DNA-dependent stages: analysis with zinc finger mutants and the progesterone antagonist ZK98299. Proc Natl Acad Sci USA 1992;89:3050–4. [19] Meyer ME, Quirin-Stricker C, Lerouge T, Bocquel MT, Gronemeyer H. A limiting factor mediates the differential activation of promoters by the human progesterone receptor isoforms. J Biol Chem 1992;267:10882–7. [20] Tung L, Mohamed MK, Hoeffler JP, Takimoto GS, Horwitz KB. Antagonist-occupied human progesterone B-receptors activate transcription without binding to progesterone response elements

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28] [29]

1003

and are dominantly inhibited by A-receptors. Mol Endocrinol 1993;7:1256–65. Vegeto E, Shabaz MM, Wen DX, Goldman ME, O’Malley BW, McDonnell DP. Human progesterone receptor A form is a celland promoter-specific repressor of human progesterone receptor B function. Mol Endocrinol 1993;7:1244–55. Sartorius CA, Tung L, Takimoto GS, Horwitz KB. Antagonistoccupied human progesterone receptors bound to DNA are functionally switched to transcriptional agonists by cAMP. J Biol Chem 1993;268:9262–6. McDonnell DP, Shahbaz MM, Vegeto E, Goldman ME. The human progesterone receptor A-form functions as a transcriptional modulator of mineralocorticoid receptor transcriptional activity. J Steroid Biochem Mol Biol 1994;48:425–32. Hovland AR, Powell RL, Takimoto GS, Tung L, Horwitz KB. An N-terminal inhibitory function, IF, suppresses transcription by the A-isoform but not the B-isoform of human progesterone receptors. J Biol Chem 1998;273:5455–560. Spitz I, Chwalisz K. Progesterone receptor modulators and progesterone antagonists in women’s health. Steroids 2000;65:807– 15. Lui Z, Auboeuf D, Wong J, Chen JD, Tsai SY, Tsai MJ, et al. Coactivator/corepressor ratios modulate PR-mediated transcription by selective receptor modulator RU486. Proc Natl Acad Sci USA 2002;99:7940–4. Smith CL, Nawaz Z, O’Malley BW. Coactivator and corepressor regulation of the agonist/antagonist activity of the mixed antiestrogen, 4-hydroxytamoxifen. Mol Endocrinol 1997;11:657–66. Shang Y, Brown M. Molecular determinants for the tissue specificity of SERMs. Science 2002;295:2465–8. Ross DT, Scherf U, Eisen MB, Perou CM, Rees C, Spellman P, et al. Systematic variation in gene expression patterns in human cancer cell lines. Nat Genet 2000;24:227–35.