The anti-aging gene KLOTHO is a novel target for epigenetic silencing in human cervical carcinoma

Background Klotho was originally characterized as an anti-aging gene that predisposed Klotho-deficient mice to a premature aging-like syndrome. Recently, KLOTHO was reported to function as a secreted Wnt antagonist and as a tumor suppressor. Epigenetic gene silencing of secreted Wnt antagonists is considered a common event in a wide range of human malignancies. Abnormal activation of the canonical Wnt pathway due to epigenetic deregulation of Wnt antagonists is thought to play a crucial role in cervical tumorigenesis. In this study, we examined epigenetic silencing of KLOTHO in human cervical carcinoma. Results Loss of KLOTHO mRNA was observed in several cervical cancer cell lines and in invasive carcinoma samples, but not during the early, preinvasive phase of primary cervical tumorigenesis. KLOTHO mRNA was restored after treatment with either the DNA demethylating agent 2'-deoxy-5-azacytidine or histone deacetylase inhibitor trichostatin A. Methylation-specific PCR and bisulfite genomic sequencing analysis of the promoter region of KLOTHO revealed CpG hypermethylation in non-KLOTHO-expressing cervical cancer cell lines and in 41% (9/22) of invasive carcinoma cases. Histone deacetylation was also found to be the major epigenetic silencing mechanism for KLOTHO in the SiHa cell line. Ectopic expression of the secreted form of KLOTHO restored anti-Wnt signaling and anti-clonogenic activity in the CaSki cell line including decreased active β-catenin levels, suppression of T-cell factor/β-catenin target genes, such as c-MYC and CCND1, and inhibition of colony growth. Conclusions Epigenetic silencing of KLOTHO may occur during the late phase of cervical tumorigenesis, and consequent functional loss of KLOTHO as the secreted Wnt antagonist may contribute to aberrant activation of the canonical Wnt pathway in cervical carcinoma.


Background
Cervical carcinoma is the second most common cause of cancer death in women worldwide [1]. Several epidemiological studies have indicated that the human papillomavirus (HPV) provides the 'initial hit' in the development of cervical cancer [2]. It has been hypothesized that deregulation of the canonical Wnt signaling pathway is the 'second hit' in the multistep process of cervical carcinogenesis [3,4]. Aberrant activation of the Wnt/βcatenin signaling pathways has been regarded as a generic pathway in a variety of human malignancies [5,6]. Gainof-function mutations centered in the N-terminus of CTNNB1 (encoding the β-catenin gene) induced these oncogenic proteins to be refractory to proteosomal degradation in many tumors. Although cytoplasmic/nuclear accumulation of β-catenin is frequently found in the malignant uterine cervix, this increase in β-catenin is not associated with a mutation in exon 3 of CTNNB1 [7]. Previous studies have revealed that epigenetic gene silencing of soluble Wnt antagonists, such as secreted Frizzledrelated proteins (SFRPs), constitutes one of the major alterations resulting in constitutive activation of the canonical Wnt pathway in many tumors, particularly colorectal [8], bladder [9], gastric [10], and breast cancer [11]. In addition, abnormal CpG island methylation of key tumor suppressor genes has been reported in invasive cervical cancer [12]. We therefore examined whether pro-moter methylation of the genes encoding secreted Wnt antagonists is responsible for stabilization of β-catenin levels in cervical carcinoma.
Klotho has been characterized as systemic anti-aging hormone [13] and was originally identified in mice homozygous for the mutated allele (kl -/-). These mice show a human-like aging-related syndrome and develop multiple disorders such as hypogonadism, ectopic calcification, osteoporosis, skin atrophy, and pulmonary emphysema. Klotho-deficient mice die around 8-9 weeks of age [14]. In contrast, transgenic mice overexpressing Klotho have an extended life span that is 30% longer in males and 20% longer in females [15]. KLOTHO encodes a single-pass type-1 transmembrane protein as well as a truncated, secreted form derived from alternative RNA splicing. The major transcript in humans produces the secreted form, and the extracellular domain of the membrane-bound form is shed and then secreted. These soluble KLOTHO products function as circulating factors that are involved in multiple signaling pathways. Although the molecular mechanism of Klotho in the suppression of aging-related phenotypes is unknown, Klotho was shown to be a secreted antagonist of the Wnt signaling pathway [16]. Klotho inhibits Wnt signaling activity by forming a complex with Wnt3, which is a manner of the SFRP class of secreted Wnt antagonists. Recently, KLOTHO was identified as a potential tumor suppressor that inhibits the IGF-1 pathway and activates the FGF pathway in human breast cancer [17]. In this study, we hypothesized that KLOTHO encoding the secreted Wnt antagonist and that acts as a tumor suppressor may be a candidate target for epigenetic silencing in human cervical carcinoma.

