[CANCER RESEARCH 64, 8980 – 8986, December 15, 2004]
Cyclin G2 Dysregulation in Human Oral Cancer
Yong Kim,1 Satoru Shintani,2 Yohko Kohno,3 Rong Zhang,1 and David T. Wong1
1
School of Dentistry and Dental Research Institute, University of California at Los Angeles, Los Angeles, California; 2Department of Oral and Maxillofacial Surgery, Ehime
University School of Medicine, Ehime, Japan; and 3Department of Oral Pathology, School of Dentistry, Showa University, Tokyo, Japan
ABSTRACT
Using expression microarray, we have previously shown that human
cyclin G2 (hCG2) is significantly down-regulated in laser capture microdissected oral cancer epithelia. Western analysis showed detectable hCG2
protein in normal (2 of 2) but not in malignant (4 of 4) oral keratinocyte
cell lines. Immunohistochemistry analysis done on oral cancers showed
that normal oral mucosa (100%, 12 of 12) and 69.1% (47 of 68) of
dysplastic oral epithelia expressed readily detectable hCG2 in the nuclei.
However, only 11.1% of oral cancer epithelia (14 of 126) showed mild
hCG2 nuclear staining. Interestingly, of the oral cancers devoid of nuclear
hCG2 (112 cases), 58 cases (52%) showed cytoplasmic hCG2 immunostaining, whereas the other 54 cases (48%) exhibited neither nuclear nor
cytoplasmic hCG2 staining. In vitro functional study by ectopic restoration
of hCG2 expression in the human malignant squamous cell carcinoma
(SCC) line SCC15 resulted in a significant inhibition of cellular proliferation (P < 0.001) and colony formation (P < 2 ⴛ 10ⴚ5) with increased
population of G1 phase and decreased in S phase (P < 0.01). Furthermore,
stable down-regulation of hCG2 by short interference RNA-based gene
silencing in immortalized normal oral keratinocytes resulted in enhanced
cell growth with increase in S and prominently in G2 phase. Because hCG2
has been implicated as a negative regulator in cell cycle progression, our
results support that hCG2 dysregulation may play an important role in
epithelial transformation and the early stages of human oral cancer
development.
INTRODUCTION
Although molecular progression models exist for oral carcinogenesis, the precise molecular targets and pathways remain unclear (1).
Chromosomal structural alterations are associated with dysplasia
(9p21, 3p21, and 17p13), carcinoma in situ (11q13, 13q21, and
14q31), and invasive carcinoma (4q26-28, 6p, 8p, and 8q; refs. 2, 3).
Recently, there has been an explosion of gene expression studies
associated with oral/head and neck carcinogenesis; however, few have
been consistently identified or functionally implicated with the oral
cancer phenotype. Certain gene alterations have a stronger association
with oral cancer development. Among these include p16ink4a, cyclin
D1, and p53, with 80%, 30%, and 50% involvement in head and neck
cancers, respectively (3). All three are located in chromosomal regions linked to oral cancer development: 9p21 (p16ink4a), 11q13
(cyclin D1), and 17p13 (p53). Several molecular events, such as
altered expression of p12CDK2-AP1 and retinoic acid receptor-␤, have
been associated with oral cancer development that has no obvious
structural chromosomal change (4, 5). Likely, there are several genes/
pathways yet to be identified as having a true causal relationship with
oral cancer development. Identification of potentially novel diagnostic
and therapeutic targets is now being addressed by DNA hybridization
arrays (6).
Cyclin G2 (CG2) is a newly identified homologue of cyclin G1 and
exhibits 60% nucleotide sequence identity and 53% amino acid sequence identity with cyclin G1, which was identified as one of the
transcriptional targets of p53 (7, 8). Murine CG2 has been shown to
be a cytoplasmic protein, whereas cyclin G1 is largely localized to the
nucleus (9). CG2 contains a protein destabilizing PEST-rich sequence
and a potential Shc phosphotyrosine binding site, implying its cell
cycle dependent temporal level. Cyclin G1 is constantly expressed
throughout the cell cycle, whereas CG2 expression peaks in the S
phase. Despite their significant sequence homology, cyclin G1 and G2
are believed to have distinct cellular functions. CG2 has been implicated in negative selection of self-reactive lymphocytes, apoptosis,
and also in DNA damage repair (10). Growth inhibitors, such as
transforming growth factor ␤1 or dexamethasone, significantly induce
CG2 mRNA expression. Further studies with B-cell antigen receptormediated cell cycle arrest have shown that CG2 may be a key negative
regulator of cell cycle progression (11). A recent study by Bennin et
al. (12) showed that CG2 inhibits cell cycle progression through
interaction with PP2A. Ectopic expression of CG2 induces the formation of aberrant nuclei and cell cycle arrest. The mechanism of
CG2 regulation and cellular function in carcinogenesis is poorly
understood, and its dysregulation during oral carcinogenesis has not
been described previously. Ito et al. (13) have recently reported
decreased expression of CG2 in papillary carcinoma of the thyroid. It
has also been shown that the expression of CG2 is down-regulated in
estrogen-treated human breast cancer cells, implying its possible
antiproliferative function (14). More recently, CG2 mRNA levels
were shown to be elevated in G0, reduced as cells enter the cell cycle,
and elevated again at the late S and G2-M phases (15). It was also
shown that the expression of CG2 mRNA could be regulated by the
FoxO family of forkhead transcription factors (15).
