ZNF367 Inhibits Cancer Progression and Is Targeted by miR-195

Background Several members of the zinc finger protein family have been recently shown to have a role in cancer initiation and progression. Zinc finger protein 367 (ZNF367) is a member of the zinc finger protein family and is expressed in embryonic or fetal erythroid tissue but is absent in normal adult tissue. Methodology/Principal Findings We show that ZNF367 is overexpressed in adrenocortical carcinoma, malignant pheochromocytoma/paraganglioma and thyroid cancer as compared to normal tissue and benign tumors. Using both functional knockdown and ectopic overexpression in multiple cell lines, we show that ZNF367 inhibits cellular proliferation, invasion, migration, and adhesion to extracellular proteins in vitro and in vivo. Integrated gene and microRNA expression analyses showed an inverse correlation between ZNF367 and miR-195 expression. Luciferase assays demonstrated that miR-195 directly regulates ZNF367 expression and that miR-195 regulates cellular invasion. Moreover, integrin alpha 3 (ITGA3) expression was regulated by ZNF367. Conclusions/Significance Our findings taken together suggest that ZNF367 regulates cancer progression.


Introduction
Our knowledge of the biologic events that inhibit or promote the metastatic spread of endocrine cancers locally and to distant sites is limited [1]. Understanding the molecular events involved in cancer progression has important implications for identifying targets for therapeutic benefit. Genomic, genetic, and epigenetic approaches have been used to identify changes in genes/pathways, transcription factors, or gene signatures by comparing normal, primary tumors, and/or metastasis. However, these analyses have not often distinguished between promoters, inhibitors, or passenger markers in cancer initiation and progression, and rarely define a specific mechanism for dysregulated gene expression.
Endocrine cancers are a diverse group of malignancies that exhibit the full spectrum of biologic behavior of malignant tumors: indolent growth to rapidly progressive cancers with poor survival. Thus, endocrine cancers represent an excellent model for studying the molecular factors that influence cancer initiation and progression. Identifying genetic changes common to these diversely behaving endocrine tumor types has been shown to play an important role in many types of human cancers. For example, BRAF mutations are highly prevalent in thyroid cancer and other cancers, and are associated with more aggressive disease in thyroid cancer [2,3]. SDHB mutations were initially identified in familial paraganglioma, but were later found to be prevalent in kidney cancers and gastrointestinal stromal tumors, and recently pituitary tumors [2,4,5]. Thus, discovering the molecular changes associated with endocrine cancer initiation and/or progression is likely to play an important role in a broad group of human malignancies.
In this study, we determined the mechanism of gene expression regulation and function of ZNF367 in a variety of endocrine cancers (papillary thyroid cancer, adrenocortical carcinoma, pheochromocytoma/paraganglioma). ZNF367 (also known as ZFF29 and CDC14B) is a member of the zinc finger protein family, with a unique Cys2His2 zinc finger motif, is expressed in embryonic or fetal erythroid tissue, and is absent in other normal adult tissue [6,7]. Recently, several zinc finger proteins have been found to be dysregulated in cancer, to function as tumor suppressors/promoters, and to cause resistance to chemotherapy [8][9][10]. However, the role of ZNF367 in cancer has not been investigated. Here, we report that ZNF367 is overexpressed in a variety of endocrine cancers and that it inhibits in vitro and in vivo growth, cellular invasion, migration, and adhesion. Moreover, ZNF367 overexpression is associated with the loss of miR-195 expression, which directly targets ZNF367. Lastly, ITGA3 is decreased with ZNF367 overexpression, establishing a miR-195-ZNF367-ITGA3 axis that functions to inhibit cancer progression.