Transcriptional repression of KLOTHO in human cervical cancer
To investigate epigenetic gene silencing of the SFRP class of secreted Wnt antagonists in human cervical carcinoma, we first examined the mRNA levels of SFRP1, SFRP2, SFRP4, SFRP5, and KLOTHO in eight cervical cancer cell lines ( Figure 1A). RT-PCR analysis showed complete loss of SFRP mRNAs in the CaSki, HeLa, and SNU-1299 cell lines and overall transcriptional downregulation of SFRPs in these cervical cancer cell lines. Complete transcriptional silencing of SFRP1, SFRP2 and SFRP5 was recently reported to be associated with promoter methylation in the CaSki and HeLa cell lines [18]. Although the SNU-703 and SNU-1160 cell lines expressed substantial levels of KLOTHO mRNA, loss of expression was found in the CaSki and SNU-1299 cell lines. In contrast with the SiHa and SNU-17 cell lines, which had extremely low levels of KLOTHO mRNA, the HeLa cell line showed a basal level of KLOTHO expres-sion. To examine whether the transcriptional repression of KLOTHO also occurs in primary cervical tumors, we performed RT-PCR for KLOTHO in 4 samples of lowgrade squamous intraepithelial lesion (LSIL), 6 samples of high-grade squamous intraepithelial lesion (HSIL), and 10 samples of invasive carcinoma ( Figure 1B). It was previously reported that the expression of KLOTHO was predominant in the placenta, kidney, small intestine, and prostate [19]. cDNAs prepared from human placenta and normal cervical epithelial tissues were used as positive controls. Four of the ten invasive cervical carcinoma samples showed complete loss of KLOTHO expression, whereas maintenance of KLOTHO mRNA was observed in all of the early, preinvasive lesions (LSILs and HSILs). Thus, KLOTHO is downregulated in human cervical cancer cell lines and in invasive cervical carcinoma.
To determine whether the loss of KLOTHO expression is associated with epigenetic gene silencing, we used RT-PCR to examine transcriptional reactivation in representative cell lines lacking KLOTHO mRNA following treatment with either the DNA methyltransferase (DNMT) inhibitor 2'-deoxy-5-azacytidine (DAC) or the histone deacetylase (HDAC) inhibitor trichostatin A (TSA) (Figure 1C). DNA demethylation by DAC induced significant restoration of KLOTHO mRNA expression in the SNU-1299 cell line, but incomplete upregulation was seen in the SiHa cell line. When each cell line was treated with TSA alone, KLOTHO restoration was detected only in the SiHa cell line. A synergistic effect of DAC and TSA was observed in the SNU-1299 cell line, in which the KLOTHO mRNA was restored only with a DNA demethylating drug. Taken together, these data imply that the downregulation of KLOTHO in cervical cancer cell lines is correlated with epigenetic inactivation mechanisms involving DNA methylation and histone deacetylation.