Using expression microarray, we recently identified down-regulation of CG2 expression is associated with the development of human
oral cancers (6). Herein, we report the validation results showing
down-regulation of CG2 in oral cancer cell lines and dysregulation of
CG2 in oral cancer tissues by immunohistochemical analysis. Furthermore, we investigated the functional role of CG2 in epithelial cell
growth examined in vitro. These data present the first evidence that
alteration of CG2 expression can be functionally associated with
oral/head and neck cancer development, and CG2 may serve as a
molecular target for diagnostic and therapeutic applications.
MATERIALS AND METHODS
Materials. Cell culture media (DMEM, DMEM-F12, and KSFM) and cell
culture supplies including antibiotics were purchased from Invitrogen (Carlsbad, CA). Fetal bovine serum was obtained from HyClone (Logan, UT). The
LipofectAMINE transfection reagent was from Invitrogen and FuGene6 was
Received 6/1/04; revised 9/1/04; accepted 10/6/04.
from Roche Diagnostics (Indianapolis, IN). All restriction enzymes used were
Grant support: Public Health Service Grant RO1 DE14857 (D. Wong) and T32
purchased from New England Biolabs (Beverly, MA). The anti-CG2 antibody
DE007296-08 (Y. Kim).
was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and the antiThe costs of publication of this article were defrayed in part by the payment of page
FLAG antibody was from Sigma-Aldrich (Sigma-Aldrich, St. Louis, MO). The
charges. This article must therefore be hereby marked advertisement in accordance with
secondary antimouse IgG-horseradish peroxidase and enhanced chemilumines18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: David T. Wong, Dental Research Institute, School of Dentistry,
cence reagent were purchased from Amersham (Piscataway, NJ). The antigoatUniversity of California at Los Angeles, 10833 Le Conte Avenue, 73-017 Center for Health
horseradish peroxidase antibody was from Sigma-Aldrich. Reagents used for
Sciences, Los Angeles, CA 90095. Phone: (310) 206-3048; Fax: (310) 794-7109; E-mail:
quantitative PCR and reverse transcription (RT)-PCR were from Bio-Rad
[email protected].
(Hercules, CA) and Perkin-Elmer (Boston, MA). The CCK-8 kit was pur©2004 American Association for Cancer Research.
8980
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2004 American Association for Cancer
Research.
DOWN-REGULATION OF CG2 IN HUMAN ORAL CANCER
reverse transcribed to synthesize the first strand of cDNA with SuperScriptII
RT (Invitrogen). The resulting cDNA was then subjected to PCR amplification
by using a primer set specific for hCG2 coupled with proper restriction site
sequences (EcoR I and XbaI) to facilitate subcloning (forward, 5⬘GCGAATTCAATGAAGGATTTGGGGGCAGAG3⬘; and reverse, 5⬘GCTCTAGACTAAGATGGAAAGCACAGTGT3⬘). The PCR product was subsequently
subcloned into the pFLAG-CMV2 cloning vector (Sigma), and selected clones
were sequence verified by automatic capillary sequencing (ABI 3100; UCLA
Department of Human Genetics DNA Sequencing Core Facility, Los Angeles,
CA). Our CG2 clone has 99% nucleotide homology with previously reported
studies, including the 1035 bp open reading frame encoding the 345 AA
protein of Mr ⬃40,000 (17). There was one silent mutation (T-⬎C substitution)
found at ⫹117 that does not alter the putative amino acid sequence.
Generation of CG2 Short Interfering RNA (siRNA) Constructs for
Gene Silencing. Several hCG2 siRNA target sequences were designed and
cloned into pSilencer 2.1-U6 hygro and pSilencer 3.1-H1 hygro vectors (Ambion). Clones were selected and verified by DNA sequencing. Plasmid DNA
for transfection was prepared by using Qiagen Endofree Maxi kit (Qiagen), and
test transfection was carried out with HeLa cells. The results were analyzed by
RT-PCR, and two different siRNA target sites for the hCG2 gene were selected
[sihCG2.1 (⫹570 -⬎ ⫹588: AGCTTGCAACTGCCGACTC) and sihCG2.2
(⫹750 -⬎ ⫹768: ATGCCTAGCCGAGTATTCT)].
