Download FEZF2, a novel 3p14 tumor suppressor gene

Carcinogenesis vol.34 no.9 pp.1984–1993, 2013
doi:10.1093/carcin/bgt165
Advance Access publication May 14, 2013
FEZF2, a novel 3p14 tumor suppressor gene, represses oncogene EZH2 and MDM2
expression and is frequently methylated in nasopharyngeal carcinoma
Xing-sheng Shu, Lili Li, Mingfang Ji1, Yingduan Cheng,
Jianming Ying, Yichao Fan, Lan Zhong, Xiaodong Liu1,
Sai Wah Tsao2, Anthony TC.Chan and Qian Tao*
Cancer Epigenetics Laboratory, Department of Clinical Oncology, State Key
Laboratory of Oncology in South China, Sir YK Pao Center for Cancer and
Li Ka Shing Institute of Health Sciences, The Chinese University of Hong
Kong and CUHK Shenzhen Research Institute, Shatin, Hong Kong, 1Cancer
Research Institute, Zhongshan Affiliated Hospital of Sun Yat-sen University,
Zhongshan, Guangdong, 528403 China and 2Department of Anatomy,
University of Hong Kong, Pokfulam, Hong Kong
*To whom correspondence should be addressed. Cancer Center, Prince Wales
Hospital, Rm 315, The Chinese University of Hong Kong, Shatin, Hong Kong.
Tel: +852 2632 1340; Fax: +852 2648 8842;
Email: [email protected]
Nasopharyngeal carcinoma (NPC) is an Epstein-Barr virus-associated tumor prevalent in southern China and southeast Asia,
with the 3p14–p12 locus reported as a critical tumor suppressor gene (TSG) region during its pathogenesis. We identified a
novel 3p14.2 TSG, FEZF2 (FEZ family zinc finger 2), for NPC.
FEZF2 is readily expressed in normal tissues including upper
respiratory epithelium, testis, brain and ovary tissues, as well
as in immortalized nasopharyngeal epithelial cell line NP69, but
it is completely silenced in NPC cell lines due to CpG methylation of its promoter, although no homozygous deletion of FEZF2
was detected. 5-Aza-2′-deoxycytidine treatment restored FEZF2
expression in NPC cell lines along with its promoter demethylation. FEZF2 was frequently downregulated in NPC tumors,
with promoter methylation detected in 75.5% of tumors, but
only in 7.1% of normal nasopharyngeal tissues. Restored FEZF2
expression suppressed NPC cell clonogenicity through inducing
G2/M cell cycle arrest and apoptosis and also inhibited NPC cell
migration and stemness. FEZF2 acted as a histone deacetylaseassociated repressor downregulating multiple oncogenes including EZH2 and MDM2, through direct binding to their promoters.
Concomitantly, overexpression of EZH2 was frequently detected
in NPC tumors. Thus, we have identified FEZF2 as a novel 3p14.2
TSG frequently inactivated by promoter methylation in NPC,
which functions as a repressor downregulating multiple oncogene
expression.
Introduction
Cancer is a disease caused by accumulated genetic and epigenetic
changes (1). Abnormal activation of oncogenes and/or inactivation
of tumor suppressor genes (TSGs) due to these alternations confer precancerous cell growth advantages, further leading to tumor
initiation and progression (2,3). Nasopharyngeal carcinoma (NPC)
is a malignancy virtually 100% associated with Epstein-Barr
virus infection and shows unique geographical and ethnic distribution with high incidence in southern China and southeast Asia
(4). Genetic alterations in tumors, such as homozygous deletion or
loss of heterozygosity, pinpoint potential locus for critical TSGs
(1). A recent new model of cancer gene island phenomenon was
proposed, revealing that hemizygous deletions preferentially occur
Abbreviations: aCGH, array-based comparative genomic hybridization; Aza,
5-aza-2′-deoxycytidine; BGS, bisulfite genomic sequencing; ChIP, chromatin
immunoprecipitation; FEZF2, FEZ family zinc finger 2; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; HDAC, histone deacetylase; MSP, methylation-specific PCR; NPC, nasopharyngeal carcinoma; RT–PCR, reverse
transcription–PCR; TSA, trichostatin A; TSG, tumor suppressor gene.
in small gene islands harboring high density of TSGs to maximize
the proliferative advantage of tumor cells (5). Loss of 3p (3p21 and
3p14–p12) is common in NPC (4) and other tumors including lung
(6), esophageal (7) and breast (8) cancers. Several critical TSGs
within this region, inactivated by genetic and/or epigenetic mechanisms, have been identified for NPC and other tumors, such as FHIT
(3p14.2) (9), PTPRG (3p21–p14) (10), ADAMTS9 (3p14.1) (11),
RASSF1A (3p21.3) (12) and MLH1 (3p21.3) (13). We have also
refined several minimal deleted regions in 3p including 3p14–p12
in NPC, using 1 Mb array-based comparative genomic hybridization (aCGH), and identified some functional TSGs, like BLU (14),
PLCD1 (15) and DLEC1 (16), as well as a novel 3p14 candidate
tumor suppressor FEZF2 (FEZ family zinc finger 2, also known as
ZNF312 or ZFP312).
First isolated as a neural-specific gene, Fezf2 is specifically
expressed in the forebrain of Xenopus, zebra fish and mouse, and regulates forebrain development through controlling neuronal differentiation (17–24). FEZF2 belongs to the FEZF (forebrain embryonic zinc
finger) protein family, which includes another member FEZF1 (also
known as ZNF312B). Like FEZF2, FEZF1 also plays a critical role
during neural development (21,25). Recent data demonstrated that
human FEZF1 acts as an oncogene mediating the transcription activation of K-RAS, thus contributing to gastric cancer progression (26).
However, the role of FEZF2 in human tumorigenesis is still unknown.
In this study, we examined the epigenetic alteration of FEZF2 in
NPC and further characterized its tumor suppressive functions and the
underlying molecular mechanisms during NPC pathogenesis.