Promoter CpG hypermethylation of KLOTHO in cervical cancer
To analyze the methylation status of the CpG islands in the KLOTHO promoter region, methylation-specific PCR (MSP) and bisulfite genomic sequencing (BGS) analysis were performed in cervical cancer cell lines. As shown in Figure 2A, a dense GC-rich region was identified around the transcription start site of KLOTHO. Positive controls for unmethylated alleles were verified with MSP using genomic DNA from human placenta. Positive controls for methylated alleles were confirmed by MSP of in vitro methylated DNA (IVD) of the same material. Unmethylated (U) products were predominant in the promoter region of control DNA from human placenta; however, methylated (M) products were exclusively amplified from an IVD template ( Figure 2B    cell lines (Figure 3). Although extensive hypermethylation of the KLOTHO promoter region was prominent in the cell lines that lacked KLOTHO expression (CaSki, SNU-17, and SNU-1299), the KLOTHO-expressing cell lines (SNU-703 and SNU-1160) tended to have completely unmethylated CpG islands. These data strongly suggest that transcriptional repression of KLOTHO is associated with aberrant promoter hypermethylation in cervical cancer cell lines.
We next examined the methylation status of primary cervical tumor samples to investigate whether KLOTHO undergoes abnormal methylation in association with the histopathologic grades of squamous intraepithelial lesions ( Figure 2C). Promoter methylation was not detected in the normal cervical samples (n = 7), the LSIL (n = 30) or the HSIL (n = 37), but extensive methylation was observed exclusively in nine of 22 (41%) of the invasive carcinoma cases. Based on these observations, the number of non-KLOTHO-expressing tissue samples (Figure 1B) may be consistent with the methylation frequency of primary cervical tumors. To confirm the MSP results, we carried out bisulfite sequencing analysis using representative cases, including two normal and two invasive carcinoma samples (Figure 3). The overall methylation of the KLOTHO promoter region in the invasive carcinoma cases was in good agreement with the MSP data, whereas a fully unmethylated pattern was observed in the normal samples. Thus, epigenetic silencing of KLOTHO due to promoter hypermethylation may occur in an invasive carcinoma phase-specific manner during cervical carcinogenesis.

Histone deacetylation is the major epigenetic silencing mechanism for KLOTHO in the SiHa cell line
A chromatin immunoprecipitation (ChIP) assay using anti-acetyl histone H3 (AcH3) and H4 (AcH4) antibodies was performed to determine whether SiHa cells utilize local histone modification as a mechanism of KLOTHO silencing. The immunoprecipitated DNA was analyzed by PCR to elucidate the histone H3 and H4 acetylation level of the KLOTHO promoter region. Figure 2A shows the promoter region around the transcription start site (+1) that was analyzed by the ChIP assay. As shown in Figure 4, marked differences were observed between the acetylation level of histones H3 and H4 in untreated SiHa cells. Although very low levels of PCR products were detected in the control SiHa cells, TSA treatment resulted in a dramatic amplification of specific DNA fragments immunoprecipitated with the AcH3 antibody. This result suggested that the KLOTHO promoter is enriched in deacetylated histone H3, leading to repressive histone modification in the SiHa cell line. On the other hand, the acetylation level of histone H4 seemed to not be associated with restoration of KLOTHO mRNA after treatment with TSA in the SiHa cell line. Suzuki et al. reported that a group of genes lacking promoter CpG methylation tended to be reactivated after HDAC inhibition alone [20]. Our study identified unmethylated CpGs present in the KLOTHO promoter in the SiHa cell line ( Figure 3). Thus, deacetylated histone H3 in the promoter region is strongly correlated with epigenetic silencing of KLOTHO in the SiHa cell line.

KLOTHO functions as a secreted Wnt antagonist in cervical cancer
To assess whether KLOTHO inhibits the Wnt signaling pathway in cervical cancer, the CaSki cell line was transfected with a cDNA expression vector that produces a secreted form of KLOTHO (pcDNA3.1/V5-His-sKL). Ectopic expression of sKL protein resulted in a decrease in total β-catenin levels and a dramatic reduction in the active form of β-catenin (ABC), which is dephosphorylated on S37 or T41 residues ( Figure 5A). The expression of representative T-cell factor (TCF)/β-catenin target genes, including c-MYC and CCND1 (encoding Cyclin D1), was markedly decreased in the CaSki cell line when the overexpression of sKL was induced ( Figure 5B). c-MYC is a specific oncogene that is commonly activated during cervical carcinogenesis [21]. These findings indicate that the canonical Wnt pathway is inhibited by a secreted form of KLOTHO in a cervical cancer cell line. Finally, we performed a colony formation assay to test the tumor-suppressor activity of KLOTHO as a result of inhibition of Wnt/β-catenin signaling. As shown in Figure  5C, CaSki cells transfected with the sKL expression vector showed 57% fewer colonies compared to the empty vector. This result shows that KLOTHO functions as a potential tumor suppressor and that loss of KLOTHO may contribute to cervical carcinogenesis.
Abnormal activation of the canonical Wnt pathway due to epigenetic deregulation of Wnt antagonists is thought to play a crucial role during multi-step cervical tumori- genesis. In addition, several extracellular antagonists of the Wnt pathway, including SFRP1, SFRP2 [18], DKK1 [22], and DKK3 [23], were reported to be downregulated by methylation in cervical cancer. Baylin et al. proposed that epigenetic changes may cause cancer cells to become "addicted" to altered signaling pathways, e.g., the Wnt/βcatenin pathway, during the early stages of tumorigenesis [24]. These early epigenetic events that occur in the secreted Wnt-antagonist gene family may allow these cells, i.e., HPV-infected cervical epithelial cells, to acquire genetic or epigenetic alterations in the same pathway, providing the cell with selective advantages that promote tumor development. In our unpublished studies, frequent epigenetic inactivation of the SFRP gene family was observed during the early, preinvasive phase of cervical carcinogenesis preceding KLOTHO silencing. Our results suggest that loss of KLOTHO during the late phase may be another important step toward invasion of cervical cancer cells with aberrant Wnt signaling. Our current study shows that KLOTHO is frequently downregulated both in cervical cancer cell lines and in invasive carcinomas of the primary cervical tumors and that transcrip- tional repression of KLOTHO is associated with promoter hypermethylation. We show here that epigenetic aberrations of KLOTHO occurred in an invasive carcinoma phase-specific manner and that a loss of function of KLOTHO induced oncogenic activation of TCF/ β-catenin target genes in a cervical carcinoma. Thus, epigenetic aberrations of the secreted Wnt antagonist KLOTHO may contribute to the development of cervical cancer. Our analysis is the first to establish that KLOTHO, the gene for a secreted Wnt-antagonist undergoes epigenetic silencing in human cancers. It remains to be determined whether KLOTHO is epigenetically inactivated in other types of human cancer. Moreover, KLOTHO may be a potential target for the treatment of invasive carcinomas and as a prognostic biomarker of cervical cancer.