Cell Culture and Transfection Studies. For ectopic overexpression of
hCG2 in SCC15, the cells were grown in DMEM/F12 mixed with a conditioned medium from NIH3T3 cell (1:1). A day before transfection, the cells
were plated on a 60-mm dish and transfected by using Lipofectin combined
with Plus reagent according to the manufacturer’s protocol (Invitrogen) with a
minor modification. After 5 hours transfection, the cells were fed with fresh
complete medium and incubated for 48 hours before analysis. To downregulate hCG2 by siRNA in immortalized human normal oral keratinocytes
OKF6tert1 (provided by Dr. Jim Rheinwald from Harvard Medical School,
Boston, MA), the cells were grown and maintained in KSFM medium (25
␮g/mL bovine pituitary extract, 0.4 mmol/L CaCl2, 0.5 ng/mL epidermal
growth factor, 1% penicillin, and 0.1% streptomycin). A day before transfection, the cells were plated on a 60-mm dish in the same medium without
antibiotics. The pSilencer siRNA construct (sihCG2.1 or sihCG2.2) was transfected by using FuGene 6 transfection reagent according to the manufacturer’s
instructions (Roche Diagnostics Corp.). Stable sihCG2 cell lines along with
control lines were established by selecting transfectants with 50 ␮g/mL hygromycin B (Invitrogen) for 3 weeks. The resulting clones were maintained in
1 ␮g/mL hygromycin.
Cellular Growth and Fluorescence-activated Cell Sorter Analysis. For
cell growth analysis, cells were seeded on a 96-well plate, and the cellular
proliferation was measured by using the CCK-8 kit (Alexis Biochemical Corp.)
following manufacturer’s instructions with a minor modification as follows.
Briefly, on the day of assay, the cells were fed with fresh medium and
preincubated for 1 hour at 37°C. The CCK-8 assay reagent (10% of the cell
culture medium) was added to each well and mixed gently. After 1 hour
incubation at 37°C, 1% SDS (1 of 10 volume of the culture medium) was
added, and the A450 nm was measured. All of the assays were done in triplicate,
and the growth rate was presented as the fold-increase over the day 1 value. For
colony forming analysis, transiently transfected cells were harvested after 48
hours transfection and plated on a 60-mm dish at 50 cells per dish. After a
14-day incubation, colonies were stained with methylene blue, washed with
deionized water, and counted. For cell cycle profiling by fluorescence-activated
cell sorter (FACS) analysis (UCLA Janis V. Giorgi Flow Cytometry Core Facility,
Los Angeles, CA), the transiently or stably transfected cells from 100-mm dishes
were harvested by trypsinization and washed twice with PBS. After resuspending
in 500 ␮L of ice-cold PBS, the cells were fixed by mixing with 5 mL of cold 95%
EtOH and kept at 4°C overnight. The cells were pelleted and washed two times
with PBS/1% fetal bovine serum and finally resuspended in 800 ␮L of PBS/1%
fetal bovine serum. The cell suspension was mixed with 100 ␮L of 10⫻ propidium
iodide [500 ␮g/mL in 38 mmol/L sodium citrate (pH7.0); Sigma] and 100 ␮L of
boiled RNaseA [10 mg/mL in 10 mmol/L Tris-HCl (pH7.5); Calbiochem]. After
incubating for 30 minutes at 37°C, the cell cycle profile was analyzed by FACS
analysis.
8981
chased from Alexis Biochemicals (San Diego, CA), and propidium iodide was
purchased from Sigma-Aldrich. RNaseA was obtained from Calbiochem (San
Diego, CA), and all of the other chemical reagents were purchased from
Sigma-Aldrich.
Northern and RT-PCR Analysis. Human tissues preserved in RNAlater
(Ambion, Austin, TX) or actively growing cell cultures were used to isolate
total RNA by using RNeasy kit (Qiagen, Valencia, CA) according to manufacturer’s instructions. For Northern analysis, equal amount of total RNA was
used and probed with CG2 cDNA as described previously (16). RT-PCR was
done for CG2 (forward, 5⬘-GAGAGCTGCTGAGACTCTCCTCTCC-3⬘; and
reverse, 5⬘-CCATCTGTATTAGCCTTGTGCCTTCTC-3⬘) and glyceraldehyde-3-phosphate dehydrogenase [(GAPDH) forward, 5⬘-AGTCCACTGGCGTCTTCACC-3⬘; and reverse, 5⬘-TGGTTCACACCCATGAC GAA-3⬘]
as follows: 1 ␮g of total RNA in 9 ␮L is mixed with 1 ␮L of 10 mmol/L
deoxynucleoside triphosphate and 1 ␮L of oligodeoxythymidylate and heated
for 5 minutes at 65°C. After chilling on ice for 3 minutes, 4 ␮L of 5⫻
first-strand buffer, 2 ␮L of 0.1 mol/L DTT, and 2 ␮L of 20 units/␮l RNase
inhibitor was added. After 2 minutes incubation at 42°C, 1 ␮L of SuperscriptII
RT (5 units/␮l) was added and incubated for 50 minutes at 42°C, 15 minutes
at 72°C, and soaked at 4°C. For PCR amplification, 1 ␮L of RT product was
mixed with 5 ␮L of 10⫻ PCRII buffer, 3 ␮L of 25 mmol/L MgCl2, 2.5 ␮l of
10 mmol/L deoxynucleoside triphosphate, 1 ␮L of primer mix (10 ␮mol/L
each), and 0.25 ␮1 of Taq (5 units/␮l) in 50 ␮L of reaction mixture. After 5⬘
heating at 94°C, amplification was carried out for 35 cycles in 94°C 45
seconds; 60°C 45 seconds; and 72°C 2 minutes.
