Progesterone Receptor–Mediated Regulation of Cellular Glucose and ¹⁸F-Fluorodeoxyglucose Uptake in Breast Cancer

Abstract

Context

Positron emission tomography imaging with 2-deoxy-2-[¹⁸F]-fluoro-D-glucose (FDG) is used clinically for initial staging, restaging, and assessing therapy response in breast cancer. Tumor FDG uptake in steroid hormone receptor–positive breast cancer and physiologic FDG uptake in normal breast tissue can be affected by hormonal factors such as menstrual cycle phase, menopausal status, and hormone replacement therapy.

Objective

The purpose of this study was to determine the role of the progesterone receptor (PR) in regulating glucose and FDG uptake in breast cancer cells.

Methods and Results

PR-positive T47D breast cancer cells treated with PR agonists had increased FDG uptake compared with ethanol control. There was no significant change in FDG uptake in response to PR agonists in PR-negative MDA-MB-231 cells, MDA-MB-468 cells, or T47D PR knockout cells. Treatment of T47D cells with PR antagonists inhibited the effect of R5020 on FDG uptake. Using T47D cell lines that only express either the PR-A or the PR-B isoform, PR agonists increased FDG uptake in both cell types. Experiments using actinomycin D and cycloheximide demonstrated the requirement for both transcription and translation in PR regulation of FDG uptake. GLUT1 and PFKFB3 mRNA expression and the enzymatic activity of glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase were increased after progestin treatment of T47D cells.

Conclusion

Thus, progesterone and progestins increase FDG uptake in T47D breast cancer cells through the classical action of PR as a ligand-activated transcription factor. Ligand-activated PR ultimately increases expression and activity of proteins involved in glucose uptake, glycolysis, and the pentose phosphate pathway.

Approximately 70% of breast cancers are estrogen receptor (ER) and/or progesterone receptor (PR) positive [1, 2]. In breast cancer, PR is expressed as 2 predominant isoforms, PR-A and PR-B, using separate promoters of the PGR gene [3]. PR-B is the longer isoform, having an additional 164 amino acids in the N-terminus compared with PR-A [3]. The receptor isoforms have similar ligand binding and DNA binding functions; however, they differ in their transcriptional activity [4–6]. Generally, PR-B is a stronger regulator of transcription in response to progesterone than PR-A [4–6]. While there are genes regulated by both isoforms, many progesterone-regulated target genes are independently controlled by one isoform or the other [7]. Separate from the classic nuclear action of PR, rapid membrane-initiated signaling events can also be stimulated by progestins [8, 9]. Crosstalk between hormonal and metabolic signaling pathways have been hypothesized to be a major driving force for endocrine therapy resistance and metastasis [10].

Cancer cells have increased glucose metabolism and switch to using aerobic glycolysis for energy generation rather than the more efficient process of oxidative phosphorylation. This metabolic reprogramming, termed the “Warburg effect” [11–13], is a cancer hallmark that can be leveraged clinically using the glucose analog, 2-deoxy-2-[¹⁸F]-fluoro-D-glucose (FDG) [14–16]. Like glucose, FDG is transported into the cell through facilitative glucose transporter (GLUT) proteins [17–20] and then phosphorylated by hexokinase to FDG-6-phosphate [21, 22]. Once phosphorylated, unlike glucose-6-phosphate, FDG-6-phosphate cannot be metabolized further through glycolysis due to the presence of ¹⁸F instead of the 2-hydroxyl group [21, 22]. Due to low levels of glucose-6-phosphatase needed for dephosphorylation, FDG-6-phosphate cannot be released and is thus trapped intracellularly [21–23]. Accumulation of FDG can be visualized and quantified using positron emission tomography (PET) imaging. Clinically, FDG PET/computed tomography imaging can be used for initial staging, restaging, and assessing therapy response for patients with breast cancer [24].

Physiologic FDG uptake in normal breast tissue is affected by hormonal factors. Studies using whole-body PET imaging, as well as high-resolution dedicated breast PET devices, have demonstrated higher physiologic FDG uptake in normal breast tissue of premenopausal women compared with postmenopausal women [25–28]. Furthermore, physiologic FDG uptake in normal breast tissue of postmenopausal women taking hormone replacement therapy was similar to premenopausal women [25]. The specific composition of the hormone replacement therapy used in that study was not reported. Furthermore, a weak but significant positive correlation between serum progesterone level and FDG uptake was demonstrated in the contralateral normal breast tissue of premenopausal women with breast cancer. No significant correlation was found between serum estradiol levels and FDG uptake [29]. A significant correlation between menstrual cycle phase and FDG uptake was seen in normal breast, with higher uptake during the secretory/luteal phase when progesterone levels peak from the corpus luteum [28, 30–32]. Collectively, these clinical studies provide indirect evidence for a role of steroid hormones, particularly progesterone, in FDG uptake of normal breast tissue.

FDG uptake in steroid hormone receptor positive breast cancer can also be influenced by fluctuating hormonal levels. A correlation between FDG uptake and menstrual cycle for luminal A subtype tumors (ER+/HER2−, Ki67 < 14%, PR ≥ 20%) was demonstrated, with maximum standardized uptake values significantly higher for patients during the periovulatory/luteal phase than for those during the follicular phase [33]. However, tumor FDG uptake was not significantly different between the menstrual phases for patients with luminal B (ER+/HER2−, Ki67 > 14%, PR < 20%) or nonluminal tumors [33].

The role of progesterone and its signaling through PR in regulating FDG uptake has not been defined. The aim of this study was to determine the role PR has in regulating glucose and FDG uptake in breast cancer cells via pharmacologic and genomic-based approaches. Given growing evidence of PR and ER crosstalk in breast cancer tumor biology [34–36] and current clinical trials testing new antiprogestin therapeutics selectively targeting PR [37–40], a better understanding of how ligand-activated PR influences FDG uptake could be helpful for using PET/computed tomography imaging to predict and monitor therapy response.

