Research Article: Estrogen and Progesterone Regulate p27kip1 Levels via the Ubiquitin-Proteasome System: Pathogenic and Therapeutic Implications for Endometrial Cancer

Date Published: September 27, 2012

Publisher: Public Library of Science

Author(s): Kuang-Tzu Huang, Savvas C. Pavlides, Jon Lecanda, Stephanie V. Blank, Khushbakhat R. Mittal, Leslie I. Gold, Irina Agoulnik.


The levels of proteins that control the cell cycle are regulated by ubiquitin-mediated degradation via the ubiquitin-proteasome system (UPS) by substrate-specific E3 ubiquitin ligases. The cyclin-dependent kinase inhibitor, p27kip1 (p27), that blocks the cell cycle in G1, is ubiquitylated by the E3 ligase SCF-Skp2/Cks1 for degradation by the UPS. In turn, Skp2 and Cks1 are ubiquitylated by the E3 ligase complex APC/Cdh1 for destruction thereby maintaining abundant levels of nuclear p27. We previously showed that perpetual proteasomal degradation of p27 is an early event in Type I endometrial carcinogenesis (ECA), an estrogen (E2)-induced cancer. The present studies demonstrate that E2 stimulates growth of ECA cell lines and normal primary endometrial epithelial cells (EECs) and induces MAPK-ERK1/2-dependent phosphorylation of p27 on Thr187, a prerequisite for p27 ubiquitylation by nuclear SCF-Skp2/Cks1 and subsequent degradation. In addition, E2 decreases the E3 ligase [APC]Cdh1 leaving Skp2 and Cks1 intact to cause p27 degradation. Furthermore, knocking-down Skp2 prevents E2-induced p27 degradation and growth stimulation suggesting that the pathogenesis of E2-induced ECA is dependent on Skp2-mediated degradation of p27. Conversely, progesterone (Pg) as an inhibitor of endometrial proliferation increases nuclear p27 and Cdh1 in primary EECs and ECA cells. Pg, also increases Cdh1 binding to APC to form the active E3ligase. Knocking-down Cdh1 obviates Pg-induced stabilization of p27 and growth inhibition. Notably, neither E2 nor Pg affected transcription of Cdh1, Skp2, Cks1 nor p27. These studies provide new insights into hormone regulation of cell proliferation through the UPS. The data implicates that preventing nuclear p27 degradation by blocking Skp2/Cks1-mediated degradation of p27 or increasing Cdh1 to mediate degradation of Skp2-Cks1 are potential strategies for the prevention and treatment of ECA.

Partial Text

Estrogen (E2) stimulates proliferation of the endometrium and progesterone (Pg) suppresses E2-driven proliferation. Aligned with the effects of these hormones on growth, E2 induces type I endometrial carcinoma (ECA; rate: 85% of all ECAs) and conversely, Pg is used as a therapeutic agent for endometrial hyperplasia, the precursor to ECA [1]. ECA is the most common gynecological malignancy with an incidence of 136,000 global cases per year [2]. At least 50% of women with endometrial atypical hyperplasia (AEH) have concurrent ECA; an additional 30% will progress to ECA [3]. As an alternative to hysterectomy, progestins reverse AEH and well-differentiated ECA leading to a high rate of successful pregnancies [4], [5]. A molecular level understanding of normal and malignant growth regulation of the endometrium by E2 and Pg is important to advance the field in terms of defining novel preventative and therapeutic molecular targets for this disease. We previously reported that the cyclin-dependent kinase (Cdk) inhibitor, p27kip1 (p27) critical to growth arrest, is absent in the glands of both AEH and ECA tissue due to rapid and perpetual degradation of p27 via the ubiquitin proteasome system (UPS) implicating loss of p27 occurs early in the oncogenesis of ECA [6]. Aligned with the opposing effects of E2 and Pg on proliferation, we further showed that E2 caused proteasomal degradation of p27 in primary EECs whereas Pg markedly increased p27 in both primary endometrial epithelial cells (EECs) and ECA cells. These data suggest that p27 is a significant molecular target involved in both the pathogenesis and treatment of ECA.