Materials and methods
Cell lines and tumor samples
A series of NPC and glioma cell lines were used (27,28). Cell lines were cultured in RPMI 1640 (Invitrogen, Grand Island, NY) supplemented with 10%
fetal bovine serum. Immortalized, non-transformed normal epithelial cell line
NP69 was used as a control (29). Cell lines were obtained either from the
American Type Culture Collection or from our collaborators. RNA samples
of three normal nasopharyngeal biopsy tissues were described previously
(30). DNA samples of normal nasopharyngeal and primary NPC tissues were
described previously. All the NPC tissues were Epstein-Barr virus positive
(28). Nasal swab DNA samples from NPC patients were also used. C15, C17
and C18, three nude mice-passaged undifferentiated NPC tumors from North
Africans, were used (31). Paired RNA samples of tumors and adjacent normal tissues were obtained from BioChain (Newark, CA) or Stratagene (Santa
Clara, CA). Human normal tissues RNA were purchased from Stratagene,
BioChain or Chemicon (Billerica, MA).
Array-based comparative genomic hybridization
The aCGH was carried out as reported previously (32), using 1 Mb resolution whole-genome arrays from Sanger Institute (Cambridge, UK) with 3040
bacterial artificial chromosome/P1-derived artificial chromosome clones. Log2
ratio ranging from −0.2 to −0.7 was considered as hemizygous deletion.
Semiquantitative and quantitative reverse transcription–PCR
Total RNA was extracted using TRI reagent. Reverse transcription (RT) using
random hexamer and RT–PCR using Go-Taq (Promega, Madison, WI) were
performed as described previously (33). Primers used for RT–PCR were listed
in Supplementary Table S1, available at Carcinogenesis Online.
Quantitative real-time PCR was performed as described previously (31).
SYBR Green master mix (Applied Biosystems, Grand Island, NY) was used.
The expression of target genes in FEZF2-transfected cells was normalized
to those transfected with vector control. Glyceraldehyde 3-phosphate
dehydrogenase (GAPDH) was used as an internal control. The sequences of
primers used in quantitative RT–PCR will be provided upon request.
Bisulfite treatment and promoter methylation analysis
Bisulfite modification of DNA, methylation-specific PCR (MSP) and
bisulfite genomic sequencing (BGS) were conducted according to our
© The Author 2013. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
1984
FEZF2 as a novel 3p14 tumor suppressor
previous report (32). The specificity of MSP primers was tested first using
DNA samples without bisulfite treatment. Thirty-five cycles of PCR reaction were performed in MSP using primers amplifying methylated gene
allele, with 40 cycles for reaction using primers amplifying unmethylated
gene allele. PCR products amplified using BGS primers were cloned into
pCR4-TOPO vector (Invitrogen), with 6–10 colonies randomly chosen and
sequenced. Primers used for MSP and BGS were listed in Supplementary
Table S1, available at Carcinogenesis Online.
Demethylation treatment using 5-aza-2′-deoxycytidine and trichostatin A
Treatment of NPC cell line using 5-aza-2′-deoxycytidine (Aza) and trichostatin A (TSA) was carried out as described previously (32). Briefly, cells were
treated with 10 μM Aza (Sigma, Ronkonkoma, NY) for 72 h and harvested
for DNA and RNA extraction. Alternatively, after 72 h of Aza treatment, cells
were incubated with 100 ng/ml of TSA for additional 24 h.
Construction of FEZF2 expression vector
The full-length open reading frame of FEZF2 was amplified from human testis
total RNA using Pfu polymerase (Stratagene) and cloned to the EcoRI and
BamHI sites of pcDNA3.1 (+) vector, with a Flag tag added to its C-terminus.
The sequence and orientation of FEZF2 open reading frame were confirmed.
FEZF2 open reading frame was also cloned into pCMV-BD vector, with the
DNA binding domain of GAL4 fused to the N-terminus of FEZF2 protein.
Monolayer and soft agar colony formation assays
For colony formation assay using monolayer culture, HNE1 and HONE1
cells were seeded in a 12-well plate at 1–2 × 105/well. The cells were then
transfected with FEZF2 expression vector or empty vector using Fugene
6 (Roche, Basel, Switzerland). After 48 h of post-transfection, the transfectants were subcultured into six-well plates for selection with 400 µg/ml
of G418 (Calbiochem, Darmstadt, Germany). After 2 weeks of selection,
surviving colonies (>50 cells per colony) were stained with gentian violet
and counted.
For soft agar assay, the transfected cells were suspended in growth
medium containing 0.35% agar and 400 µg/ml of G418 in 24-well plates.
Surviving colonies were photographed and counted after about 2 weeks of
selection.
Immunoprecipitation
HEK293T cells were transfected with control vector or FLAG-FEZF2. After
48 h of transfection, cells were washed with ice-cold phosphate-buffered
saline once and incubated with lysis buffer (50 mmol/l Tris–HCl, pH 8.0;
150 mmol/l NaCl and 0.5% NP40) on ice for 30 min. After centrifugation,
the supernatant was collected and 200 μg of protein was applied for immunoprecipitation using 40 µl anti-FLAG M2 affinity gel (Sigma) at 4°C overnight. Precipitated protein was eluted with 150 ng/μl of 3× FLAG peptide at
4°C for 1 h, and then used for immunoblotting analysis with 5% cell lysate
as input control.
For immunoprecipitation using histone deacetylase 1 (HDAC1) antibody,
200 μg of total protein lysate was incubated with 1 µg HDAC1 antibody at 4°C
overnight. The lysate was then incubated with 20 μl Protein G Sepharose beads
at 4°C for 4 h. Sepharose beads were washed three times using RIPA lysis
buffer and resuspended in 4× sodium dodecyl sulfate protein loading buffer.
After boiling for 5 min, the supernatant was collected and applied to immunoblotting analysis with 5% cell lysate as input control.