Materials and Methods

Cell Culture

Experiments were performed under a protocol approved by the University of Wisconsin-Madison Office of Biological Safety. T47D, MDA-MB-231, and MDA-MB-468 cells were obtained from American Type Culture Collection. Modified T47D cell lines were also used including those with CRISPR knockout of the PGR gene (PR KO) and the scrambled (Scr) control [41], and cell lines expressing only the PR-A protein isoform (T47D YA) or the PR-B protein isoform (T47D YB) [42]. Cell lines were authenticated using short tandem repeat analysis and tested negative for murine pathogens and Mycoplasma contamination. All T47D cells were cultured in Roswell Park Memorial Institute media (RPMI 1640 with L-glutamine; Corning 10-040-CV). MCF-7, MDA-MB-231, and MDA-MB-468 cells were cultured in Dulbecco's Modified Eagle medium (DMEM with 4.5 g/L glucose, L-glutamine, sodium pyruvate; Corning 10-013-CV). Media was supplemented with 10% fetal bovine serum (VWR) and 1% penicillin–streptomycin (GeminiBio). Cells were maintained at either 5% or 10% CO₂ at 37 °C. For hormone deprivation, cells were grown in phenol red–free media (Corning DMEM 17-205-CV or RPMI 17-205-CV) supplemented with 5% charcoal–dextran stripped fetal bovine serum, 2% L-glutamine, and 1% penicillin–streptomycin. Cells were trypsinized using phenol red–free trypsin (0.05%, Gibco).

FDG Cell Uptake Assay

Cells were hormone starved for 3 days, then seeded into a 24-well plate at a density of 4 × 10⁴ cells in a total of 0.5 mL of medium. Parallel wells were seeded for protein measurement. The following day, cells were treated with various hormones: 0.01 to 100 nM promegestone (R5020, Perkin Elmer), 100 nM medroxyprogesterone acetate (MPA, Steraloids), 100 nM megestrol acetate (MA, Steraloids), 100 nM mifepristone (RU486, Sigma), 100 nM ulipristal acetate (UPA, Sigma), 100 nM progesterone (Sigma), or 100 nM dexamethasone (Sigma). For experiments testing the requirement of transcription and translation, cells were pretreated with actinomycin D (1 μM; Sigma) or cycloheximide (10 μg/mL; Sigma) for 1 hour prior to the addition of hormone. Thirty minutes prior to the assay, cells were washed twice with phosphate-buffered saline (PBS) and placed into either no glucose or high glucose (4.5 g/L) DMEM medium without glutamine or phenol red (Gibco A14430-01 or 31053-028, respectively) containing the same treatment. FDG (1 μCi, Sofie) was added to cells and incubated for 40 minutes at 37 °C. After the incubation, the medium was removed, cells were washed with PBS, and then lysed with 1 N NaOH. Radioactivity was measured using a gamma counter (PerkinElmer) and decay corrected and normalized to protein content. Specific FDG uptake was determined by subtracting the values of the high glucose medium lysates (nonspecific activity) from the corresponding no glucose medium lysates (total activity). Protein concentration was determined using the Bradford assay.

Glucose Consumption and Glucose Uptake Assays

Cells were hormone deprived for 3 days and then seeded at a density of 1 × 10⁵ cells per well. Cells were treated with 100 nM R5020, MPA, MA, or ethanol vehicle (0.1%, v/v) for 24 hours. Cellular glucose consumption was measured using a point-of-care glucometer (AccuChek) and determined by subtracting the glucose value measured in wells with medium alone, without cells, from the glucose value measured in medium from wells containing cells with the respective treatments. Cells were counted and the amount of glucose consumed was determined on a per-cell basis. Glucose uptake was also determined by measuring 2-deoxyglucose-6-phosphate (2DG6P) using the nonradioactive Glucose Uptake-Glo bioluminescent assay (Promega). Cells were hormone deprived for 3 days and seeded at a density of 4 × 10⁴ per well in a 96-well white plate with a clear bottom. Cells were treated with or without 100 nM R5020 for 24 hours and the assay was performed following the manufacturer's protocol. Data were normalized to protein content and expressed as fold change relative to ethanol control.

Reporter Gene Assays

Cells were hormone deprived for 3 days, seeded, and cotransfected with a progesterone response element (PRE) firefly luciferase reporter gene plasmid [43] and a beta-galactosidase control plasmid [44] using Lipofectamine 3000 for 5 hours (Invitrogen). The next day, cells were treated with hormones. After treatment, cells were washed and harvested. Luciferase activity (Promega) and beta-galactosidase activity (Applied Biosystems) assays were performed using the manufacturers’ protocols.

Flow Cytometry

T47D cells were hormone deprived for 3 days. Cells were treated for 24 hours with either 100 nM progesterone, R5020, or ethanol. Cells were harvested, washed with PBS, and fixed in 70% ethanol. Fixed cells were stained with 50 µg/mL propidium iodine (Sigma-Aldrich) containing 100 µg/mL RNase (Sigma-Aldrich) overnight at 4 °C. Cell cycle analysis was performed using Attune NxT (ThermoFisher) flow cytometer and data were analyzed using FlowJo software (FlowJo, LLC).

Western Blot Analysis

Cells were lysed in radioimmunoprecipitation assay buffer containing protease inhibitors (Sigma) and phosphatase inhibitors (Sigma). Protein concentration was measured using Bradford Reagent (Bio-Rad) and equal amounts of protein were run on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis gels. Protein was transferred to PVDF and incubated with antibodies against PR (1:1000; Leica Biosystems Cat# NCL-L-PGR-312, RRID:AB_563967) or β-actin (1:10 000; Sigma-Aldrich Cat# A1978, RRID:AB_476692). Protein expression was quantified using ImageJ (1.48 V).

Quantitative Real-Time Polymerase Chain Reaction

Cells were grown in hormone depleted media for 72 hours and plated in 6 well plates at a density of 3.75 × 10⁵ cells per well. Cells were treated with 10 nM R5020 for varying incubation times. RNA was extracted using RNeasy (Qiagen) and cDNA was synthesized using iScript DNA synthesis (Bio-Rad). Primer sequences are as follows—FKBP54: forward 5′-AGAACCAAACGGAAAGGAGAG-3′ and reverse 5′-TCACCGCCTGCATGTATTT-3′; HK2: forward 5′-GGGACAATGGATGCCTAGATG-3′ and reverse 5′-GTTACGGACAATCTCACCCAG-3′; G6PD: forward 5′-CCCGGAAACGGTCGTACACT-3′ and reverse 5′-CATGACGCTGTCTGCGCTTC-3′; PFKFB3: forward 5′-GGCAAGACCTACATCTCCAAG-3′ and reverse 5′-ATGGCTTCCTCATTGTCGG-3′; SLC2A1 (GLUT1): forward 5′-TCATCGTGGCTGAACTCTTC-3′ and reverse 5′-GATGAAGACGTAGGGACCAC-3′; SLC2A4 (GLUT4): forward 5′-TCGGTTCTTTCATCTTCGCC-3′ and reverse 5′-AGAACACAGCAAGGACCAG-3′; SLC2A12 (GLUT12): forward 5′-TTGCTTGTTTATGTTGCTGCT-3′ and reverse 5′-AGATTGATGCCCCAGTTCATG-3′; and RPLP0: forward 5′-CCATTCTATCATCAACGGGTACAA-3′ and reverse 5′-TCAGCAAGTGGGAAGGTGTAATC-3′. C_T values were normalized to the housekeeping gene, RPLP0. Relative fold change was calculated using the formula 2^−ΔΔC_T [45].