The present studies provide new insights into hormone regulation of cell proliferation. Specifically, we describe novel mechanisms by which E2 and Pg have opposite effects on endometrial growth through the ubiquitin-proteasome system to regulate the levels of nuclear p27, important for G1 arrest (Figure 8). We conclude that: 1) the levels of the Cdh1 component of the APC/Cdh1 and p27 vary directly and are inverse to the Skp2 and Cks1 components of the SCF-Skp2/Cks1 in response to E2 and Pg with stimulatory and inhibitory effects, respectively; 2) the regulation of Cdh1, Skp2, and Cks1 of the UPS and p27 is not at the level of transcription; 3) the ECA cell lines used herein are appropriate paradigms for this study since their responses to E2 and Pg were identical to primary EECs. Specifically, we show that E2 has two major effects: 1) induces MAPK/ERK2-dependent phosphorylation of p27 on T187, which only occurs in the nucleus and thus, targets the ubiquitylation and degradation of p27 by SCF-Skp2/Cks1 shown to take place in this subcellular compartment [7], [9], [30] 2) causes a decrease in Cdh1 to prevent degradation of Skp2 and Cks1; this raises the levels of Skp2/Cks1 causing degradation of p-p27(T187) for stimulation of cell proliferation. The decrease in p27 stability induced by E2 is underscored by a 1.5 h faster turn-over rate at a 6 h time point following treatment with CHX, which was commensurate with an increase in Skp2 by 1.3 fold. Conversely, Pg has the full opposite effect as it increases Cdh1 and its binding to the APC/Cdh1 for degradation of Skp2 and Cks1 thus, leaving nuclear p27 intact for G1 arrest. To substantiate this effect, Pg increased the stability of p27 protein as the levels were increased by 1.5-fold at 6 h after treatment with CHX whereas Skp2 turn-over was decreased by 6.2 h. The decrease in Skp2 was at least, in part, due to the degradation of Skp2 by the Pg-induced increase in Cdh1. The effects of E2 and Pg on the levels of p27, Skp2, Cks1, and Cdh1 are dose and time-dependent and are consistent with their expected effect on cell cycle distribution. Importantly, knocking-down Skp2 completely obviates both E2-induced degradation of nuclear p27 and stimulation of cell proliferation. Therefore, E2-induced Skp2 E3 ligase activity is required for nuclear p27 degradation and involves a functional interaction between Skp2/Cks1 and p27. These data provide proof of principle for the role of E2 in the pathogenesis of ECA and explain the action of progestins (e.g., Megace® etc) as successful therapeutic agents for AEH and ECA in vivo [4], [5]. Studies show that an increase in nuclear p27 predicts positive outcomes from Pg therapy [20], [34] underscoring p27 in the physiological regulation of endometrial growth and as a significant molecular target for ECA. As perpetual proteasomal degradation of p27 is an early event in ECA oncogenesis [6], the current studies implicate that blocking proteasomal degradation of nuclear p27 to regain growth control is a rational preventative and therapeutic approach to ECA particularly, in patients lacking PR. However, p27 can be mislocalized to the cytoplasm in ECA [18], [19] and other human cancers thereby abolishing its function as a Cdk inhibitor [7], [15], [17], [35]. Phosphorylation of p27 on Ser 10 by kinase interacting stathmin (KIS) causes its binding to CRM-1 for exportation to the cytoplasm [7], [12], [17]. In addition, a high frequency of PTEN inactivating mutations in type I ECA (over 50%) leads to increased Akt activity [1]; Akt phosphorylates p27 on Thr157 [13], [36] blocking its nuclear import. Interestingly, PTEN and p27 are both decreased with a concomitant increase in Skp2 [37], [38], [39]. In the cytoplasm, p27 can be further phosphorylated on, T198 by AGC kinase, downstream from PI3K-Akt stabilizing the molecule in this cellular compartment [14]. Cytoplasmic p27 represses RhoA thereby affecting cytoskeleton dynamics with effects on migration and metastasis [7], [16], [17]. As we show an increase in p27 mainly in the nucleus in response to Pg, Pg therapy might increase p27 shuttling into the nucleus.