Western blot
Western blot was carried out as described previously (34). Membranes were
incubated with primary antibody at 4°C overnight, followed by incubation
with secondary antibody at room temperature for 1 h. Immunoreactive bands
were detected using western blot luminol reagent (GE Healthcare, Waukesha,
WI). The antibodies used were anti-FEZF2 (ab69436; Abcam, Cambridge,
UK), anti-FLAG (F7425; Sigma), anti-HDAC1 (5356; Cell Signaling, Boston,
MA), anti-MDM2 (sc-813; Santa Cruz, Dallas, TX), anti-p53 (M7001; Dako,
Glostrup, Denmark), anti-α-tubulin (MS-581; Thermo Lab Vision, Kalamazoo,
MI) and anti-GAPDH (MAB374; Millipore, Billerica, MA).
Flow cytometry analysis
Cells stably transfected with FEZF2 or empty vector were collected, washed
twice with phosphate-buffered saline and stained for 1.5 h at 37°C using 50 µg/
ml propidium iodide (Sigma). Cells were again washed with phosphate-buffered saline and applied to flow cytometry analysis using Beckman Cell Lab SC
Quanta. Results were analyzed using ModFit LT.
Wound healing assay
Wound healing assay to evaluate cell migration ability was performed as
described previously (35). Briefly, cells were transfected with empty vector or
FEZF2 construct and allowed to grow until confluent (>95%). Cell scratches
were then created using 20 µl sterile tips and washed twice with culturing
medium. After indicated time points of incubation, cells were imaged under
a phase contrast microscope. The experiments were performed in duplicate.
Immunofluorescence
Immunostaining of HONE1 and HNE1 cells was performed as described previously (28). Briefly, the cells were seeded on coverslips and transfected with
FLAG-FEZF2 plasmid. After 24 h of transfection, cells were fixed and incubated with primary antibody anti-FLAG M2 (F3165; Sigma) at 4°C overnight, and then stained with secondary fluorescein isothiocyanate-conjugated
antibody (F0313; Dako, Glostrup, Denmark) at 37°C for 1 h. Cell nuclei
were then stained with 4′,6-diamidino-2-phenylindole and imaged using
a confocal microscope (Leica TCS SP5; Leica Microsystems, Mannheim,
Germany).
Chromatin immunoprecipitation
Chromatin immunoprecipitation (ChIP) was carried out using ChIP-IT
Express Kit from Active Motif (#53008; Carlsbad, CA). For each ChIP reaction, 25 µg of total chromatin was incubated with 20 µl of Protein G magnetic beads and 1 µg of FLAG-M2 antibody (F3165; Sigma) at 4°C overnight.
Both input and precipitated DNA were purified with QIAamp DNA Mini Kit
(Qiagen, Valencia, CA) for subsequent real-time PCR. The relative enrichment of precipitated DNA was normalized to its input. Following primers
were used for ChIP assay: MDM2ChIPF, 5′-CATTTGGGTACAACTCCAGC
and MDM2ChIPR, 5′-TGGAAACTGCGACAAATGCG; EZH2ChIPF,
5′-AAATTAGTCGGGTGTGGTGG and EZH2ChIPR, 5′-AAACGGAGTCTC
ACACTGTC and GAPDHChIPF, 5′-TATCAGGTCCAGGCTACAG and
GAPDHChIPR, 5′-GGCTCTGCGGTAGTGACAC.
Dual-luciferase reporter assay
Cells were co-transfected with indicated reporter constructs together with
either FEZF2 expression vector or empty vector in 24-well plate. After 48 h of
transfection, cells were collected and analyzed by the Dual-Luciferase Assay
Kit (Promega, Madison, WI). Each experiment was repeated in triplicate wells
for three times independently.
Immunohistochemistry
Immunohistochemistry was performed using the ChemMate EnVision
Detection Kit (#GK500705; Dako, Carpinteria, CA) according to manufacturer’s instruction. Primary EZH2 antibody used for staining was from Zymed
(#18–7395; Invitrogen, Grand Island, NY). A total of 67 primary NPC and 17
normal nasopharynx or chronic nasopharyngitis cases were used. Cells with
strong nuclear staining were taken as EZH2 positive. Percentage of EZH2
positive cells was scored by a pathologist blindly in five random fields (100
cells per field) for each sample.
Statistical analysis
Results were presented as mean ± SD. Statistical analysis was carried out with
Student’s t-test and P < 0.05 was considered as statistically significant.
Results
Identification of FEZF2 as a novel 3p14.2 TSG candidate for NPC
To search for novel TSGs for NPC, we refined 3p14–p12, a critical
TSG hotspot at 3p, as a major deleted region in NPC cell lines
using 1 Mb aCGH (Figure 1A). Hemizygous deletions represented
by three bacterial artificial chromosome clones (bA170K19,
bA204J18 and bA108A8), all located at 3p14.2, were detected in
all NPC cell lines studied (Supplementary Figure S1, available
at Carcinogenesis Online), indicating the existence of critical
TSGs within this locus. We thus examined the expression profile
of candidate genes residing in this locus by semiquantitative RT–
PCR and found that FEZF2 was silenced in all five NPC cell lines
(Figure 1D). However, FEZF2 was readily expressed in normal
larynx and trachea tissues (Figure 1B), normal nasopharyngeal
tissues (Figure 1C) and immortalized nasopharyngeal epithelial
cell line NP69 (Figure 1D). Downregulation of FEZF2 in NPC
cell lines, compared with normal tissues, was further confirmed
by quantitative real-time PCR (Figure 1H) and western blot at the
protein level (Figure 1I).
We also evaluated FEZF2 expression in other human normal tissues. FEZF2 was highly expressed in both adult and fetal brain
1985
X.-s.Shu et al.