Glucose-6-Phosphate Dehydrogenase and 6-Phosphogluconate Dehydrogenase Activity Assays

T47D cells were hormone deprived for 3 days prior to experimentation. Cells were seeded and treated with 100 nM R5020 for 24 hours. The cells were washed twice with ice cold PBS, scraped into 1 mL of PBS, and pelleted. Glucose-6-phosphate dehydrogenase (G6PD) and 6-phosphogluconate dehydrogenase (6PGD) activity assays were performed at the University of Iowa Free Radical and Radiation Biology core. Cell pellets were lysed in diethylenetriaminepentaacetic acid (DETAPAC) buffer. For each sample, 2 cuvettes were prepared—one for total activity (G6PD + 6PGD) and another for 6PGD activity. For total activity, substrate was added that consisted of 750 μL of nicotinamide adenine dinucleotide phosphate (NADP) reagent (2 mM NADP⁺ in 100 mM Tris buffer, pH 8.0 with 1 mM MgCl₂), 100 μL of G6P (12 mM D-glucose 6 phosphate in ddH₂O), and 100 μL of 6PGA (12 mM 6-phosphogluconic acid in ddH₂O). For the 6PGD activity, 750 μL of NADP reagent, 100 μL of ddH₂O, and 100 μL of 6PGA were added. Cuvettes were incubated for 5 minutes at 37 °C prior to samples being added. Within each cuvette, 50 μL of sample was added and mixed. NADPH formation was measured at 340 nm for 5 minutes. G6PD activity was calculated by subtracting the 6PGD activity from the total activity.

PR Binding Site Identification and Motif Analysis

Raw sequencing data of PR DNA-binding sites following treatment with R5020 in T47D cells was acquired from a previous publication (GEO accession number GSE68359) [35]. Reads were aligned to the hg38 genome using Bowtie2 [46]. Peaks were called with MACS2 [47] using input data as a control and false discovery rate of 0.05. PR binding sites were annotated to the nearest gene using Genomic Regions Enrichment of Annotations Tool (GREAT, version 4.0.4 [48]) with a gene association rule of “two nearest genes” and default settings. Binding sites were then filtered for those within 100 kb of the nearest transcription start site. Motif analysis was performed using the Find Individual Motif Occurrences (FIMO) software tool [49] in the Multiple Expectation maximizations for Motif Elicitation (MEME) Suite [50].

Chromatin Immunoprecipitation

T47D cells were hormone deprived for 3 days then were treated with either 10 nM R5020 or ethanol vehicle for 3 hours. Samples were prepared for chromatin immunoprecipitation (ChIP) as previously described [51]. Precleared chromatin was subjected to immunoprecipitation with PR A/B rabbit monoclonal antibody (1:100; Cell Signaling Cat# 8757, RRID:AB_2797144) or rabbit IgG control and incubated overnight at 4 °C while rotating. A bolus (45 µL) of 50% Protein Sepharose A beads (Cytiva) was then added and incubated for 1 hour at 4 °C. The precipitated complexes were washed once with 1 mL of TSE I buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 150 mM NaCl), once with 1 mL of TSE II buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 500 mM NaCl), once with 1 mL of TSE III Buffer (0.25 M LiCl, 1% NP-40, 1% sodium deoxycholate, 10 mM Tris-HCl pH 8.1), and twice with 1 mL of TE (10 mM Tris-HCl pH 8.1, 1 mM EDTA). Chromatin complexes were extracted from beads by incubating in 75 µL of extraction buffer (1% SDS, 0.1 M NaHCO₃) for 30 minutes. Extraction was repeated 3 times and the extractions were pooled. Samples were reverse cross-linked by adding 10 µL of 5 M NaCl to each sample and incubating at 65 °C overnight. The next day, 4.2 µL of Proteinase K (Qiagen) was added and samples were incubated at 65 °C for 1 hour. DNA was purified using the QIAquick PCR Purification Kit (Qiagen) and 1 µL of the purified DNA was analyzed via quantitative real-time polymerase chain reaction (qPCR). Data are calculated as a percent of input. Primers used for ChIP-qPCR were PFKFB3: forward 5′-CCCAGCATCCCCTTACACAG-3′ and reverse 5′-CGTCCCAGGTTGAAGGACAT-3′; SLC2A1 (GLUT1): forward 5′-TGGTGCCTTCCTTCCCATTG-3′ and reverse 5′-CTGCTGACTTGGGACAGGTT-3′; and SYNE1: forward 5′-AACGGGAATCCAACTAAGCCT-3′ and reverse 5′-CCATAACCTGTCGCAGAAGC-3′.

Statistical Analysis

Data represent the mean ± SEM of 3 independent experiments. Statistical analysis was performed using GraphPad Prism 7.0 software. Unpaired t tests and 1-way analysis of variance with multiple comparison post-tests were used to compare treatment effect relative to the ethanol controls. An unpaired t-test was used to compare PR protein levels in T47D YA and T47D YB cells.

