O-GlcNAc-Dependent Regulation of Progesterone Receptor Function in Breast Cancer

Abstract

Emerging clinical trial data implicate progestins in the development of breast cancer. While the role for the progesterone receptor (PR) in this process remains controversial, it is clear that PR, a steroid-activated nuclear receptor, alters the transcriptional landscape of breast cancer. PR interacts with many different types of proteins, including transcriptional co-activators and co-repressors, transcription factors, nuclear receptors, and proteins that post-translationally modify PR (i.e., kinases and phosphatases). Herein, we identify a novel interaction between PR and O-GlcNAc transferase (OGT), the enzyme that catalyzes the addition of a single N-acetylglucosamine sugar, referred to as O-GlcNAc, to acceptor serines and threonines in target proteins. This interaction between PR and OGT leads to the post-translational modification of PR by O-GlcNAc. Moreover, we show that O-GlcNAcylated PR is more transcriptionally active on PR-target genes, despite the observation that PR messenger RNA and protein levels are decreased when O-GlcNAc levels are high. O-GlcNAcylation in breast cancer is clinically relevant, as we show that O-GlcNAc levels are higher in breast cancer as compared to matched normal tissues, and PR-positive breast cancers have higher levels of OGT. These data predict that under conditions where O-GlcNAc levels are high (breast cancer), PR, through an interaction with the modifying enzyme OGT, will exhibit increased O-GlcNAcylation and potentiated transcriptional activity. Therapeutic strategies aimed at altering cellular O-GlcNAc levels may have profound effects on PR transcriptional activity in breast cancer.

Fig. 1

PR interacts with OGT. a STRING network of the top 60 PR interactors identified in RIME experiments. The size of the node increases proportionally to the number of identified peptides, and thick edges denote high confidence STRING interactions (0.7–0.99). RIME was performed in T47D-YB cells on two biological replicates, and proteins identified in PR-null cells were not included in the PR-specific interactor list. b Sequence coverage of PR and OGT in both replicate RIME experiments. Green highlights high confidence peptides at FDR < 1%. PR and OGT have been identified by 27 and 12 unique peptides, respectively. c OGT was immunoprecipitated from T47D-YB cell lysates (+/− R5020), and the resulting associated protein complexes were analyzed by Western blotting. Right panels represent total input cell lysates (color figure online)

Fig. 2

PR is post-translationally modified by O-GlcNAc. a PR was immunoprecipitated from T47D-co cell lysates pre-treated with TMG for 18 h, followed by R5020 treatment for 60 min. The resulting associated protein complexes were analyzed for O-GlcNAcylation by Western blotting. b Experiments were performed as in a, but the following T47D cell line variants were used: T47D-YB, T47D-YA, T47D-S79/81A, T47D-co, and T47D-Y (PR-null). All cells were pre-treated with TMG for 18 h. c Experiments were performed as in b in MCF7 cells. All experiments were performed in triplicate, and representative experiments are shown here

Fig. 3

PR levels are lower in TMG-treated cells. a T47D-co cells were continually passed in the presence of TMG (or vehicle) for 3 weeks, followed by R5020 or vehicle (EtOH) for 1 h. Nuclear, cytoplasmic, and whole cell lysates were separated and blotted with antibodies to PR and tubulin (cytoplasmic marker and loading control). b Cells were treated with TMG as in a, and RNA was isolated after 3 weeks of TMG treatment. Isolated RNA was analyzed for PR. Gene values were normalized to an internal control (β-actin). Error bars represent standard deviation between biological triplicates. Asterisks represent statistical significance between the vehicle and TMG-treated groups; p < 0.05, as determined using an unpaired Student’s t test. The experiments shown here were performed in triplicate, and a representative experiment is shown here

Fig. 4

PR-target genes are differentially regulated by O-GlcNAcylated PR. Starved T47D-co cells were continually passed in the presence of TMG (or vehicle) for 3 weeks, followed by R5020 or vehicle (EtOH) for 6 h. Isolated RNA was analyzed for select genes. Gene values were normalized to an internal control (β-actin). Error bars represent standard deviation between biological triplicates. Asterisks represent statistical significance; p < 0.05, as determined using an unpaired Student’s t test. This experiment was performed in triplicate, and a representative experiment is shown here

Fig. 5

O-GlcNAc levels are high in breast cancer. a Tissue microarray (TMA) analysis was performed using immunohistochemical staining with an O-GlcNAc antibody. Select benign and normal cases from the same patient are shown here at × 20 magnification. b Staining intensity changes between normal and cancer are shown for each patient in the TMA for which matched benign and cancer tissue was collected (n = 35). Data is shown in bar graph format; each bar represents an individual patient. Blue bars (top) represent an increase in staining intensity when comparing cancer to benign; red bars represent a decrease in staining intensity when comparing cancer to benign. No bar indicates unchanged. The Wilcoxon signed rank test was used to compare O-GlcNAc levels between matched benign and cancer samples, as described in the “Methods” section (color figure online)

Fig. 6

OGT levels are higher in PR-positive breast cancers. Tissue microarray analysis was performed using immunohistochemical staining with PR and OGT antibodies. Select PR-negative (left) and PR-positive (right) breast cancer (BrCa) cases are shown here at × 20 magnification. The Wilcoxon rank sum test was used to compare OGT levels in PR-positive vs PR-negative breast cancer, as described in the “Methods” section