Tissue samples and animal studies
Tissue samples were procured on an Institutional Review Board-approved clinical protocol after obtaining written informed consent (ClinicalTrials.gov Identifier: NCT01005654). The tissue samples were collected at the time of surgery, and they were immediately snap frozen and stored at 280uC. One-hundred-andfive human adrenocortical (19 normal adrenal cortex, 79 cortical adenomas, 7 adrenocortical carcinomas), 47 thyroid tissue (8 normal, 39 papillary thyroid cancer), and 68 (19 normal adrenal medulla, 28 benign, 21 malignant pheochromocytoma/paraganglioma) tissue samples were used in this study. All diagnosis was confirmed by an endocrine pathologist, and tumor samples were confirmed to contain $ 80% tumor cells/nuclei. The diagnosis of adrenocortical carcinoma and malignant pheochromocytoma/ paraganglioma was based on the presence of gross local invasion and/or distant metastasis and a Weiss score of . 3 for adrenocortical carcinoma. Normal adrenal cortex and medulla were collected from healthy organ donors using laser capture microdissection under an Institutional Review Board-approved protocol. Two publicly available datasets of genome-wide gene expression analysis in adrenocortical tumor samples were used to evaluate ZNF367 mRNA expression (http://www.ebi.ac.uk/ arrayexpress/experiments/E-GEOD-10927 and http://www.ebi. ac.uk/arrayexpress/experiments/E-TABM-311).
The National Cancer Institute Animal Care and Use Committee approved the protocols for animal care and handling in the present study. Any mouse experiencing significantly abnormal neurological signs, bleeding from any orifice, impaired mobility, rapid weight loss, debilitating diarrhea, rough hair coat, hunched posture, labored breathing, lethargy, persistent recumbence, jaundice, anemia, self-induced trauma, becomes moribund or otherwise becomes unable to obtain food or water, or with a tumor 2 cm or greater in diameter has been immediately euthanized by CO 2 chamber.

Immunohistochemistry
Formalin-fixed and paraffin-embedded tissue sections were deparaffinized and then rehydrated using xylene and ethanol. Antigen retrieval was completed using a 10% citrate buffer pH 6.0 (Thermo Scientific, Lafayette, CO) in a pressure cooker at 120uC for 10 minutes. Tissue sections were incubated with 6% hydrogen peroxide to quench endogenous peroxidase activity for 30 minutes (Dako, Carpinteria, CA), followed by incubation with serum for an hour (Dako, Carpinteria, CA). Primary anti-ZNF367 rabbit polyclonal antibody (Sigma, St. Louis, MO; HPA015785) at 5 mg/ml dilution was used overnight at 4uC. Anti-rabbit secondary antibody was used (Dako Envision anti-rabbit, Carpinteria, CA) for 1 hour at room temperature. Sections were developed using 3,39-diaminobenzidine DAB as the chromogen (Dako, Carpinteria, CA), and they were counterstained with hematoxylin. The sections were dehydrated and mounted with vectamount mounting medium (Vector Laboratories, Burlingame, CA). The immunostaining of ZNF367 was evaluated by light microscopy (Nikon, Tokyo, Japan), and images were scanned at 20X magnification.
RNA preparation, reverse transcription, and real-time quantitative PCR RNA was isolated using the TRIzol reagent, according to the manufacturer's instructions (Invitrogen Inc., Carlsbad, CA). Total RNA (200-500 ng) was reverse-transcribed using a High Capacity Reverse Transcription cDNA kit, and cDNA was amplified according to the manufacturer's instructions (Applied Biosystems, Foster City, CA). The PCR primers and probes for ZNF367 (Hs00400665_m1), ITGA3 (Hs01076873_m1), and GAPDH (Hs_99999905_m1) were obtained from Applied Biosystems.

Cell proliferation
Cells were seeded at a concentration of 2,000 cells in a 96-well plate in six replicates. The CyQUANT TM assay kit (Invitrogen) was used to evaluate the cell number, according to the manufacturer's instructions.

Clonogenic assay
HEK293 cells (8610 5 ) were transfected with the empty vector or ZNF367 construct. After 24 hours, cells were trypsinized, re-suspended in the media, and counted. The cells were re-seeded in 6-well plates at 250, 500, and 1,000 cells. After 10 days, media was removed from the wells and washed twice with ice-cold PBS. The colonies were fixed with 4% paraformaldehyde for 20 minutes and stained with 2 ml of crystal violet for 60 minutes on a rocking platform. The plates were washed three times with PBS and airdried, and the colonies were photographed in FluorChem Imager (San Jose, CA).