Results

PR Agonists Stimulate FDG Uptake and Glucose Uptake in T47D Breast Cancer Cells

To determine the effect of ligand-activated PR on FDG uptake, three PR agonists were used—promegestone (R5020), MPA, and MA. T47D, MDA-MB-231, and MDA-MB-468 cells were hormone deprived for 3 days and then were treated with either R5020, MPA, MA (each 100 nM), or ethanol (0.1% v/v) for 24 hours. Cells were then incubated with 1 μCi of FDG for 40 minutes and retained radioactivity was measured. In the PR-positive T47D cells, FDG uptake was significantly increased by R5020 (12.9 ± 1.2 × 10⁵ CPM/mg protein, P = .01), MPA (16.6 ± 1.2 × 10⁵ CPM/mg protein, P = .003), and MA (16.3 ± 1.2 × 10⁵ CPM/mg protein, P = .003) vs ethanol control (6.9 ± 0.8 × 10⁵ CPM/mg protein; Fig. 1A). However, no progestin hormonal effect was observed in the PR-positive MCF-7 cell line (Fig. 1B) or PR-negative cell lines MDA-MB-231 (Fig. 1C) or MDA-MB-468 (Fig. 1D). PR transactivation by R5020 was confirmed using a reporter gene assay (Fig. 1E) which showed stronger PR transcriptional activity in T47D cells than in MCF-7 cells and no progestin effect in MDA-MB-231 and MDA-MB-468 cells. Treatment with the natural PR agonist, progesterone (100 nM for 24 hours), increased FDG uptake in T47D cells similar to the response observed with synthetic progestins (1.8 ± 0.2 fold increase compared with ethanol control, P = .02). However, treatment of T47D cells with 100 nM dexamethasone, a glucocorticoid receptor (GR) agonist, for 24 hours had no effect on FDG uptake compared with ethanol (1.0 ± 0.2 fold change vs 1.0 ± 0.2, P = 1.0, Fig. 1F). Treatment with R5020 or progesterone for 24 hours significantly decreased the percentage of T47D cells in G0/G1 and increased the percentage of cells in G2 compared with ethanol control (Table 1). Taken together, FDG uptake is increased after PR agonist treatment in T47D cells.

Figure 1.

PR agonist induction of FDG uptake in T47D cells. Cells were steroid hormone deprived for 3 days prior to experimentation then seeded, and the next day treated with 100 nM R5020, medroxyprogesterone acetate (MPA), megestrol acetate (MA), or ethanol (EtOH) control. After treatment, FDG uptake assay was performed in (A) T47D, (B) MCF-7, (C) MDA-MB-231, and (D) MDA-MB-468 cells. Thirty minutes prior to the addition of 1 μCi FDG, cells were washed with PBS and placed in medium with or without glucose plus the treatment. Cells were incubated with FDG for 40 minutes, then washed and lysed, and radioactivity was measured. Specific FDG uptake was calculated and normalized to protein content. (E) Reporter gene assay for PR transcriptional activation after 24 hours treatment with 100 nM of R5020 or ethanol control. (F) FDG uptake was measured in T47D cells after 24 hours treatment with 100 nM progesterone (Pg) or dexamethasone (Dex). Data represents the mean ± SEM of 3 independent experiments. *P < .05 compared with ethanol.

Table 1.

Cell cycle analysis in T47D cells

	G0/G1	S	G2
EtOH	58.0 ± 1.0	18.9 ± 1.1	13.7 ± 0.5
R5020	43.5 ± 1.3*	17.9 ± 0.9	27.5 ± 1.3*
Pg	39.5 ± 3.0*	21.7 ± 0.9	25.5 ± 0.9*

Values represent percentage of cells in the various stages of the cell cycle. Data represent the mean ± SEM of 3 individual experiments

^* P < 0.05 compared with ethanol (EtOH) control.

Glucose Uptake in T47D Cells Is Increased After PR Agonist Treatment

To confirm FDG uptake, glucose consumption was also measured in T47D cells after 24 hours treatment with PR agonists using a glucometer. There was a significant increase in glucose consumed after R5020 and MPA treatment compared with ethanol (1.9 ± 0.3 fold change, P = .04 and 1.9 ± 0.2 fold change, P = .02 vs 1.0 ± 0.2, respectively, Fig. 2A). MA treatment resulted in a nonsignificant increase in glucose consumption (1.7 ± 0.3 fold increase). Additionally, glucose uptake based on 2DG6P formation was significantly increased in T47D cells treated with R5020 (100 nM) compared with ethanol (1.7 ± 0.2 fold change vs 1.0 ± 0.07, P = .03; Fig. 2B). Taken together, these data show that PR agonists increase glucose consumption and glucose uptake in T47D cells in confirmation with FDG uptake.

Figure 2.

PR agonist induction of glucose uptake in T47D cells. T47D cells were steroid hormone deprived for 3 days then treated with 100 nM of R5020, MPA, or MA, or ethanol control for 24 hours. (A) Glucose consumption was measured in T47D cells using a glucometer. (B) 2-deoxyglucose uptake was measured in T47D cells using the Glucose Uptake-Glo assay. Values were normalized to the ethanol control. Data represent the mean ± SEM of 3 independent experiments. *P < .05 compared with ethanol.

Dynamics of R5020 Induction of FDG Uptake

To define the dynamics of ligand activated PR on FDG uptake, dose titration and time course experiments were performed. Hormone-deprived T47D cells were treated with R5020 (10⁻¹¹ to 10⁻⁷M) for 24 hours. FDG uptake was measured and PR transactivation confirmed with a reporter gene assay (Fig. 3A and 3B). Sigmoidal dose–response curves were observed for both FDG uptake (Fig. 3A) and PR transcriptional activity (Fig. 3B) with half maximal effective concentration values of 0.2 nM (95% CI 0.03-1.6 nM) and 0.07 nM (95% CI 0.04-1.2 nM), respectively. FDG uptake and PR transcriptional activity were then measured in T47D cells treated with a saturating dose of 100 nM R5020 for 0.5, 4, 8, 12, 24, 36, and 48 hours (Fig. 3C and 3D). A significant increase in FDG uptake was observed at 12 hours with a 1.9-fold increase compared with ethanol control and was sustained through 48 hours of treatment. Similar to the results obtained with R5020, treatment of T47D cells with 100 nM progesterone for 0.5 hours did not cause an increase in FDG uptake compared with ethanol (0.8 ± 0.1 fold change vs 1.0 ± 0.09; P = .9).

Figure 3.