Western Analysis. Cell lysate was isolated and homogenized in 10
volumes of radioimmunoprecipitation assay buffer [10 mmol/L Tris-HCl
(pH 7.5) 150 mmol/L NaCl; 5 mmol/L EDTA; 0.5% SDS; 1.0% NP40; and
1.0% sodium deoxycholate-containing protease inhibitors]. The homogenate was incubated on ice for 30 minutes and centrifuged for 15 minutes at
14,000 rpm, 4°C. The supernatant was recovered and kept at ⫺80°C. Equal
amounts of cell tissue lysate were mixed with SDS-PAGE sample buffer
and boiled for 5 minutes before loading onto a discontinuous SDS-PAGE
gel. After electrophoresis, the gel was incubated in 3-(cyclohexylamino)propanesulfonic acid (CAPS) buffer [10 mmol/L 3-cyclohexylamino-1propane-sulfonic acid (pH 11.0) 10% methanol] for 5 minutes, and proteins
were electroblotted onto a polyvinylidene difluoride membrane (Bio-Rad
Laboratories) in CAPS buffer. The blot was then blocked by incubating in
5% nonfat dried milk/TBS-Tween for 1 hour at room temperature and
incubated overnight with a polyclonal goat antihuman CG2 antibody (1:500
in 5% nonfat dried milk/TBS-Tween; Santa Cruz Biotechnology) at 4°C.
On the next day, the blot was washed 3⫻ 20 minutes in TBS-Tween at
room temperature and additionally incubated with a secondary antigoat
immunoglobulin conjugated with horseradish peroxidase (Sigma, St. Louis,
MO; 1:1,000 in 5% nonfat dried milk/TBS-Tween) for 1 hour at room
temperature. After washing 3⫻ 30 minutes in TBS-Tween at room temperature, the signal was detected by using enhanced chemiluminescence
(Amersham Life Science, Arlington Heights, IL). Loading of protein was
normalized by reprobing with a ␤-actin polyclonal antibody (Sigma).
Immunohistochemistry Analysis. Four-micron sections were cut from the
specimens, deparaffinized in xylene, and rehydrated in graded alcohol. Endogenous peroxidase was blocked by incubating tissue sections in 0.1%
hydrogen peroxidase in absolute methanol for 20 minutes. Before applying the
antihuman CG2 antibody, sections were heated in an autoclave for 15 minutes
in citrate buffer (pH 6.0). The CG2 antibody was used at 1:500 dilution, and
the incubation time was 16 hours at 4°C. Immunodetection was done with the
Envision system (DAKO, Carpinteria, CA). Peroxidase activity was visualized
by applying diaminobenzidine chromogen, containing 0.05% hydrogen peroxidase. The sections were counterstained with methyl green, dehydrated, and
mounted. Negative control staining was carried out by substituting nonimmune
serum for primary antibodies. Immunostaining was evaluated by three independent
observers. For each section, 10 high-power fields were chosen, and a total of at
least 1,000 cells were evaluated. The results were expressed as the percentage of
positive cells counted. To confirm the reproducibility, 25% of the slides were
chosen randomly and scored twice. All duplicates were similarly evaluated.
Cloning of the Full-length Human CG2 (hCG2) Coding Sequence. On
the basis of the published sequence (17), we have cloned the full-length hCG2
coding region with a PCR-based approach. OKF4 human normal oral keratinocyte total RNA was primed with an oligodeoxythymidylate primer and
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2004 American Association for Cancer
Research.
DOWN-REGULATION OF CG2 IN HUMAN ORAL CANCER
Fig. 1. Expression microarray data showed down-regulation of hCG2 in oral cancers.
The hybridization intensity of CG2 from the expression microarray analysis was normalized against the intensity of ␤-actin. The relative intensity of the five-paired cases of oral
SCC was compared, and there was a significant reduction of CG2 expression level in
tumors compared with normal cases. The case number 2 was found to be a bony invasive
variant form of oral SCC.
RESULTS
Expression Microarray Analysis on Human Oral Cancers
Showed Down-regulation of CG2. In an attempt to establish a
molecular gene expression profile in oral/head and neck cancer and
concurrently to identify specific gene sequences the expression of
which are differentially regulated in oral cancer, we have previously
done expression microarray analysis on laser capture microdissected
normal and oral cancer tissues (6, 18). Through extensive analyses of
Affymetrix GeneChip microarray data, we have identified about 600
genes that are associated with oral cancer. These include oncogenes,
tumor suppressors, transcription factors, xenobiotic enzymes, metastatic proteins, differentiation markers, and also genes that have not
been implicated in oral cancer (6). Additional validation of the expression microarray data led us to select genes that are linked to cell
cycle regulation. Fig. 1 shows that hCG2 is significantly downregulated (4 of 5 cases; P ⬍ 0.0375) in oral cancer tissues. It should
be noted that we later found that the case number 2 was a bony
invasive variant of oral squamous cell carcinoma (SCC) and thus
clinically behaves differently. CG2 was initially identified as a homologue of cyclin G1, and it has been implicated in the regulation of
G2-M phase. The growth inhibitory effect of CG2 has also been
reported, but the detailed cellular function of CG2 has not been
established in general, and certainly not in oral cancer. The significance of CG2 in cancer development has recently been explored. Ito
et al. (13) reported that CG2 expression is significantly down-regulated in papillary carcinoma of the thyroid. Our microarray data
strongly support CG2 expression is down-regulated in human oral
cancer.