Cell invasion and migration assay
Cellular invasion and migration were assessed using the BD BioCoat Matrigel Invasion Chamber (BD Biosciences) according to the manufacturer's protocol. 1610 5 cells were seeded onto the inserts (8-mM pore-sized polycarbonate membranes) with and without a thin layer of Matrigel Basement Membrane Matrix (BD Biosciences) coat. The inserts were placed into bottom wells, with 10% serum-containing culture medium as a chemoattractant. The plates were incubated for 48 hours at 37uC. Cells that invaded the Matrigel matrix or that migrated through the pores without Matrigel matrix to the lower surface of the membrane were fixed and stained with Diff-Quik (Dade Behring, Newark, NJ) and counted under a light microscope in four separate fields with Image J software (NIH, Bethesda, MD).

Cell adhesion assay
SW13 cells were transfected with ZNF367 siRNA and the negative control. After 120 hours of transfection, cells were trypsinized, and 1610 5 cells were plated in a 48-well plate containing five adhesion molecules (Fibronectin, Collagen I, Collagen IV, Laminin I, and Fibrinogen) and incubated at 37uC for 90 minutes according to the manufacturer's instructions (CytoSelect, Cell Biolabs, Inc., San Diego, CA). The wells were washed twice with PBS, and 200 ml of the staining solution was added and incubated for 10 minutes. The wells were washed twice with deionized water. The wells were air-dried, and 200 ml of the extraction solution was added and incubated for 10 minutes. A total of 150 mL was transferred from each extracted sample to a 96-well plate, and absorbance measured at 560 nm on a SpectraMax M5e microplate reader (Sunnyvale, CA).
Genome-wide mRNA expression microarray SW13 cells were transfected with ZNF367 siRNA and the negative control in triplicate. Following transfection, total RNA was extracted using the RNeasy (Qiagen, Valencia, CA) kit, according to the manufacturer's instructions. cDNA reverse transcription, synthesis, amplification, fragmentation, and terminal labeling of 150 ng total RNA were performed using the GeneChip WT Sense Target Labeling and Control Reagents (Affymetrix, Santa Clara, CA). A total of 25 ng/mL of cDNA was hybridized to the Affymetrix Human Gene 1.0 ST Array GeneChip. Pathway analysis was performed using the DAVID bioinformatics resources (http://david.abcc.ncifcrf.gov, Frederick, MD).

In silico identification of microRNA targeting ZNF367
MiR databases (Target Scan (http://www.targetscan.org) and mirDB (http://mirdb.org/miRDB/) were used to identify micro-RNA that may target ZNF367. Candidate microRNAs predicted to target ZNF367 were then analyzed to determine if they were differentially expressed in adrenocortical carcioma and papillary thyroid cancer.

Luciferase assay
Wild-type ZNF367 39UTR was cloned into the GoClone construct (Switchgear Genomics, Menlo Park, CA). The mutant construct of ZNF367 39UTR was obtained by introducing the mutation into the first three nucleotides of the seed region (143-150, GCTGCTA --CGAGCTA) for miR-195. Wild-type ZNF367 39UTR or mutant ZNF367 39UTR and the empty 39UTR vector with pre-miR-195 or pre-NC were co-transfected into SW13 cells (Switchgear Genomics). The normalization method for transfection efficiency in luciferase reporter assay was according to the recommendation of SwitchGear Genomics (Menlo Park, CA) with empty vector used as positive control for the transfection. The empty 39UTR vector contains a constitutive promoter (found on all UTR constructs) and the luciferase gene (RenSP). This construct served as a positive control for the transfection. Cells were plated in 24-well plates and co-transfected with 5 nM of miR-195 or the negative control (Applied Biosystems), 0.12 mg of the vector (SwitchGear Genomics), and 0.75 ml of Lipofectamine (Invitrogen). The luminescence was read after 24 hours with the Light Switch assay system, using a SpectraMax M5e microplate reader.