Fig. 1. Identification of FEZF2 as a 3p14.2 candidate TSG in NPC. (A) FEZF2 locus and gene structure. 3p14–p12 was detected as a frequent deletion in NPC
by aCGH as indicated by a black arrowed line. RT–PCR (F/R) and multiplex genomic-DNA-PCR (F/int2R) primer sites are indicated. Transcription start site is
indicated by a curved arrow. A CpG island spanning FEZF2 promoter and exon 1 is shown at the bottom. Each vertical bar represents a CpG site. MSP primer
sites and BGS region are labeled. (B) FEZF2 expression in human normal adult and fetal tissue panel by RT–PCR, with GAPDH as a control. Sk. Muscle,
skeletal muscle. (C) FEZF2 is expressed in normal nasopharyngeal tissues as evaluated by RT–PCR. (D) FEZF2 was silenced by promoter methylation in
NPC cell lines as determined by RT–PCR and MSP but expressed and unmethylated in immortalized normal cell line NP69. M, methylated; U, unmethylated.
(E) Multiplex genomic DNA-PCR showed no homozygous deletion of FEZF2 in NPC cell lines. GAPDH was used as a control. Normal peripheral blood
mononuclear cells were used as positive controls. (F) High-resolution BGS methylation analysis of FEZF2 promoter. Each row of circles represented an
individual promoter allele. Filled circle, methylated CpG site; open circle, unmethylated CpG site. (G) Pharmacologic demethylation restored FEZF2 expression
in methylated and silenced NPC cell lines. A + T, treatment with Aza and TSA. (H) Quantitative RT–PCR further confirmed FEZF2 downregulation in NPC cell
lines compared with normal tissues and its restoration after demethylation treatment. The expression level of each sample was normalized to internal control
GAPDH. Fold change of FEZF2 expression was calculated relative to that of larynx tissue. Experiments were performed three times independently. (I) Western
blot showed downregulation of FEZF2 protein in NPC cell lines compared with normal trachea tissue. α-Tubulin was used as a loading control.
tissues, consistent with its important role in the development of neural
system. In addition, FEZF2 expression was also observed in adult testis and ovary tissues, whereas barely detectable in other human tissues
(Figure 1B). The expression of FEZF2 in upper respiratory tract tissues and its silencing in NPC cell lines suggested that FEZF2 is likely
a candidate tumor suppressor for NPC.
Silencing of FEZF2 due to its promoter CpG methylation in NPC
As TSGs can be inactivated by genetic deletions, we first determined
the homozygous deletion of FEZF2 in NPC. Deletion of FEZF2 exon
2 was assessed by multiplex genomic DNA-PCR, using GAPDH as an
internal control. Results showed no homozygous deletion of FEZF2
in NPC cell lines and normal peripheral blood mononuclear cell controls (Figure 1E), suggesting that other mechanism is responsible for
FEZF2 silencing in NPC.
1986
We thus studied whether promoter CpG methylation was involved
in silencing FEZF2 in NPC. Bioinformatic analysis revealed a typical
CpG island spanning the promoter and exon 1 of FEZF2 (Figure 1A).
MSP assay showed that FEZF2 promoter was methylated in all NPC
cell lines, but not in immortalized nasopharyngeal epithelial cell line
NP69 (Figure 1D). No methylation was detected in unbisulfited DNA
of NPC cell lines, verifying the specificity of MSP (Supplementary
Figure S2A, available at Carcinogenesis Online). MSP results were
further confirmed by high-resolution BGS analysis of 31 CpG sites
in FEZF2 promoter CpG island. Most CpG sites were intensively
methylated in silenced NPC cell lines, whereas rarely methylated in
NP69 cells (Figure 1F). These results revealed a strong correlation
between FEZF2 silencing and its promoter methylation in NPC cells.
To determine whether methylation directly contributes to the
silencing of FEZF2, we treated silenced NPC cell lines with DNA
FEZF2 as a novel 3p14 tumor suppressor
methyltransferase inhibitor Aza, alone or in combination with HDAC
inhibitor TSA. After Aza treatment, FEZF2 expression was restored
(Figure 1G, upper panel and Figure 1H), accompanied by significant
increase of unmethylated promoter alleles (Figure 1G, bottom panel).
Demethylation of FEZF2 promoter in C666-1 and HONE1 cells was
confirmed by BGS analysis (Supplementary Figure S2B, available at
Carcinogenesis Online). Thus, promoter CpG methylation directly
mediates FEZF2 silencing in NPC.
FEZF2 is frequently downregulated and methylated in NPC tumors
We also examined FEZF2 expression in paired tumor and adjacent
normal tissues. FEZF2 was downregulated in larynx tumor compared
with the adjacent normal larynx tissue (Figure 2A), but not expressed
in other common tumors, including lung, colon, rectum, liver, kidney, breast and gastric tumors, as well as their adjacent normal tissues
(Supplementary Figure S2C, available at Carcinogenesis Online).
Consistently, we observed frequent FEZF2 downregulation in primary NPC tissues (Figure 2A). We further detected FEZF2 methylation in 75.5% of primary NPC tumor tissues (37 of 49) and three NPC
tumors passaged from nude mice, but only in 7.1% of normal nasopharyngeal tissues (1 of 14) (Figure 2B). FEZF2 methylation in NPC
tumors was further confirmed by BGS (Figure 2C). Furthermore,
FEZF2 methylation was detected in 75% (12 of 16) of nasal swab
samples from NPC patients (Figure 2D), whereas only one of the
seven nasopharyngitis tissues from healthy individuals was detected
having weak methylation (Supplementary Figure S2D, available at
Carcinogenesis Online). These results suggest that FEZF2 is frequently downregulated and methylated in NPC in a tumor-specific
manner.
FEZF2 is a transcriptional repressor
FEZF2 contains an EH1 repressor domain and six tandem C2H2type zinc fingers, which are highly conserved among its homologs
(Supplementary Figure S3, available at Carcinogenesis Online).