Validation of CG2 Dysregulation in Oral SCC In vitro and In
vivo. Additional validation by the real-time PCR on the laser capture
microdissected samples confirmed the reduction of CG2 expression in
cancer tissues (data not shown). In addition, Northern analysis on
paired normal and oral cancer tissues revealed that CG2 mRNA was
down-regulated or undetectable in 5 independent pairs of human oral
cancer (Fig. 2A). Findings from microarray studies were additionally
validated in cultured oral keratinocytes. Establishing cell model system is important and valuable for the functional analysis and also to
understand the mechanism of altered gene expression. To confirm
dysregulation of CG2 in oral cancer cells, human normal oral keratinocytes and malignant SCC lines were cultured in serum-free KSFM
medium. Total RNA was prepared from growing normal oral keratinocytes (OKB2 and OKF4) and malignant SCC lines (SCC13, 15, and
25) and analyzed for CG2 mRNA expression by semiquantitative
RT-PCR. The result showed that CG2 mRNA expression is dramatically decreased in SCC lines compared with normal oral keratinocytes (Fig. 2B). We also examined the level of CG2 protein by
Western analysis. Total protein lysate was prepared from normal and
malignant oral keratinocytes and subjected to immunodetection of
CG2. As shown in Fig. 2C, normal oral keratinocytes (OKF4 and
OKF6) express readily detectable and comparable levels of CG2,
whereas all of the malignant oral SCC lines (SCC4, 15, 66, and 105)
do not express detectable levels of CG2 protein. To examine the in
vivo expression and cellular localization of CG2 protein in oral cancer
tissues, immunohistochemistry was done on 12 normal human oral
mucosa, 68 epithelial dysplasia (mild to severe), and 126 oral cancer
tissues obtained from patients. There was strong CG2 staining in the
nuclei in normal oral mucosa, but there was a progressive loss of
nuclear CG2 staining and a transition to cytoplasmic localization of
CG2 as cancer progressed (Fig. 3). This finding differs from the
previously reported cytoplasmic localization of murine CG2 (9, 19).
However, those findings were based on the in vitro ectopic expression
Fig. 2. Dysregulation of CG2 expression in vivo
and in vitro. A. Northern analysis on paired normal
(N) and oral cancer (T) tissues revealed that CG2
mRNA was down-regulated or undetectable in five
independent pairs of human oral cancer. B. RTPCR analysis showed a decreased CG2 mRNA in
oral SCC lines. Total RNA was prepared from
human normal (OKB2 and OKF4) and malignant
(SCC13, SCC15, and SCC25) oral keratinocytes,
and 1 ␮g of total RNA was used for RT-PCR
analysis by using a specific primer set for hCG2. A
primer set for GAPDH was used to normalize the
amplification. C. Western analysis of human normal and malignant oral keratinocytes for CG2 protein production was done. Twenty-five micrograms
of lysate was prepared for each sample. The blot
was incubated with 1:500 dilution of polyclonal
antihuman CG2 antibody overnight at 4°C. A Mr
40,000 signal, the expected size of hCG2, was
detected in the lysate of the two normal human oral
keratinocyte cell lines (OKF4 and OKF6) but not
in malignant human oral keratinocytes (SCC4,
SCC15, SCC66, and SCC105). ␤-actin (Mr 42,000)
was probed on the same membrane to correct for
sample loading.
8982
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2004 American Association for Cancer
Research.
DOWN-REGULATION OF CG2 IN HUMAN ORAL CANCER
nuclear CG2 (112 cases), 58 cases (52%) showed cytoplasmic CG2
immunostaining, whereas the other 54 cases (48%) exhibited neither
nuclear nor cytoplasmic CG2 staining (Fig. 3B). These data jointly
suggest that CG2 expression is dysregulated during human oral cancer
development. There is a progressive loss of CG2 expression during
oral carcinogenesis and an altered cellular localization from nuclear in
normal cells to cytoplasmic to no expression in cancer cells. Because
CG2 has been implicated as a negative regulator in cell cycle progression, our results suggest that CG2 dysregulation may play a role
in the early stages of human oral cancer development.
Ectopic Expression of CG2 in Oral SCC Line Inhibited Growth
and Colony Formation with Increased G1 Phase Cells. Previous
studies on possible cellular functions of CG2 have shown that ectopic
expression of murine CG2 in HEK293 and CHO cell lines induced
growth arrest in G1 phase with decreased population in S and G2
phases after 21 hours post-transfection (12). It was also reported that
ectopic expression of CG2 resulted in aberrant and fragmented nuclei.