Dose titration and time course of FDG uptake and PR transcriptional activation in response to R5020. T47D cells were steroid hormone deprived for 3 days prior to experimentation. Cells were seeded and treated with increasing doses of R5020 (0.01-100 nM) for 24 hours and (A) FDG uptake and (B) reporter gene assays for PR transcriptional activation were performed. Cells were treated with 100 nM R5020 for varying times (0.5-48 hours) and data normalized to the ethanol control. (C) FDG uptake assay and (D) reporter gene assay for PR transcriptional activation were performed. Data represent the mean ± SEM of 3 independent experiments. *P < .05 compared with ethanol.

Progestins Increase FDG Uptake Through PR-A or PR-B

While PR exists in several isoforms, the two predominant isoforms with distinct roles in breast cancer biology are PR-A and PR-B. To determine if one isoform is responsible for the increase in FDG uptake due to PR agonists, T47D cells that stably express either PR-A (T47D YA) or PR-B (T47D YB) were used. Cells were hormone deprived for 3 days and then treated with PR agonists at 100 nM for 24 hours (Fig. 4A). T47D YA cells treated with R5020, MPA, and MA had an increase in FDG uptake compared with ethanol vehicle (7.4 ± 0.5 × 10⁵ CPM/mg protein, P = .02; 8.0 ± 0.6 × 10⁵ CPM/mg protein, P = .01; 7.1 ± 1.2 × 10⁵ CPM/mg protein, P = .03 vs 3.4 ± 0.8 × 10⁵ CPM/mg protein, respectively). Similarly, R5020 (13.0 ± 0.6 × 10⁵ CPM/mg protein, P = .0001), MPA (12.5 ± 1.2 × 10⁵ CPM/mg protein, P = .0002), and MA (13.1 ± 0.5 × 10⁵ CPM/mg protein, P < .0001) increased FDG uptake in T47D YB cells compared with ethanol (5.5 ± 0.8 × 10⁵ CPM/mg protein). PR protein expression in T47D YA and YB cells were also measured (Fig. 4B). T47D YA cells had higher levels of PR protein than T47D YB cells (1.0 ± 0.3 vs 0.2 ± 0.1, respectively, P = .04). Additionally, PR transcriptional activity was measured in T47D YA and YB cells that were hormone deprived and then treated with 10 nM PR agonists for 24 hours (Fig. 4C and 4D). R5020, MPA, and MA caused a significant increase in PR-A transcriptional activation compared with ethanol control in T47D YA cells (81.6 ± 10.5 fold change, P < .0001; 65.1 ± 4.8 fold change, P = .0005; and 64.9 ± 12.5 fold change, P = .0005 vs 1.0 ± 0.2, relatively). PR-B transcriptional activation was also increased in T47D YB cells after R5020 (2772 ± 150.4 fold change, P = 0.0001), MPA (1618 ± 61.4 fold change, P = .007), and MA (1843 ± 465.3 fold change, P = 0.003) compared with ethanol control (1.0 ± 0.1). Thus, either PR isoform can increase FDG uptake in T47D cells treated with PR agonists.

Figure 4.

PR-A and PR-B isoforms increase FDG uptake after agonist treatment. T47D YA and YB cells were steroid hormone deprived for 3 days, seeded, and the next day treated with 100 nM R5020, MPA, MA, or ethanol control for 24 hours. (A) FDG uptake assay was performed and PR protein expression was measured by Western blot analysis (B) in T47D YA and YB cells. Values in B were normalized to the actin loading control then to T47D YA PR expression. PR transcriptional activation was confirmed with a reporter gene assay for T47D YA (C) and T47D YB (D) cells. Data represent the mean ± SEM of 3 independent experiments. *P < .05 compared with ethanol in (A), (C), and (D) or YA in (B).

PR Antagonists Mitigate R5020-Mediated Increase in FDG Uptake

To further investigate the role ligand-activated PR plays in mediating FDG uptake, the antagonists mifepristone (RU486) [52, 53] and ulipristal acetate (UPA; CBD-2914) [54] were used to pharmacologically inhibit PR transcriptional activity. T47D cells were treated with either 100 nM RU486 or UPA in the presence of 1 nM R5020 for 24 hours. Both RU486 and UPA inhibited R5020-mediated increase in FDG uptake compared with R5020 alone (7.6 ± 0.4 × 10⁵ CPM/mg protein, 7.3 ± 0.3 × 10⁵ CPM/mg protein vs 12.7 ± 0.8 × 10⁵ CPM/mg protein, respectively, Fig. 5A). Inhibition of PR transcriptional activity by these antagonists was confirmed using a reporter gene assay (Fig. 5B). Thus, pharmacologic of inhibition of PR transcriptional activity mitigates R5020-mediated increase in FDG uptake.

Figure 5.

Pharmacologic inhibition of PR transcriptional activity prevents R5020-mediated FDG uptake. T47D cells were steroid hormone deprived for 3 days prior to experimentation. Cells were seeded, then cotreated for 24 hours with 100 nM mifepristone (RU486) or ulipristal acetate (UPA) in the presence or absence of 10 nM R5020. (A) FDG uptake and (B) reporter gene assays for PR transcriptional activation were performed. FDG uptake is normalized to ethanol treated cells. Data represents the mean ± SEM of 3 independent experiments. *P < .05 compared with ethanol.