Immunofluorescence staining showed that FEZF2 is localized in
the nucleus (Figure 3A and Supplementary Figure S4, available at
Carcinogenesis Online). To determine whether FEZF2 acts as a
transcriptional activator or repressor, a GAL4 luciferase reporter
system was used. Fused to the GAL4 DNA binding domain, FEZF2
was recruited to the luciferase gene promoter containing five GAL4
binding sites. Results showed that BD-FEZF2 dramatically repressed
the reporter activities in HNE1 and HONE1 cells. Moreover,
FEZF2 repressed the reporter activity in a dose-dependent manner
in HEK293T cells (Figure 3B). Furthermore, we found that FEZF2
physically associates with HDAC1 (Figure 3C and 3D), a common
core component of repressor complexes like LSD-CoREST and
Fig. 2. FEZF2 is downregulated and methylated in NPC tissues. (A) Left: FEZF2 was downregulated in primary larynx tumor compared with the paired adjacent
normal larynx tissue. N, normal; T, tumor. Right: FEZF2 was frequently silenced in primary NPC tumor tissues, evaluated by RT–PCR. (B) Representative MSP
analysis of FEZF2 methylation in normal nasopharyngeal tissues, NPC tumor tissues and NPC tumors from nude mice. (C) BGS analysis confirmed FEZF2
methylation in primary NPC tissues. Filled circle, methylated CpG site; open circle, unmethylated CpG site. (D) FEZF2 methylation in nasal swab samples from
NPC patients, examined by MSP. M, methylated; U, unmethylated.
1987
X.-s.Shu et al.
Fig. 3. FEZF2 is a transcriptional repressor. (A) Upper panel: Functional domains of FEZF2 protein. ZnF, zinc finger. Bottom panel: Confocal microscopy of
FEZF2 subcellular localization in HONE1 cells. 4′,6-Diamidino-2-phenylindole was used for nucleus staining. Scale bar: 5 μm. (B) Lower left: GAL4 luciferase
reporter assay showed that FEZF2 suppressed the transcription of reporter gene in HNE1 and HONE1 cells. *P < 0.05; **P < 0.01. Lower right: Transcription
repression mediated by FEZF2 was dose dependent in HEK293T cells. (C and D) Co-immunoprecipitation assays showed that HDAC1 binds to the transfected
(C) and endogenous (D) FEZF2 protein. IP, immunoprecipitation; IB, immunoblotting.
NuRD. These data suggest that FEZF2 is a transcriptional repressor
and a component of some repressor complexes.
FEZF2 suppresses NPC cell growth through inducing cell cycle
arrest and apoptosis
To investigate the potential tumor suppressive properties of FEZF2,
we performed cell colony formation assays of methylated/silenced
NPC cell lines, transfected either with a empty vector or a plasmid
encoding FEZF2. Results showed that, compared with control groups,
1988
FEZF2 significantly inhibited the anchorage-dependent colony formation of monolayer-cultured NPC cells and also suppressed the
anchorage-independent growth of NPC cells in soft agar (Figure 4A
and 4B). Restored expression of FEZF2 in these cells was confirmed
by RT–PCR (Figure 4B), which showed FEZF2 expression levels
in transfected cells are similar to normal tissues, such as larynx and
trachea.
We then studied the underlying mechanisms of FEZF2-mediated
growth inhibition. Flow cytometry analysis showed an increase of
FEZF2 as a novel 3p14 tumor suppressor
Fig. 4. FEZF2 suppresses NPC cell growth through inducing cell cycle arrest and apoptosis. (A) Representative monolayer culture and soft agar colony
formation assay of HONE1 and HNE1 cells. (B) Upper panel: Restored expression of FEZF2 in transfected cell lines as confirmed by RT–PCR; bottom panel:
quantitative analysis of colony numbers. Data are presented as mean ± SD of three independent experiments. **P < 0.01. (C) Flow cytometry analysis of FEZF2expressing HNE1 cells showed increased G2/M phase cell percentage compared with controls. Quantitative representation of the results is shown at right side.
*P < 0.05. (D) Left: A sub-G1 (or apoptosis) cell peak was observed in FEZF2-expressing cells; right: immunofluorescence staining of HNE1 cells showed
nucleus condensation and segmentation of some FEZF2-expressing cells. Scale bar: 2.5 μm. *P < 0.05. (E) Left: RT–PCR showed the downregulation of
cell cycle regulators (CDC25A and E2F3) and antiapoptotic genes (BCL2 and cIAP2) by FEZF2. Right: Western blot showed BCL2 downregulation and p53
upregulation by FEZF2. GAPDH was used as a loading control.
G2/M phase cell percentage in tumor cells stably transfected with
FEZF2 compared with control cells (Figure 4C). In addition, we
observed the downregulation of two critical cell cycle regulators
(CDC25A and E2F3), which promote G2/M transition, by FEZF2
(Figure 4E). A sub-G1 peak was observed in FEZF2-transfected
cells but not in control cells (Figure 4D), suggesting that FEZF2
expression induced tumor cell apoptosis. Immunofluorescence
staining showed that some FEZF2-expressing cells had abnormal
nucleus condensation and segmentation, a typical mark of cells
undergoing apoptosis (Figure 4D). Moreover, the mRNA expression of two antiapoptotic genes, BCL2 and cIAP2, was downregulated in FEZF2-expressing cells (Figure 4E). Consistently,
we also observed downregulation of BCL2 protein and upregulation of the key apoptosis regulator p53 by FEZF2 (Figure 4E).
We further found that p53 expression was detectable in most
NPC cell lines, whereas only a single point mutation of p53 was
reported in two NPC cell lines (Supplementary Figure S5, available at Carcinogenesis Online), indicating that epigenetic silencing of FEZF2 likely contributes critically to the perturbation of
p53 signaling pathway in NPC cells. These results suggest that
FEZF2 suppresses NPC cell growth through inducing cell cycle
arrest and apoptosis.
FEZF2 inhibits the migration and stemness of NPC cells
We further found that FEZF2 inhibited the migration of NPC cells
by wound healing assay (Figure 5A). Considering the pivotal role of
FEZF2 in neural cell differentiation and the reported stem-like properties of NPC cells (32,35), we also assessed whether it regulates the
1989
X.-s.Shu et al.