After longer times in culture, a broad G1-S–phase distribution and
alteration in G2-M phase profile suggestive of possible aneuploidy
were observed (12). Overexpression of hCG2 in HeLa cells also
resulted in reduced colony formation (20). As an indirect way of
assessing the functional consequences of down-regulation of CG2 in
oral cancers, hCG2 was ectopically overexpressed in a SCC line
(SCC15) that does not show expression of endogenous hCG2. After
transiently transfecting the pFLAG-CMV2– hCG2 into SCC15 cells
along with a control null vector, the hCG2 transfectants were analyzed
by cell proliferation analysis, colony forming assay, and FACS analysis and compared the results to control transfectants. Western analysis was used to confirm the expression of the exogenous FLAGtagged hCG2. Total cell lysate was prepared from transiently
transfected cells and immunoblotted with anti-CG2 antibody (Santa
Cruz Biotechnology). As shown in Fig. 4A, the expression of recombinant FLAG-hCG2 was detected by anti-CG2 or anti-FLAG antibody
Fig. 3. Immunohistochemical analyses of CG2 expression in human oral cancers. A.
The in vivo expression of CG2 was examined by immunohistochemistry with anti-CG2
antibody (Santa Cruz Biotechnology). Normal human oral mucosa and dysplastic oral
epithelia expressed detectable CG2 in the nuclei. But CG2 expression was greatly reduced
in cancer tissues. For negative control, the slide was stained with secondary antibody only.
B. The percentage of positive cells in each case was semiquantitatively evaluated into one
of the following five groups: (a) immunoreactivity completely absent (negative, 0%); (b)
⬍5%; (c) ⬍25%; (d) ⬍50%; and (e) up to 100%. In the present study, cases showing
⬎5% of nuclear positive cells were defined as “nuclear positive.” And cases showing
⬎5% of cytoplasmic positive cells were defined as “cytoplasmic positive.” The result was
examined by Fisher’s exact test. Normal epithelia and dysplasia: P ⫽ 0.0300, dysplasia
versus SCC: P ⬍ 0.0001. (OSCC, oral SCC)
Table 1
Expression of CG2 protein in normal, premalignant, and oral cancer
Dysplasia
CG2 staining
Normal
(n ⫽ 12)
Mild
(n ⫽ 28)
Moderate
(n ⫽ 19)
Severe
(n ⫽ 23)
SCC
(n ⫽ 126)
Nuclei
Cytoplasm
Negative
12 (100%)
0 (0%)
0 (0%)
24 (85.7%)
0 (0%)
4 (14.3%)
14 (73.7%)
1 (5.3%)
4 (21.0%)
9 (39.1%)
3 (13.0%)
11 (47.8%)
14 (11.1%) *
58 (46.0%) †
54 (42.9%)
NOTE. Data was analyzed by using StatView software (Cary, NC) and P values were
determined by using ␹2 test.
* P ⫽ 0.00022.
† P ⫽ 0.0027.
of the exogenous recombinant murine CG2 in cultured cells, which
could be different from the in vivo environment. As summarized in
Table 1 and depicted in Fig. 3B, all of the normal oral mucosa
examined (100%, 12 of 12) and 69.1% (47 of 68) of the dysplastic oral
epithelial expressed readily detectable CG2 in the nuclei. However,
only 11.1% of oral cancer epithelia (14 of 126) showed mild CG2
nuclear staining. Thus, 88.8% of oral cancers (112 of 126) showed
loss of nuclear CG2. Interestingly, of the oral cancers devoid of
Fig. 4. Transient ectopic overexpression of CG2 in SCC line SCC15 inhibited cellular
proliferation and colony formation. A. The oral SCC line SCC15 that does not show
expression of endogenous hCG2 was transiently transfected with pFLAG-hCG2 (Lane 3)
along with control construct (Lane 1: no transfection and Lane 2: pFLAG). Western
analysis showed overexpression of FLAG-hCG2 detectable by both anti-hCG2 and
anti-FLAG antibody. B. The cellular proliferation of transiently transfected cells was
monitored by using the Cell Counting kit (CCK-8, Alexis Biochemical), and overexpression of hCG2 in SCC15 cells induced significant inhibition (P ⬍ 0.001) of cell growth
compared with null transfectant (pFLAG). C. pFLAG-hCG2 transfectant showed greatly
reduced colony forming cells compared with the null transfectant (P ⬍ 2 ⫻ 10-5). Bars,
⫾SD.
8983
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2004 American Association for Cancer
Research.
DOWN-REGULATION OF CG2 IN HUMAN ORAL CANCER
Fig. 5. Induced apoptosis and cell cycle arrest by hCG2 in SCC15 cells. A. Transient
ectopic expression of CG2 induced significant apoptosis in SCC15 cells after 24 hours (*,
P ⫽ 0.00053) and 48 hours (**, P ⬍ 0.0001). B. Cell cycle profiling by FACS analysis
showed that overexpression of hCG2 in SCC15 resulted in an increase in G1 population
at 24 hours (*, P ⫽ 0.01) and 48 hours post-transfection (**, P ⫽ 0.007). The data were
analyzed by using paired student’s t test. Bars, ⫾SD.
only in cells transfected with the pFLAG-hCG2 plasmid (Lane 3).
Note that endogenous hCG2 was not detectable in SCC15. The level
of hCG2 achieved by transient transfection was ⬃3-fold higher than
an endogenous level of CG2 in normal oral keratinocytes. The effect
of hCG2 restoration in the growth of SCC15 cells is shown in Fig. 4B.
