Krüppel-like factor 9 loss-of-expression in human endometrial carcinoma links altered expression of growth-regulatory genes with aberrant proliferative response to estrogen

Abstract

Endometrial cancer is the most commonly diagnosed female genital tract malignancy. Krüppel-like factor 9 (KLF9), a member of the evolutionarily conserved Sp family of transcription factors, is expressed in uterine stroma and glandular epithelium, where it affects cellular proliferation, differentiation, and apoptosis. Deregulated expression of a number of Sp proteins has been associated with multiple types of human tumors, but a role for KLF9 in endometrial cancer development and/or progression is unknown. Here, we evaluated KLF9 expression in endometrial tumors and adjacent uninvolved endometrium of women with endometrial carcinoma. KLF9 mRNA and protein levels were lower in endometrial tumors coincident with decreased expression of family member KLF4 and growth-regulators FBJ murine osteosarcoma viral oncogene homolog (FOS) and myelocytomatosis viral oncogene homolog (MYC) and with increased expression of telomerase reverse transcriptase (TERT) and the chromatin-modifying enzymes DNA methyltransferase 1 (DNMT1) and histone deacetylase 3 (HDAC3). Expression of estrogen receptor alpha (ESR1) and the tumor-suppressor phosphatase and tensin homolog deleted in chromosome 10 (PTEN) did not differ between tumor and normal tissue. The functional relevance of attenuated KLF9 expression in endometrial carcinogenesis was further evaluated in the human endometrial carcinoma cell line Ishikawa by siRNA targeting. KLF9 depletion resulted in loss of normal cellular response to the proliferative effects of estrogen concomitant with reductions in KLF4 and MYC and with enhancement of TERT and ESR1 gene expression. Silencing of KLF4 did not mimic the effects of silencing KLF9 in Ishikawa cells. We suggest that KLF9 loss-of-expression accompanying endometrial carcinogenesis may predispose endometrial epithelial cells to mechanisms of escape from estrogen-mediated growth regulation, leading to progression of established neoplasms.

FIG. 1

Analyses of KLF9 expression in human EC tissues. A) Representative H&E-stained sections of adjacent NE and EC tissues from two patients undergoing hysterectomy after diagnosis with EC. GE, glandular epithelium; LE, luminal epithelium; ST, stroma. Original magnification ×10. B) Messenger RNA transcript levels of KLF9 and KLF family members were quantified by QPCR and normalized to those of TATA box-binding protein (TBP). Data (mean ± SEM) are expressed as the fold-change relative to corresponding NE (n = 13 samples/tissue type). C) Western blot analyses of nuclear proteins isolated from NE and EC tissues using rabbit anti-rat KLF9 and anti-GAPDH (loading control) antibodies. A representative blot (left) with the quantified data from densitometric scans of normalized immunoreactive bands from NE and EC samples pooled from individual patients (right) is shown. *P < 0.05 (Student t-test).

FIG. 2

The expression of growth-associated genes in EC versus NE. The expression levels of ESR1, FOS, MYC, TERT, PTEN, CDH1, and VIM (A) and of DNMT1, HDAC3, HDAC6, and HDAC9 (B) were quantified in NE and EC tissues by QPCR. Data (mean ± SEM, n = 13 samples/tissue type) were normalized to TATA box-binding protein (TBP) and renormalized to values for NE. *P < 0.05 (Student t-test).

FIG. 3

Altered gene expression upon KLF9 gene silencing in human Ishikawa cells treated with E₂. The transcript levels of KLF9 and other genes for which expression were altered in EC versus NE (see Fig. 2) were quantified in Ishikawa cells transfected with either control (Scr) siRNA or siRNA directed against KLF9 and treated with vehicle (0.1% DMSO in treatment medium; data not shown) or 100 nM E₂ in vehicle for 72 h. Data (mean ± SEM) were normalized to TBP and are expressed as the fold-change relative to vehicle-treated, scRNA-transfected controls (n = 4 individual wells/transfection). Inset) KLF9 knockdown resulted in more than 90% loss of KLF9 protein levels. *P < 0.05 (Student t-test).

FIG. 4

Proliferative and apoptotic responses of human Ishikawa cells to KLF9 gene silencing. A) Proliferation status of Ishikawa cells transfected with either Scr or KLF9 siRNAs and treated with vehicle (0.1% DMSO in treatment medium) or 100 nM E₂ was measured using the MTT assay after 3 days. Data (mean ± SEM) are from five independent experiments. B) Apoptotic status of Ishikawa cells transfected with either Scr or KLF9 siRNAs and treated with vehicle (0.1% DMSO in treatment medium) or 100 nM E₂ for 72 h was measured using the caspase 3/7-Luciferase assay kit and expressed as relative light units (RLUs). Data (mean ± SEM) are from five independent experiments. Means with different letters differed significantly (P < 0.05, two-way ANOVA followed by Tukey test).

FIG. 5

Analyses of gene expression as a function of KLF4 gene silencing in Ishikawa cells treated with E₂. Expression of KLF4, MYC, and TERT in Ishikawa cells transfected with either Scr or KLF4 siRNAs and treated with vehicle (0.1% DMSO in treatment medium; data not shown) or 100 nM E₂ for 72 h was quantified by QPCR. Data (mean ± SEM) were normalized to TBP and are expressed as the fold-change relative to vehicle-treated, scRNA-transfected controls (n = 4 individual transfections/treatment). *P < 0.05 (Student t-test).