In vivo xenograft assay
Athymic nude female mice (five to six weeks old; body weight: 20-22 g) were obtained from the Frederick National Laboratory for Cancer Research animal facilities (Frederick, MD). After 48 hours of transfection with ZNF367 siRNA and the negative control, three million cells were re-suspended in 100 ml of DMEM and Matrigel (1:1), and they were injected into the flank of athymic nude mice.

Statistical analyses
Continuous data is presented as mean 6 standard deviation (SD) or standard error of mean (SEM). The Kruskal-Wallis test was used to compare nonparametric data from three or more groups. The Student's t-test or Mann-Whitney test was used to compare differences between two groups for parametric and nonparametric variables, respectively. The Pearson correlation test was used to identify correlations between two groups. A p-value , 0.05 was considered statistically significant. Statistical analysis was completed using Graph Pad Prism 5.0 statistical software.

ZNF367 is overexpressed in cancer
ZNF367 is overexpressed in adrenocortical carcinoma, papillary thyroid cancer, and malignant pheochromocytoma/paraganglioma, compared to benign and normal tissue samples for each tumor type (p,0.05; Figure 1). ZNF367 protein expression was present in the nucleus and cytoplasm. There was stronger ZNF367 staining in the nucleus than in the cytoplasm in each cancer sample (adrenocortical cancer, papillary thyroid cancer and malignant pheochromocytoma/paraganglioma). To confirm the elevated expression of ZNF367 in adrenocortical carcinoma in a larger sample set, we analyzed expression profiling data from publicly available datasets deposited in gene expression omnibus. In two independent genome-wide gene expression data sets, ZNF367 was overexpressed in adrenocortical carcinoma compared to adrenal cortical adenoma and normal tissue samples (p, 0.001; Figure S1). These findings suggest that ZNF367 is overexpressed in a variety of cancers, with consistent results across different analysis and genomic platforms used.

ZNF367 inhibits cellular proliferation, invasion, migration, and adhesion
Because ZNF367 is overexpressed in endocrine cancer, we investigated its effect on cellular proliferation, invasion, and migration, events that are necessary for cancer progression. ZNF367 was expressed in adrenocortical carcinoma (SW13, BD140A), papillary thyroid cancer (TPC-1), and HEK293 cell lines. Thus, we used both ZNF367 knockdown and overexpression approaches to determine the effect of ZNF367 on cellular proliferation, invasion, and migration. SiRNA knockdown achieved up to 80% knockdown of ZNF367 mRNA and protein expression compared to siRNA negative control ( Figure S2).
ZNF367 knockdown in SW13 cells, increased cellular proliferation (30-40%) in vitro and (3.5-fold) in vivo (p,0.05; Figure 2). ZNF367 knockdown also increased cellular invasion and migration (p,0.05; Figure 3A). Given the dramatic effect of ZNF367 on cellular invasion and migration in SW13 cells, the effect of ZNF367 knockdown on cellular invasion and migration was also evaluated in BD140A, TPC-1, and HEK293 cell lines. The knockdown had a similar effect on the cellular invasion and migration of BD140A, TPC-1, and HEK293 cell lines (p,0.05; Figure 3, B-D). The effect of ZNF367 on cellular growth, invasion, and migration was also confirmed by overexpressing ZNF367 in HEK293 cells. ZNF367 overexpression decreased cellular invasion, migration, and colony formation compared to the empty vector control ( Figure 3E-G).
Because cellular invasion and migration require cell adhesion to the extracellular matrix, we evaluated the effect of ZNF367 knockdown on cellular adhesion to extracellular matrix proteins. We found increased cellular adhesion, with ZNF367 knockdown, to Laminin I (two-three-fold high) and fibrinogen (three-six-fold high) (p,0.05; Figure 3H). These proteins are known to function  to enhance cellular attachment to the basement membrane or extracellular matrix through integrin receptors, which allow cancer cell adhesion and migration.