Overexpression of FLAG-hCG2 in SCC15 greatly reduced cell proliferation rates (P ⬍ 0.001; Fig. 4B) and also induced a floating cell
population, suggesting that CG2 overexpression may induce apoptosis. We also measured clonogenicity of the transient transfectants, and
ectopic expression of CG2 resulted in significantly reduced formation
(P ⬍ 2 ⫻ 10-5) of methylene blue stained colonies as shown in Fig.
4C. To examine if ectopic expression of CG2 induced apoptosis, we
measured percentage of apoptotic cell population at 24 and 48 hours
post-transfection by using Annexin V-EGFP FACS analysis. Consistent with the observation of increased floating cells in hCG2 transfectant, we observed significantly induced apoptosis after hCG2 transfection compared with null vector transfection (P ⫽ 0.00053 at 24
hours and P ⬍ 0.0001 at 48 hours; Fig. 5A). Also, to examine if
ectopic expression of CG2 altered the cell cycle profile in SCC15
cells, we determined the cell cycle positions of the transient transfectants at 24 and 48 hours post-transfection. Compared with control
transfectants, hCG2-overexpressing cells showed a decrease in S
phase and an increase in G1 phase at 24 hours post-transfection
(P ⫽ 0.01), and the effect was more dramatic at 48 hours posttransfection (P ⫽ 0.007; Fig. 5B). These data supports the growth
inhibitory function of CG2.
Reduction of CG2 Expression in Human Normal Oral Keratinocyte Resulted in an Increased Cellular Proliferation. Our initial
expression microarray analysis and additional follow-up studies
showed that there is a significant down-regulation of CG2 expression
in both mRNA and protein level in oral cancer tissues and squamous
cell carcinoma lines. Therefore, it will be important to examine the
effect of down-regulation of CG2 level in normal oral tissues or cell
lines. As a way to address this, we examined the effect of downregulation of CG2 in normal oral keratinocytes. The siRNA was used
to specifically down-regulate CG2 in an immortalized human normal
oral keratinocyte (OKF6tert1) cell line. OKF6tert was established by
expressing hTERT in normal oral keratinocyte (OKF6) to rescue cells
from senescence (21). OKF6tert1 cells express readily detectable
levels of CG2. The hCG2 siRNA sequences (sihCG2) were designed
by using the Ambion target finder web-based program, and their
specificity was analyzed by NCBI BLAST search. Selected hCG2
siRNA sequences were prepared by DNA synthesis (MWG Biotech
Inc., High Point, NC) and subsequently cloned into pSilencer 2.1-U6
hygro and pSilencer 3.1-H1 hygro vectors (Ambion). The effects of
chosen siRNA sequences were initially tested by transiently transfecting them into HeLa cells. Reduction of hCG2 mRNA was evaluated
by semiquantitative RT-PCR and Northern analysis (data not shown).
The selected clones from hCG2.1 siRNA construct were additionally
analyzed by Western analysis with anti-CG2 antibody (Santa Cruz
Biotechnology) to assess the extent of gene silencing by siRNA. As
shown in Fig. 6A, compared with control clones [vector alone (V) or
negative siRNA with random scrambled sequence (N)], sihCG clones
showed various reduced levels of hCG2 expression (from 30 to 70%
reduction after normalized against ␤-actin; P ⬍ 0.05). Clones with
significant reduction of hCG2 were selected, and the effect of hCG2
down-regulation on cell growth was examined by cellular proliferation assay. As shown in Fig. 6B, the sihCG2 clones showed increased
rate of cellular proliferation (1.3- to 1.5-fold; P ⬍ 0.03) compared
with siRNA negative control clones. We have additionally examined
the cell cycle profiles of these sihCG2 clones by DNA FACS analysis.
As shown in Fig. 6C, the sihCG2 clones showed significant alteration
in cell cycle profile (P ⬍ 0.003). There was a decrease in G1 phase (17
to 36% decrease compared with controls) along with increase in S
phase (18 to 33% increase compared with controls) and prominent
increase in G2 phase (0.9- to 2.6-fold increase compared with controls). Taken together, these data strongly support that CG2 plays a
physiologic role in epithelial cell growth and cell cycle control, and
alteration of its expression could significantly affect normal cellular
Fig. 6. Gene silencing of hCG2 by siRNA induced cellular proliferation in human
normal oral keratinocytes. The effect of gene silencing of hCG2 was examined by
generating stable siRNA clones from normal oral keratinocytes, OKF6tert1. A. Western
analysis on the selected stable clones showed significantly reduced expression of hCG2
when normalized against ␤-actin level (P ⬍ 0.05). B. Cellular proliferation of sihCG2
stable clones (hCG C3, C5, C9, and C11) was compared with the negative siRNA (H1N1
and H1N2) clones by using the Cell Counting Kit (CKK-8, Alexis Biochemical). The
result showed the increase in cellular proliferation by gene silencing of hCG2 (P ⬍ 0.03).