PR Expression Is Required for Increased FDG Uptake After R5020 Treatment

Since pharmacologic inhibition of PR prevents FDG uptake after R5020 treatment, the direct role of PR was evaluated by using T47D PR KO cells and their scrambled control (T47D Scr) [41]. First, PR KO was confirmed by Western blot which showed T47D PR KO cells had no detectable PR-B or PR-A protein (Fig. 6A). To confirm PR transcriptional activity, T47D WT, Scr, and PR KO cells were treated with either ethanol control or 10 nM R5020 for 24 hours and PR transactivation was measured using a reporter gene assay (Fig. 6B). There was a significant increase in PR transcriptional activity after R5020 treatment compared with ethanol in T47D WT (2.2 ± 0.06 vs 0.0004 ± 0.00002, P < .0001) and T47D Scr cells (2.7 ± 0.1 vs 0.0008 ± 0.0002, P < .0001). However, the PR KO cells demonstrated no increase in transcriptional activity in response to R5020. Additionally, the PR-regulated endogenous target gene [7, 55], FKBP5 (FK506 binding protein 5) mRNA expression was measured to confirm PR KO (Fig. 6C). Cells were hormone deprived for 3 days and then treated with 10 nM R5020 for 12 and 24 hours. T47D WT cells had an increase in FKBP5 expression at 12 hours (51.3 ± 0.2 fold change; P < .0001) and at 24 hours (29.6 ± 0.2 fold change; P < .0001) compared with ethanol. Similarly, T47D Scr cells had a significant increase in FKBP5 at 12 hours (67.4 ± 0.7 fold change; P < .0001) and at 24 hours (41.4 ± 0.2 fold change; P < .0001). T47D PR KO cells had no measurable FKBP5 expression. To determine if PR is directly responsible for PR agonists increasing FDG uptake, these cells were treated with 100 nM R5020, MPA, or MA for 24 hours (Fig. 6D). In T47D PR KO cells, FDG uptake did not increase after treatment with R5020, MPA, or MA (0.83 ± 0.14 fold, 0.98 ± 0.12 fold, or 1.0 ± 0.08 fold, respectively). For comparison, treating the T47D Scr cells with R5020, MPA, or MA increased FDG uptake by 2.02 ± 0.11 fold (P = .0002), 1.88 ± 0.10 fold (P = .0006), and 1.84 ± 0.13 fold (P = .0008) compared with ethanol control, which was similar to T47D WT. Taken together, these data suggest that PR is essential for mediating increased FDG uptake in T47D cells.

Figure 6.

Genetic knockout of PR inhibits R5020-mediated FDG uptake. PR protein expression in T47D cells with CRISPR/Cas9 PGR gene knockout (PR KO), their scrambled control (Scr), and wild-type (WT) was confirmed with Western blot analysis (A). Transcriptional activation after 24 hours treatment with 10 nM R5020 or ethanol vehicle control (EtOH) was determined using a reporter gene assay (B) and measurement of an endogenous PR-regulated gene, FKBP5, via qPCR (C). Cells were steroid hormone deprived for 3 days and treated with 100 nM R5020, MPA, MA, or ethanol vehicle for 24 hours and (D) FDG uptake was performed. Data is normalized to ethanol control and expressed as relative fold change. Data represent the mean ± SEM of 3 independent assays for B and D. Data represent the mean ± SD for C. *P < .05 compared with ethanol.

Transcription and Translation Are Required for Ligand-Activated PR Induction of FDG Uptake

To investigate whether new RNA and protein synthesis are required for PR agonist–induced FDG uptake, the inhibitors actinomycin D and cycloheximide were used, respectively. Actinomycin D binds to DNA and inhibits RNA polymerase elongation thereby preventing RNA synthesis. Cycloheximide inhibits protein synthesis by preventing translational elongation. T47D cells were pretreated with actinomycin D (1 μM) or cycloheximide (10 μg/mL) for 1 hour prior to the addition of 100 nM R5020. FDG uptake was measured after 24 hours treatment (Fig. 7A). Both actinomycin D and cycloheximide were able to inhibit R5020-induced FDG uptake (P = .85 and P = .65, respectively). Thus, ligand-activated PR induction of FDG uptake requires both transcription and new protein synthesis.

Figure 7.

Progestin-stimulated gene expression and enzymatic activity of proteins involved in glucose uptake, glycolysis, and the pentose phosphate pathway. T47D cells were steroid hormone deprived for 3 days. (A) Cells were pretreated with 1 μM actinomycin D (Act D) or 10 μg/mL cycloheximide (Chx) for 1 hour and then treated with 100 nM R5020 for 24 hours. FDG uptake was measured and normalized to protein. Data were normalized to vehicle control. T47D WT (B), Scr (C), and PR KO (D) cells were treated with 100 nM R5020 for 12 and 24 hours. Data were normalized to RPLP0, a housekeeping gene, then normalized to each ethanol vehicle control (0 hours). ChIP-qPCR was performed on T47D cells treated for 3 hours with 100 nM R5020 or ethanol control using (E) PR antibody or (F) rabbit immunoglobulin (IgG). Data represents the mean ± SEM of 3 independent assays for A, E, and F. Data represents the mean ± SD for B, C, and D. *P < .05 compared with ethanol.

R5020 Treatment Increases Expression and Activity of Proteins Involved in Glucose Uptake, Glycolysis, and Pentose Phosphate Pathway

T47D WT, Scr, and PR KO cells were hormone deprived for 3 days, then treated with 10 nM R5020 for 12 and 24 hours and mRNA was measured for selected candidate genes involved in glucose uptake, glycolysis, and the pentose phosphate pathway (Fig. 7B-D). After 12 hours of treatment, there was a significant increase in GLUT1 and PFKFB3 mRNA expression compared with ethanol in T47D WT cells (3.5 ± 0.4, P < .0001 and a 4.7 ± 0.7, P < .0001, respectively). GLUT1 mRNA expression remained increased at 24 hours with a 2.2 ± 0.4 fold change (P = .01). GLUT4 expression was also significantly increased at 24 hours (2.2 ± 0.7, P = .009). There was no significant change in expression of GLUT12 or HK2. For T47D Scr cells, R5020 significantly increased GLUT1 expression at 12 hours (3.0 ± 0.5, P = .0012) and PFKFB3 expression at 12 and 24 hours (8.3 ± 0.9, P < .0001 and 4.6 ± 0.8, P < .0001, respectively). However, there was no significant increase in mRNA expression of these genes in T47D PR KO cells.

PR binding sites in the presence of R5020 in T47D cells near these genes were identified using a published ChIP sequencing (ChIP-Seq) data set (GEO accession number GSE68359) [35] (Table 2). GLUT1, PFKFB3, G6PD, and 6PGD all have PR binding sites containing a full PRE located within 100 kb of their respective transcription start sites, suggesting that these genes may be directly regulated by PR in response to R5020. For confirmation, ChIP-qPCR was performed focusing on the PR binding sites upstream of the GLUT1 and PFKFB3 genes identified from the published ChIP-Seq data set. SYNE1 was used as a negative control for PR binding. There was a significant increase in PR protein binding to the SLC2A1 (GLUT1) and PFKFB3 genes in response to R5020 treatment (Fig. 7E and 7F).

Table 2.