Fig. 5. FEZF2 inhibits NPC cell migration and stemness. (A) Wound healing assay demonstrated a slower wound closure of FEZF2-expressing HNE1 and
HONE1 cells compared with controls. (B) The mesenchymal-like phenotype of HNE1 and HONE1 cells was changed to epithelial-like phenotype after FEZF2
expression. After 48 h of post-transfection, cells were harvested and seeded in six-well plate with the selection of G418 (400 μg/ml). After 1–2 weeks, cells
were imaged under a microscope. (C) Expression of several stem cell markers (KLF4, CD44 and BMI1) was decreased after FEZF2 expression, as evaluated by
RT–PCR.
stemness of NPC cells. Indeed, we observed the change of mesenchymal-like phenotype to epithelial-like phenotype in FEZF2-expressing
NPC cells (Figure 5B) and the downregulation of several stem cell
markers including KLF4, CD44 and BMI1, by FEZF2 (Figure 5C).
These data indicated that FEZF2 inhibits the migration and stemness
of NPC cells.
FEZF2 represses multiple oncogene expression including EZH2
and MDM2
To identify cancer genes regulated by FEZF2, we examined the
expression of a panel of oncogenes in NPC cells transfected with
FEZF2 construct. Quantitative RT–PCR showed that the expression
of multiple oncogenes, including EZH2 and MDM2, was significantly downregulated in FEZF2-transfected HONE1 and HNE1 cells
(Figure 6A; Supplementary Table S2, available at Carcinogenesis
Online).
1990
Recently, the genome-wide binding sites of zebra fish FEZF2
were characterized, with a core CnnCAnCn sequence as the putative consensus FEZF2 binding motif (36). As FEZF2 is evolutionarily well conserved, the human FEZF2 is likely to target similar
sites in the genome. We thus analyzed the promoter regions of
EZH2 and MDM2 and found multiple potential FEZF2 binding
sites (the CnnCAnCn motif) in the promoter. We then performed
ChIP assay and detected the recruitment of FEZF2 to MDM2 and
EZH2 promoters (Figure 6B). Furthermore, the promoter activity of MDM2 and its protein level was significantly decreased in
FEZF2-expressing cells (Figure 6C). EZH2 was also found to be
upregulated in most NPC cell lines compared with normal immortalized nasopharyngeal cell line NP69, which is negatively correlated with FEZF2 expression status in these cell lines (Figure 6D).
Immunohistochemistry further confirmed the overexpression
of EZH2 in NPC tumors compared with normal nasopharyngeal
FEZF2 as a novel 3p14 tumor suppressor
Fig. 6. FEZF2 represses multiple oncogene expression. (A) Real-time RT–PCR showed FEZF2 repressed multiple oncogenes expression in HONE1 and HNE1
cells. The gene expression level in vector-expressing cells was set to 1.00. Fold change of expression in FEZF2-expressing cells was calculated by normalizing to
vector-transfected cells. GAPDH was used as an internal control. Data are shown as mean ± SD of three independent experiments. *P < 0.05; **P < 0.01.
(B) Top panel: Putative FEZF2 binding sites at MDM2 and EZH2 promoters. Fragments amplified in ChIP–PCR were indicated by bottom black lines. Lower
panel: ChIP–quantitative PCR showed enrichment of FEZF2 binding at MDM2 and EZH2 promoter regions. GAPDH was used as a negative control. The signal
of precipitated DNA was normalized to their input DNA. *P < 0.05; **P < 0.01. (C) FEZF2 repressed MDM2 promoter activity and downregulated its protein
expression, as determined by luciferase reporter assay and western blot, respectively. **P < 0.01. (D) RT–PCR showed upregulation of EZH2 in NPC cell
lines compared with normal cell line NP69. Expression and methylation status of FEZF2 in these cell lines are also indicated at the bottom. (E) Representative
immunohistochemistry results for EZH2 in normal nasopharynx and NPC tissues. Two magnifications (×20 and ×40) of the same sample are shown.
1991
X.-s.Shu et al.
tissues, showing that EZH2 was highly expressed in 71.6% of NPC
tumors (48/67), but not overexpressed in 17 normal nasopharynx
or chronic nasopharyngitis tissues examined (Figure 6E). Thus,
frequent methylation and silencing of FEZF2 contributed to EZH2
overexpression in NPC cells.
Funding
Discussion
Acknowledgements
Our aCGH study and others showed that 3p14 is one of the most
frequently deleted regions in NPC cell lines (32,37). Previously,
several 3p14 genes located adjacent to FEZF2, like FHIT (9,38–
40), PTPRG (10,41,42) and ADAMTS9 (11), have been identified as functional TSGs for NPC and other cancers. Our study
demonstrated that FEZF2 is another functional novel 3p14 TSG.
Thus, FHIT-PTPRG-FEZF2 appears to be a small TSG island at
3p14. Moreover, a study using cytogenetic detection method FISH
even proposed 3p12.3–p14.2 deletion as a prognostic marker for
advanced NPC patients (43).
Epigenetic alterations, especially promoter CpG methylation,
have been shown to be as important as genetic abnormalities during tumor initiation and progression (44–46). We found that promoter methylation is responsible for FEZF2 silencing in NPC.
Importantly, we observed frequent FEZF2 methylation in primary
NPC tumor tissues and nasal swab samples from NPC patients, but
not in normal nasopharyngeal tissues, indicating its methylation is
tumor specific.