ZNF367 regulates ITGA3 expression
We were interested in identifying the gene(s) and pathway(s) regulated by ZNF367, given the effect of ZNF367 on cellular invasion, migration, and adhesion, and because it is a transcription factor predicted to regulate gene expression. Using genome-wide mRNA expression analysis in cells transfected with ZNF367 and negative control siRNAs, we identified two candidate genes (ITGA3, serpin peptidase inhibitor, clade B (ovalbumin), member 9 [SERPINB9]) possibly regulated by ZNF367 based on applying several filter criteria (FDR,0.25, fold-change.1.5, common to all siRNA knockdown, and strong correlation in expression in human tumor samples) ( Figure S3). ITGA3 was chosen as a promising candidate for further study because it plays a significant role in cellular adhesion and invasion and was validated to be upregulated up to 2.5-fold with ZNF367 knockdown (p,0.05; Figure 4A) [11]. Moreover, ITGA3 mRNA expression was inversely correlated with ZNF367 mRNA expression in human adrenocortical tumor samples (r = -0.37, p = 0.015; Figure 4B). ITGA3 protein and tumor growth was measured weekly. The Y axis represents the tumor volume and X axis the weeks of tumor measurement after flank injection. *p,0.05 and error bars represent 6 SD. doi:10.1371/journal.pone.0101423.g002  expression was also increased with siRNA knockdown of ZNF367 ( Figure 4C). On the other hand, overexpression of ZNF367 reduced ITGA3 expression compared to the empty vector ( Figure 4D-E). These data suggest that ZNF367 regulates cellular adhesion, invasion, and migration through its effect, at least in part, on ITGA3 expression.

MiR-195 directly targets ZNF367 and regulates cellular invasion
To understand the mechanism by which ZNF367 expression is dysregulated in cancer, we postulated that microRNAs may be responsible for ZNF367 overexpression. To identify candidate microRNAs that may target ZNF367, we queried the Target scan and miRDB databases for microRNAs predicted to target ZNF367. A total of 3 out of 14 microRNAs, predicted to target ZNF367, were found to be differentially expressed in adrenocor-tical carcinoma (Table S1). However, only miR-195 was significantly inversely correlated with ZNF367 expression (r = -0.44; Figure 5A).
To investigate if miR-195 regulates ZNF367 expression, cells were transfected with pre-miR-195 and pre-negative control (pre-NC), which showed reduced ZNF367 mRNA and protein expression ( Figure 5B). Moreover, overexpression of miR-195 showed increased ITGA3 mRNA expression (two-fold) and cellular invasion compared to the negative control (p,0.05; Figure 5, C-D). To confirm that ZNF367 is a direct target of miR-195, we performed luciferase assays to determine whether miR-195 binds to the 39UTR of ZNF367 mRNA. We observed significant reduced luciferase activity with miR-195 overexpression in wild-type 39UTR compared to negative control and mutated ZNF367 39UTR-transfected cells (p,0.01; Figure 5E). Thus, these observations suggest that miR-195 negatively regulates the The binding site of miR-195 in the ZNF367 39UTR, along with the mutant construct in the predicted seed region. In the right panel, the luciferase assay demonstrates decreased luminescence in SW13 cells co-transfected with miR-195 and the negative control at 5 nM, with the empty vector, wild-type ZNF367 39UTR, or MUT-ZNF367 39UTR vector (mutated in the first three nucleotides of the seed sequence). The luminescence was read after 24 hours of transfection. The Y axis represents the ratio of ZNF367 39UTR to the empty vector. *p value , 0.05 compared to the negative control. Error bars represent 6 SEM. doi:10.1371/journal.pone.0101423.g005 expression of ZNF367 by directly targeting its 39UTR, resulting in decreased cellular invasion, the most dramatic effect of ZNF367 that we observed in our functional studies.