PR binding sites near selected genes in the presence of R5020 for T47D cells

Gene	Full PRE (distance to TSS)	Half PRE (distance to TSS)	No PRE
SLC2A1 (GLUT1)	−72 605, −33 300, +11 624, +12 483, +20 008	−2062, +7764, +21 169, +26 266, +27 031, +98 711	None
SLC2A4 (GLUT4)	None	None	None
SLC2A12 (GLUT12)	−50 554	−64 838, −57 179, +3369, +45 402	None
HK2	−72 418, +11 503	None	+32 638
PFKFB3	+26 654, +45 996, +54 553, +94 351	−31 266, −29 290, −26 683, +23 304, +30 665, +51 066, +52 987	None
G6PD	−5005	+4101	None
6PGD	−10 666, +29 900	None	None

The locations of PR binding sites following treatment with R5020 that are within 100 kb of the TSS of selected genes. The presence of a half-PRE (RGnACA or TGTnCY) or full-PRE (RGnACAnrnTGTnCY) sequence in the binding sites is noted. Data were obtained from a previously published dataset using T47D cells (GEO accession number GSE68359 [35]).

Abbreviations: 6PGD, 6-phosphogluconate dehydrogenase; G6PD, glucose-6-phosphate dehydrogenase; HK2, hexokinase 2; PFKFB3, 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3; PRE, progesterone receptor response element; TSS, transcription start site.

Additionally, the activity of G6PD and 6PGD, two enzymes in the pentose phosphate pathway, were measured. R5020 treatment significantly increased G6PD activity compared with ethanol (50.4 ± 1.6 vs 39.4 ± 1.6 mU/protein, P = .007; Table 3). 6PGD activity was also significantly increased in T47D cells after R5020 treatment (6.3 ± 0.08 vs 4.5 ± 0.5 mU/protein, P = .02; Table 3). Taken together, R5020 affected the expression and activity of proteins involved in glucose uptake, glycolysis, and the pentose phosphate pathway.

Table 3.

G6PD and 6PGD enzymatic activity in T47D cells

	G6PD	6PGD
EtOH	37.8 ± 4.7	4.5 ± 0.5
R5020	47.8 ± 5.4*	6.3 ± 0.08*

Values are mU/protein. Data represent the mean ± SEM of 3 individual experiments.

^* P < 0.05 compared with EtOH control.

Abbreviations: 6PGD = 6-phosphogluconate dehydrogenase; G6PD, glucose-6-phosphate dehydrogenase; EtOH, ethanol.

Discussion

The purpose of this study was to evaluate the role of PR in regulating glucose and FDG uptake in breast cancer cells with the future translational goal of using FDG PET imaging for predicting and assessing early response to new PR-targeted endocrine therapy agents in clinical trials. This study focused on T47D breast cancer cells since they express high levels of PR protein and allow the study of PR function in an estrogen-free system, which is important given the independent, confounding effects of estrogen on glycolysis and FDG uptake [56–59]. We report that progesterone and progestins increase FDG uptake in T47D cells, which can be inhibited by pharmacologic antagonists of PR transcriptional activity. Progestin-stimulated FDG uptake is dose dependent (half maximal effective concentration 0.2 nM), requires at least 12 hours of treatment for effect, can be mediated by either PR-B or PR-A, and requires both transcription and translation. Despite lower PR-B expression than PR-A in the cell model used, there was a stronger impact on FDG uptake in PR-B–expressing cells. This phenomenon could be attributed to PR-B being a stronger transactivator than PR-A [4–6]. Progestin-stimulated FDG uptake is independent of GR signaling since treatment with dexamethasone, a GR selective agonist, showed no effect. Additionally, R5020 increased the mRNA expression of GLUT1 and PFKFB3 in T47D WT and Scr cells but not in PR KO cells. PRE sites were identified for both GLUT1 and PFKFB3 suggesting direct transcriptional regulation, which is supported by ChIP-qPCR results confirming PR binding. T47D WT cells also had an increase in G6PD and 6PGD enzymatic activity within the pentose phosphate pathway. Together, these findings support a classic genomic PR signaling mechanism that mediates progesterone and progestin stimulated FDG uptake (Fig. 8).

Figure 8.

Proposed model for PR regulation of glucose and FDG uptake in T47D breast cancer cells. Progesterone receptor (PR) is activated by progesterone and progestins which promotes transcription of GLUT1 and PFKB3, increased glucose transport into the cell via GLUT1, and increased glycolysis through PFKFB3. Activated PR also increases glucose-6-phosphate-dehydrogenase (G6PD) and 6-phosphogluconate dehydrogenase (6PGD) enzymatic activity which promotes increased glucose uptake. ¹⁸F-fluorodeoxyglucose (FDG) is transported into the cell via GLUT1 and is then phosphorylated by hexokinase [39]. Phosphorylated-FDG is not processed further leading to accumulation of FDG within the cell. Phosphofructokinase (PFK), 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3), mifepristone (RU486), and ulipristal acetate (UPA).

The results of our study expand upon previous work initially reporting an effect of progesterone on GLUTs. In the study by Medina et al, progesterone treatment of ZR-75-1 breast cancer cells increased GLUT1 mRNA and protein, GLUT3 mRNA and protein, and GLUT4 protein [60]. GLUT2 mRNA and protein expression were unaffected by progesterone treatment [60]. The direct mechanistic role for PR was not explored in this prior study. Our results also demonstrate an increase in GLUT1 mRNA by progestin treatment using T47D cells and expand existing knowledge by demonstrating a classic genomic PR-signaling mechanism that mediates FDG uptake in part through increased GLUT1 expression. Overexpression of GLUT1 is common in breast cancer [17, 20, 61–63] and correlates positively with FDG uptake [64]. Furthermore, GLUT1 expression has been associated with higher tumor grade [65–67] and reduced patient survival, thus a poor prognostic factor [67–69].