As FEZF2 is a critical gene for vertebrate forebrain development, we also examined whether it is a candidate TSG for brain
tumor. Through analyzing the microarray data from online database (Oncomine; Compendia Bioscience, Ann Arbor, MI) (47),
we found that FEZF2 expression was significantly reduced in
brain tumors compared with normal brain tissues (Supplementary
Figure S6A, available at Carcinogenesis Online), with highgrade brain tumors showing lower FEZF2 expression than lowgrade brain tumors (Supplementary Figure S6B, available at
Carcinogenesis Online). We also found that FEZF2 expression
levels in normal brain and glioblastoma tissues were negatively
correlated with EZH2 levels (Supplementary Figure S6C, available at Carcinogenesis Online), indicating that FEZF2 silencing also contributes to EZH2 overexpression in brain tumors. In
addition, we detected FEZF2 downregulation in glioma cell lines
(Supplementary Figure S6D, available at Carcinogenesis Online),
although no promoter methylation was observed, suggesting that
other mechanism like genetic deletions might be responsible for its
inactivation in these cell lines.
So far, two FEZF family genes, FEZF2 and FEZF1 (Supplementary
Figure S7, available at Carcinogenesis Online), have both been linked
to cancer development, although displaying distinct properties. FEZF1
acts as a oncogene promoting gastric cancer progression (26) and is
activated by abnormal epigenetic modifications in gastric cancer (48).
In this study, FEZF2 was found to possess tumor suppressor properties and is silenced by promoter methylation in NPC. The unique
proline-rich domain of FEZF1 mediates its translocation to nucleus
and the activation of K-RAS oncogene (26), whereas FEZF2 localizes
in nucleus and represses the transcription of multiple oncogenes.
In conclusion, we identified and functionally characterized a novel
3p14.2 TSG, FEZF2, in NPC pathogenesis. FEZF2 is frequently
silenced by promoter methylation in NPC, and functions as a TSG
suppressing the NPC growth through inducing cell cycle arrest and
apoptosis and inhibiting NPC cell migration and stemness. Finally,
FEZF2 was shown to be an HDAC-associated transcription repressor,
downregulating multiple oncogenic genes, which control cell cycle
progression, apoptosis and cell stemness (Supplementary Figure S8,
available at Carcinogenesis Online).
We thank Drs T.J.Seng and C.Langford for their help with array-CGH.
Supplementary material
Supplementary Figures S1–S8 and Tables S1 and S2 can be found at
http://carcin.oxfordjournals.org/
1992
Hong Kong Research Grants Council (GRF 474710); The Chinese
University of Hong Kong and China National Natural Science
Foundation (81071634 and 81172582).
Conflict of Interest Statement: None declared.
References
1.Balmain,A. et al. (2003) The genetics and genomics of cancer. Nat. Genet.,
33, 238–244.
2.Baylin,S.B. et al. (2006) Epigenetic gene silencing in cancer - a mechanism
for early oncogenic pathway addiction? Nat. Rev. Cancer, 6, 107–116.
3.Vogelstein,B. et al. (2004) Cancer genes and the pathways they control.
Nat. Med., 10, 789–799.
4.Tao,Q. et al. (2007) Nasopharyngeal carcinoma: molecular pathogenesis
and therapeutic developments. Expert Rev. Mol. Med., 9, 1–24.
5.Solimini,N.L. et al. (2012) Recurrent hemizygous deletions in cancers may
optimize proliferative potential. Science, 337, 104–109.
6.Zabarovsky,E.R. et al. (2002) Tumor suppressor genes on chromosome
3p involved in the pathogenesis of lung and other cancers. Oncogene, 21,
6915–6935.
7.Ogasawara,S. et al. (1995) Frequent microsatellite alterations on chromosome 3p in esophageal squamous cell carcinoma. Cancer Res., 55,
891–894.
8.Martinez,A. et al. (2001) Chromosome 3p allele loss in early invasive
breast cancer: detailed mapping and association with clinicopathological
features. Mol. Pathol., 54, 300–306.
9.Ko,J.Y. et al. (2002) Definition of three minimal deleted regions by comprehensive allelotyping and mutational screening of FHIT,p16(INK4A), and
p19(ARF) genes in nasopharyngeal carcinoma. Cancer, 94, 1987–1996.
10. Cheung,A.K. et al. (2008) Functional analysis of a cell cycle-associated,
tumor-suppressive gene, protein tyrosine phosphatase receptor type G, in
nasopharyngeal carcinoma. Cancer Res., 68, 8137–8145.
11. Lo,P.H. et al. (2010) Extracellular protease ADAMTS9 suppresses esophageal and nasopharyngeal carcinoma tumor formation by inhibiting angiogenesis. Cancer Res., 70, 5567–5576.
12. Lo,K.W. et al. (2001) High frequency of promoter hypermethylation of
RASSF1A in nasopharyngeal carcinoma. Cancer Res., 61, 3877–3881.
13. Wong,T.S. et al. (2004) Quantitative plasma hypermethylated DNA markers of undifferentiated nasopharyngeal carcinoma. Clin. Cancer Res., 10,
2401–2406.
14. Qiu,G.H. et al. (2004) The candidate tumor suppressor gene BLU, located
at the commonly deleted region 3p21.3, is an E2F-regulated, stress-responsive gene and inactivated by both epigenetic and genetic mechanisms in
nasopharyngeal carcinoma. Oncogene, 23, 4793–4806.
15. Hu,X.T. et al. (2009) Phospholipase C delta 1 is a novel 3p22.3 tumor suppressor involved in cytoskeleton organization, with its epigenetic silencing
correlated with high-stage gastric cancer. Oncogene, 28, 2466–2475.
16. Ying,J. et al. (2009) DLEC1 is a functional 3p22.3 tumour suppressor
silenced by promoter CpG methylation in colon and gastric cancers. Br.
J. Cancer, 100, 663–669.
17. Matsuo-Takasaki,M. et al. (2000) Cloning and expression of a novel
zinc finger gene, Fez, transcribed in the forebrain of Xenopus and mouse
embryos. Mech. Dev., 93, 201–204.
18. Levkowitz,G. et al. (2003) Zinc finger protein too few controls the development of monoaminergic neurons. Nat. Neurosci., 6, 28–33.
19. Yang,Z. et al. (2001) A zebrafish forebrain-specific zinc finger gene can
induce ectopic dlx2 and dlx6 expression. Dev. Biol., 231, 138–148.