Conclusions
This study suggests that transcriptional repression of KLOTHO is correlated with promoter CpG hypermethylation in cervical cancer. Frequent (41%) promoter methylation of KLOTHO occurred in invasive carcinoma but not in normal cervical tissues or during the early, preinvasive phase of primary cervical tumors. Our study also showed that histone deacetylation was responsible for silencing of KLOTHO in the SiHa cell line. Functional loss of KLOTHO due to epigenetic silencing may contribute to aberrant activation of the canonical Wnt pathway in cervical carcinoma.

Cell culture and drug treatment
The human cervical cancer cell line C-33A was obtained from the American Type Culture Collection (Manassas, VA), and CaSki, HeLa, SiHa, SNU-17, SNU-703, SNU-1160, and SNU-1299 were obtained from the Korean Cell Line Bank (Seoul, Korea) [25]. The C-33A and CaSki cell lines were derived from carcinomas, SiHa and all the SNU-series cell lines were derived from squamous cell carcinomas, and the HeLa cell line was derived from an adenocarcinoma. The HPV-negative cell line C-33A, the HPV-16-positive cell line SiHa, and the HPV-18-positive cell line HeLa were grown in Dulbecco's Modified Eagle Medium (Invitrogen) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100 μg/ml streptomycin, and 250 ng/ml amphotericin B. The HPV-16-positive cell lines CaSki, SNU-17, SNU-703, and SNU-1299, and the HPV-18-positive cell line SNU-1160 were grown in RPMI-1640 medium (Invitrogen) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, 100 μg/ml streptomycin, and 250 ng/ml amphotericin B. All cells were maintained at 37°C in a humidified incubator with 5% CO 2 . We treated the cervical cancer cell lines with DAC (Sigma) for 72 h and TSA (Sigma) for 24 h.

Tissue samples and nucleic acid preparation
The uterine cervical tissues were obtained either by punch biopsy or hysterectomy from patients with cervical intraepithelial neoplasm. The tissues were fixed in 10% neutral buffered formalin for 6 to 12 h, and then embedded in paraffin. The histopathologic diagnoses of tissues were made by a pathologist. All patients provided informed consent, and the study was approved by the institutional review board of Soonchunhyang University Cheonan Hospital. After identification of the uterine cervical epithelial lesions on hematoxylin and eosin-stained slides, a pathologist manually dissected the epithelial portions with a 26 gauge needle using a light microscope and added the cells to lysis buffer (10 mM Tris-HCl, pH 8.0; 0.1 mM EDTA, pH 8.0; 2% SDS; 0.15 mg proteinase K). Samples were incubated at 60°C until they were completely lysed, and then genomic DNA was extracted with phenol:chloroform:isoamyl alcohol (25:24:1) and ethanol precipitated at -20°C overnight. The DNA pellet was then dissolved and quantified using a NanoDrop ® ND-1000 Spectrophotometer (Nanodrop Technologies, USA). After the same tissue samples were treated with lysis buffer at 55°C for 3 h, total RNA was extracted using TRIzol Reagent (Invitrogen) according to the manufacturer's protocol.