In addition to stimulation of GLUT1 mRNA expression, the results from our study demonstrated an increase in PFKFB3 mRNA expression in response to R5020 treatment in T47D cells. PFKFB3 is a key regulator of glycolysis and is one of four isoenzymes responsible for catalyzing and degrading fructose 2,6-bisphosphate [70]. Fructose 2,6-bisphosphate is an activator of 6-phosphofructo-1-kinase (PFK-1), the most important regulatory enzyme for glycolysis [71, 72]. PFKFB3 has been shown to have a higher kinase to phosphatase ratio that favors glycolytic flux [73]. In addition to promoting glycolysis, PFKFB3 also promotes cell cycle progression, survival, angiogenesis, immunosuppression, and metastasis in cancer [74–77]. Our results agree with previous studies demonstrating that treatment of T47D and MCF-7 cells with the synthetic progestins ORG2058 or norgestrel increased expression of PFKFB3, which was inhibited by the PR antagonist mifepristone [70, 78]. In the study by Novellasdemunt et al, chromatin immunoprecipitation experiments demonstrated recruitment of PR to the DNA region flanking a partial PRE sequence within the PFKFB3 promoter in response to R5020 treatment in T47D cells supporting a classic genomic PR signaling mechanism [78]. The authors also demonstrated rapid, nongenomic progestin signaling through the ERK (extracellular signal–regulated kinase)/RSK (ribosomal S6 kinase) pathway resulting in phosphorylation of PFKFB3 protein and increased enzymatic activity as a dual mechanism for progestin activation of glycolysis in breast cancer cells [78]. A recently published study also showed that progestin treatment shifts metabolism towards glycolysis in T47D cells and a patient-derived cell line, UCD65, which are highly PR-positive [79], consistent with our findings with FDG uptake.

The results of our study also demonstrated that ligand-activated PR increases the enzymatic activity of G6PD and 6PGD within the pentose phosphate pathway, an important pathway in cancer cell metabolism. G6PD is the rate-limiting enzyme for the pentose phosphate pathway, which produces ribose 5-phosphate, for nucleotide biosynthesis, and NADPH, for biosynthesis of fatty acids and for redox homeostasis [80–82]. Upregulation of G6PD expression or enzymatic activity is a common phenotype in many malignancies including breast cancer and may be associated with poor prognosis and endocrine resistance [80, 83, 84]. Our results using T47D breast cancer cells agree with previous studies using MCF-7 breast cancer cells demonstrating that the PR agonist R5020 increases G6PD activity [85] and is inhibited by the antagonist, RU486 [86]. In addition, a PR binding site with a PRE sequence can be found upstream of the G6PD transcription start site in T47D cells treated with R5020 (Table 1).

This work focused on upstream proteins involved in glucose metabolism that could impact FDG uptake and retention for potential clinical translation. FDG PET imaging is FDA approved for clinical use and increased FDG uptake in response to brief treatment with estradiol or tamoxifen (ie, metabolic flare) in patients with advanced breast cancer has been shown to be associated with clinical response to ER-targeted endocrine therapy agents [87–90]. Given the role for PR regulation in glucose uptake, metabolic imaging with FDG PET could be similarly investigated for assessing early treatment response to new anti-progestins in clinical development.

A different type of metabolic imaging uses hyperpolarized carbon-13 (¹³C) magnetic resonance imaging (MRI), which largely focuses on lactate production from hyperpolarized [1-¹³C]-pyruvate occurring further downstream from glucose influx and phosphorylation by hexokinase [91]. Pyruvate is the final product of glycolysis and is converted to lactate by the enzyme lactate dehydrogenase (LDH). Dynamic hyperpolarized ¹³C MRI can be used to visualize and quantify apparent LDH activity via the pyruvate to lactate conversion rate (k_PL) [91]. Increased LDH activity in T47D and MCF-7 breast cancer cells treated with R5020 has been demonstrated, which is likely mediated via classic genomic PR functional activity since the effect was inhibited by RU486 and by inhibitors of transcription and translation [92–95]. Thus, hyperpolarized ¹³C MRI may represent another potential noninvasive imaging approach for assessing in vivo PR functional activity in breast cancer. Hyperpolarized ¹³C MRI of pyruvate metabolism to lactate has been demonstrated in preclinical rodent models of breast cancer and may enable early detection of treatment response and resistance [96–99]. Although still investigational and not widely available, recent pilot studies have demonstrated safety and feasibility of translating hyperpolarized ¹³C MRI for measuring pyruvate metabolism to patients with early-stage breast cancer [100, 101]. With combined PET/MRI scanners now available and validated for breast cancer imaging [102, 103], simultaneous characterization of PR regulation of upstream glucose uptake and hexokinase activity with FDG and downstream regulation of LDH activity via hyperpolarized ¹³C MRI is technically possible.

Collectively, the results of this study suggest that monitoring alterations in FDG uptake and trapping via progestin stimulation or antiprogestin inhibition could provide insight into PR function in breast cancer. These findings provide the rationale for subsequent in vivo testing using tumor xenografts and small animal FDG PET imaging. In conclusion, progesterone and progestins increase FDG uptake in T47D breast cancer cells through the classical action of PR as a ligand-activated transcription factor. Ligand-activated PR ultimately increases expression and activity of proteins involved in glucose uptake, glycolysis, and the pentose phosphate pathway.

Abbreviations

2DG6P

2-deoxyglucose-6-phosphate

6PGD

6-phosphogluconate dehydrogenase

ChIP

chromatin immunoprecipitation

CPM

counts per minute

DMEM

Dulbecco's modified Eagle’s medium

ER

estrogen receptor

FDG

2-deoxy-2-[¹⁸F]-fluoro-D-glucose

GLUT

glucose transporter

GR

glucocorticoid receptor

LDH

lactate dehydrogenase

MA

megestrol acetate

MPA

medroxyprogesterone acetate

PBS

phosphate-buffered saline

PET

positron emission tomography

G6PD

glucose-6-phosphate dehydrogenase

MRI

magnetic resonance imaging

NADPH

nicotinamide adenine dinucleotide phosphate

PR

progesterone receptor

PRE

progesterone response element

PR KO

knockout of the PGR gene

RPMI

Roswell Park Memorial Institute

Scr

scrambled

SDS

sodium dodecyl sulfate

qPCR

quantitative real-time polymerase chain reaction

UPA

ulipristal acetate

Funding

University of Wisconsin Carbone Cancer Center Support Grant P30 CA014520 and the Department of Radiology, University of Wisconsin School of Medicine and Public Health. This work was partially funded by an NIH R01 grant to ETA (CA260140). Funding sources for AMF include the American Cancer Society (RSG-22-015-01-CCB) and the National Cancer Institute (R01 CA272571).

Disclosures

K.S., R.M.R., and E.T.A. have nothing to declare. A.M.F. receives book chapter royalty from Elsevier, Inc. and has served on an advisory board for GE Healthcare. The Department of Radiology receives research support from GE Healthcare.

Data Availability

Original data generated and analyzed during this study are included in this published article or in the data repositories listed in References.