20. Rouaux,C. et al. (2010) Fezf2 directs the differentiation of corticofugal
neurons from striatal progenitors in vivo. Nat. Neurosci., 13, 1345–1347.
21. Shimizu,T. et al. (2010) Zinc finger genes Fezf1 and Fezf2 control neuronal differentiation by repressing Hes5 expression in the forebrain.
Development, 137, 1875–1885.
22. Chen,J.G. et al. (2005) Zfp312 is required for subcortical axonal projections and dendritic morphology of deep-layer pyramidal neurons of the
cerebral cortex. Proc. Natl Acad. Sci. USA, 102, 17792–17797.
23. Chen,B. et al. (2005) Fezl regulates the differentiation and axon targeting of layer 5 subcortical projection neurons in cerebral cortex. Proc. Natl
Acad. Sci. USA, 102, 17184–17189.
FEZF2 as a novel 3p14 tumor suppressor
24. Shim,S. et al. (2012) Cis-regulatory control of corticospinal system development and evolution. Nature, 486, 74–79.
25. Shimizu,T. et al. (2009) Formation and patterning of the forebrain and
olfactory system by zinc-finger genes Fezf1 and Fezf2. Dev. Growth
Differ., 51, 221–231.
26. Song,I.S. et al. (2009) Human ZNF312b promotes the progression of gastric cancer by transcriptional activation of the K-ras gene. Cancer Res., 69,
3131–3139.
27. Cui,Y. et al. (2008) OPCML is a broad tumor suppressor for multiple carcinomas and lymphomas with frequently epigenetic inactivation. PLoS One,
3, e2990.
28. Cheng,Y. et al. (2010) KRAB zinc finger protein ZNF382 is a proapoptotic tumor suppressor that represses multiple oncogenes and is commonly
silenced in multiple carcinomas. Cancer Res., 70, 6516–6526.
29. Tsao,S.W. et al. (2002) Establishment of two immortalized nasopharyngeal
epithelial cell lines using SV40 large T and HPV16E6/E7 viral oncogenes.
Biochim. Biophys. Acta, 1590, 150–158.
30. Feng,P. et al. (2000) Expression of Epstein-Barr virus lytic gene BRLF1 in
nasopharyngeal carcinoma: potential use in diagnosis. J. Gen. Virol., 81(Pt
10), 2417–2423.
31. Shu,X.S. et al. (2011) The epigenetic modifier PRDM5 functions as a
tumor suppressor through modulating WNT/β-catenin signaling and is frequently silenced in multiple tumors. PLoS One, 6, e27346.
32. Ying,J. et al. (2006) Functional epigenetics identifies a protocadherin
PCDH10 as a candidate tumor suppressor for nasopharyngeal, esophageal
and multiple other carcinomas with frequent methylation. Oncogene, 25,
1070–1080.
33. Tao,Q. et al. (2002) Defective de novo methylation of viral and cellular
DNA sequences in ICF syndrome cells. Hum. Mol. Genet., 11, 2091–2102.
34. Li,L. et al. (2010) The tumor suppressor UCHL1 forms a complex with
p53/MDM2/ARF to promote p53 signaling and is frequently silenced in
nasopharyngeal carcinoma. Clin. Cancer Res., 16, 2949–2958.
35. Li,L. et al. (2011) The human cadherin 11 is a pro-apoptotic tumor suppressor modulating cell stemness through Wnt/beta-catenin signaling and
silenced in common carcinomas. Oncogene, 23, 3901–3912.
36. Chen,L. et al. (2011) Genomic selection identifies vertebrate transcription factor Fezf2 binding sites and target genes. J. Biol. Chem., 286,
18641–18649.
37. Hui,A.B. et al. (2005) Array-based comparative genomic hybridization
analysis identified cyclin D1 as a target oncogene at 11q13.3 in nasopharyngeal carcinoma. Cancer Res., 65, 8125–8133.
38. Sozzi,G. et al. (1996) The FHIT gene 3p14.2 is abnormal in lung cancer.
Cell, 85, 17–26.
39. Huebner,K. et al. (2003) Cancer and the FRA3B/FHIT fragile locus: it’s a
HIT. Br. J. Cancer, 88, 1501–1506.
40. Loyo,M. et al. (2011) A survey of methylated candidate tumor suppressor
genes in nasopharyngeal carcinoma. Int. J. Cancer, 128, 1393–1403.
41. Della Peruta,M. et al. (2010) Protein tyrosine phosphatase receptor type
{gamma} is a functional tumor suppressor gene specifically downregulated
in chronic myeloid leukemia. Cancer Res., 70, 8896–8906.
42. Shu,S.T. et al. (2010) Function and regulatory mechanisms of the candidate
tumor suppressor receptor protein tyrosine phosphatase gamma (PTPRG)
in breast cancer cells. Anticancer Res., 30, 1937–1946.
43. Sheu,J.J. et al. (2009) Chromosome 3p12.3-p14.2 and 3q26.2-q26.32 are
genomic markers for prognosis of advanced nasopharyngeal carcinoma.
Cancer Epidemiol. Biomarkers Prev., 18, 2709–2716.
44. Jones,P.A. et al. (2002) The fundamental role of epigenetic events in cancer. Nat. Rev. Genet., 3, 415–428.
45.Robertson,K.D. (2005) DNA methylation and human disease. Nat. Rev.
Genet., 6, 597–610.
46. Shu,X.S. et al. (2012) Chromatin regulators with tumor suppressor
properties and their alterations in human cancers. Epigenomics, 4,
537–549.
47. Rhodes,D.R. et al. (2004) ONCOMINE: a cancer microarray database and
integrated data-mining platform. Neoplasia, 6, 1–6.
48. Song,I.S. et al. (2011) Human ZNF312b oncogene is regulated by Sp1
binding to its promoter region through DNA demethylation and histone
acetylation in gastric cancer. Int. J. Cancer, 129, 2124–2133.
Received December 7, 2012; revised May 4, 2013; accepted May 10, 2013
1993