Reverse transcription-PCR
Total RNA was isolated from cell lines using an RNeasy Mini kit (Qiagen) following the manufacturer's instructions, and cDNA synthesis was performed with 1 μg of the total RNA using the ImProm-II™ Reverse Transcription System (Promega). The primer sequences for RT-PCR of the KLOTHO cDNA were 5'-ACTCCCCCAGT-CAGGTGGCGGTA-3' (forward) and 5'-TGGGCCCGG-GAAACCATTGCTGTC-3' (reverse). All amplifications were carried out with AccuPower PCR PreMix (Bioneer). Human ACTB (encoding β-actin) was amplified as an endogenous control.

Methylation analysis
Genomic DNA was extracted from cell lines using a DNeasy Tissue kit (Qiagen), and bisulfite modification of genomic DNA was performed using an EZ DNA Methylation kit (Zymo Research), following the manufacturers' instructions. Bisulfite-converted genomic DNA was amplified using specific MSP primer sets that discriminate between unmethylated (U) and methylated (M) promoter region of KLOTHO. The sequences of the Uspecific primer set were 5'-AGAGGATGTGTGGTAGG-TAAAGAG-3' (forward) and 5'-ACAAACCAAAAC-TACCTCCACCCT-3' (reverse). The sequences of the Mspecific primer set for the M1 region were 5'-AGAG-GACGCGCGGTAGGTAA-3' (forward) and 5'-ACGAACCGAAACTACCTCCGC-3' (reverse). We treated the human placental DNA with SssI methylase (New England Biolabs) following the manufacturer's protocol for use as positive control for methylation analysis. Bisulfite genomic sequencing was performed to determine the methylation density of the 357 bp fragment containing 54 CpGs at the 5'-end region of KLOTHO. The primer sequences were 5'-GGGAGTTGGGA-GAAATAGGTGT-3' (forward) and 5'-CCAAACCCAA-CAACACCAACAAC-3' (reverse). PCR products were purified and subcloned into the pGEM-T Easy Vector (Promega) for subsequent sequencing analysis.

Chromatin immunoprecipitation analysis
ChIP analysis was carried out using an EZ ChIP kit (Millipore) following the manufacturer's protocol. SiHa cells (4 × 10 5 cells) were plated in a 100-mm culture dish and incubated for 24 h. Formaldehyde-treated cells were resuspended in SDS lysis buffer, and the cell lysates were sheared by sonication. The chromatin fragments were immunoprecipitated with antibodies against acetylated histone H3 (Millipore) or H4 (Millipore), and then the purified DNA was analyzed by PCR using the primer pairs: 5'-AGTCCCGGCTCGCAGGTAATTATTG-3' (forward) and 5'-AGGAGGCCGCGAGAAACGGG-3' (reverse), which were designed to amplify nucleotides -134 to +155 of KLOTHO.

Transfection, immunoblotting, and colony formation assay
The secreted form of human KLOTHO (sKL) cDNA cloned into pcDNA3.1/V5-His expression vector (Invitrogen) was a generous gift from Michael J. Econs (Indiana University School of Medicine, Indianapolis, IN, USA). CaSki cells were plated at 1 × 10 6 cells/60-mm culture dish 24 h prior to transfection and were transfected with 5 μg of either the sKL expression vector or an empty control vector for 5 h in serum-free medium using Lipofectamine 2000 (Invitrogen). After replacing the DNA-Lipofectamine complex-containing medium with complete growth medium, transfected cells were incubated for 72 h. For immunoblotting, 20 μg of protein per lane were separated on polyacrylamide gels and blotted onto PVDF membranes (Millipore). Antibodies against KLOTHO (Santa Cruz), total β-catenin (Cell Signaling), and active β-catenin (Millipore) were used to probe the blot. For the colony formation assay, CaSki cells were plated in 6-well plates after transfection and selected with 1 mg/ml G418 (Invitrogen) for 17 d. Subsequently, colonies were stained with crystal violet, and duplicate wells were counted.