Discussion
To our knowledge, this is the first study to characterize the function of ZNF367 in cancers. We found that ZNF367 was overexpressed in a variety of endocrine cancers (adrenocortical carcinoma, papillary thyroid cancer, malignant pheochromocytoma/paraganglioma) compared to benign and or normal tissue samples. Functional in vitro and in vivo knockdown and overexpression studies demonstrated that ZNF367 regulates cellular proliferation, invasion, migration, and adhesion. We used genome-wide gene expression analysis to identify candidate genes that are regulated by ZNF367, and we identified ITAG3 as a gene that likely mediates the effect of ZNF367 on cellular adhesion, invasion, and migration. Moreover, in silico analysis of human tumor sample genome-wide gene expression analysis demonstrated an inverse relationship between ZNF367 and ITAG3 expression. Lastly, we demonstrate that dysregulated ZNF367 expression was associated with the loss of miR-195 expression in tumor samples and that this microRNA directly targets ZNF367 and regulates cellular invasion, providing an understanding of the mechanism for dysregulated ZNF367 expression. Based on these findings, we propose that there is a miR-195-ZNF367-ITAG3 axis that regulates endocrine cancer progression. This is the first study to characterize the expression and function of ZNF367 in cancer. ZNF367 knockout in mice has shown no significant abnormalities [12]. According to the Catalogue of Somatic Mutations in Cancer (COSMIC) database (http://cancer. sanger.ac.uk/cosmic/gene/overview?ln = ZNF367), missense, nonsense, and insertion mutations of ZNF367 have been identified at a low rate (16 of 5,089 samples) in a variety of cancers. Nonetheless, using both knockdown and overexpression strategies, we found that ZNF367 inhibited cancer growth and invasion, which suggests that the elevated levels in cancer samples are counter-regulatory and inhibit cancer progression. This finding is consistent with recent studies showing the diverse functions of other ZNF proteins in cancer biology and in mediating resistance to chemotherapy [9,10,[13][14][15]. Moreover, we demonstrate that the increased levels of ZNF367 are likely a result of a loss in expression of miR-195.
MiR-195 is a member of the miR-15 family of microRNAs that have been shown to have a role in cancer [16]. We and others have previously reported that miR-195 is downregulated in adrenocortical carcinoma and thyroid cancer, and that it is a prognostic marker in cancer [17][18][19][20][21]. Moreover, miR-195 has been shown to have a tumor suppressive function by regulating cell proliferation and apoptosis [22][23][24]. We identified the loss of miR-195 expression to be associated with increased ZNF367 expression in human cancer samples based on integrated analysis of microRNA expression data and the target prediction database, and we demonstrated that miR-195 directly targets ZNF367. MiR-195 overexpression resulted in increased cellular invasion, recapitulating the effect of ZNF367 on cellular invasion, and is consistent with its effect on cellular invasion in trophoblasts, in contrast to its effect on osteosarcoma and glioblastoma cells [22,25,26]. Such divergent effects of microRNAs on cellular invasion have previously been observed for other microRNAs, suggesting that miR-195 has a cell-specific effect. We propose that the loss of miR-195 expression results in the increased expression of ZNF367 and a miR-195-ZNF367 axis that is important in inhibiting cancer progression.
Given the significant effect of ZNF367 on cellular adhesion, migration, and invasion, which are the necessary steps for cancer progression, we investigated the possible gene(s) that may inhibit this progression as a result of high levels of ZNF367 in cancer, and we found that ITGA3 expression is modulated by ZNF367 and miR-195, and that it is inversely correlated in human tumor samples with ZNF367 expression. ITGA3 is a member of the integrin family of proteins, which function as a cell surface adhesion molecules and has a role in cancer invasion [27]. The increased ITGA3 levels from ZNF367 knockdown and increased adhesion of cells to Laminin I and Fibronectin suggest that decreased ITGA3 may mediate the inhibitory effect of ZNF367 on cellular adhesion and invasion. Thus, we propose a miR-195-ZNF367-ITGA3 axis that regulates cancer progression.
In summary, ZNF367 inhibits cancer progression by modulating cellular growth and invasion. ZNF367 expression is directly regulated by miR-195, and the loss of miR-195 is associated with increased ZNF367 in human cancer samples. The effect of ZNF367 on cellular invasion, at least in part, may be mediated by ITGA3. Taken together, this data suggest that ZNF367 regulates endocrine cancer progression.