C. Cell cycle analysis on sihCG2 stable clones (C3, C5, C9 and C11) by FACS showed
decrease in G1 along with increase in S and G2 phases compared with vector controls (V1
and V2; P ⬍ 0.003). The data were analyzed by using paired student’s t test. Bars, ⫾SD.
8984
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2004 American Association for Cancer
Research.
DOWN-REGULATION OF CG2 IN HUMAN ORAL CANCER
physiology. Furthermore, our data supports that down-regulation of
CG2 is a functionally important event in cellular transformation of
oral epithelial cells.
DISCUSSION
Cancer biology is governed by alterations of complex pathways and
cellular programs. Many mechanisms potentially important to the
management of the cancer patient remain undiscovered. The dysregulation of the molecular events governing cell cycle control is emerging
as a central theme of oral carcinogenesis (22). Regulatory pathways
responding to extracellular signaling or intracellular stress converge
on the cell cycle apparatus. Abrogation of mitogenic and antimitogenic response regulatory proteins, such as the retinoblastoma tumor
suppressor protein (pRB), cyclin D1, cyclin-dependent kinase 6
(CDK6), and cyclin-dependent kinase inhibitors (p21WAF1/CIP1,
p27KIP1, and p16INK4a), occur frequently in human oral cancers (22).
Cellular response to metabolic stress or genomic damage through p53
and related pathways that block cell cycle progression are also altered
during oral carcinogenesis. In addition, new pathways and cell cycle
regulatory proteins, such as p12CDK2-AP1, are being discovered (4, 5).
By using expression microarray on laser capture microdissectednormal and oral cancer tissues, we have identified CG2 as a potentially important molecule in the development of oral cancer. As a
follow-up validation process, we have confirmed the dysregulation of
CG2 in malignant oral SCC lines in vitro and also down-regulation of
CG2 expression in vivo in oral cancer tissues. More importantly, the
functional significance of differential expression of CG2 in epithelial
cell growth has been addressed. Data presented in this report strongly
support that down-regulation of CG2 is a functionally significant
event in epithelial transformation and may have biological significance in the early development of oral cancer. Dysregulation of CG2
in cancers has been emerging recently; however, the functional significance of dysregulation of CG2 in cancer development has never
been addressed. CG2 may control cellular growth to maintain proper
balances in G2-M phase of cell cycle. As shown by Bennin et al. (12),
if CG2 is involved in proper control and maintenance of mitosis and
cytokinesis, the alteration of CG2 will result in genomic instability
from accumulation of aberrant DNA content (Fig. 7). Especially under
the circumstances that CG2 is altered to be down-regulated, there will
be a loss of growth control combined with alteration in downstream
interactions with other cellular partners, such as phosphatases or
kinases. This will result in cellular transformation through altered
mitosis/cytokinesis.
It remains to be tested if down-regulation of CG2 in vivo will
eventually have a significant effect on tumor growth. Our data
showing a growth inhibitory function of CG2 in a human malignant
oral SCC line set the stage for animal tumor model study to
evaluate the in vivo effect of CG2 on tumor regression. At this
point, it is not known mechanistically how CG2 expression is
down-regulated in oral cancers. Future studies should be directed
toward identification of any genomic alterations of CG2 (Ccng2)
gene that is located at chromosome 4q21.22, investigation on the
gene regulation mechanisms of CG2, and also the cellular mechanisms of CG2 function. Preliminary genomic loss of heterozygosity analysis done on a panel of normal and malignant oral keratinocytes has implied that there is a comparable degree of genomic
alteration at the Ccng2 locus.4 In addition, based on our findings
that CG2 mRNA level is differentially regulated in oral cancers in
vitro and in vivo, we are investigating a possible mechanism of
down-regulation of CG2 in human SCC lines at the transcriptional
4
X. C. Zhou, unpublished results.
Fig. 7. Working hypothesis for the role of CG2 in the epithelial cell cycle control. CG2
plays a role in the epithelial cell cycle control by modulating the progress through the
G2-M cycle. The molecule is induced at the late S phase and exerts its growth inhibitory
effect through interactions with other cellular partners to allow cells to properly progress
through mitosis. If the expression of CG2 is down-regulated by genetic alteration, the cell
loses control over G2-M phase and progresses through mitosis without proper cellular
control checks, resulting in genomic instability and downstream cellular transformation.
and post-transcriptional levels. To answer how dysregulation of
CG2 renders cellular growth inhibitory effect, we are now identifying CG2 interacting cellular partners.
The multistep process of oral carcinogenesis likely involves functional alteration of cell cycle regulatory members combined with
escape from cellular senescence and apoptotic signaling pathways
(23). Detailing the molecular alterations and understanding the functional consequences of the dysregulation of the cell cycle apparatus in
the oral cancer cells may uncover novel diagnostic and therapeutic
approaches. Our data strongly suggest that CG2 be one of the promising candidate genes.
REFERENCES
1. Califano A, Stolovitzky G, Tu Y. Analysis of gene expression microarrays for
phenotype classification. Proc Int Conf Intell Syst Mol Biol 2000;8:75– 85.
2. Califano J, van der Riet P, Westra W, et al. Genetic progression model for head and
neck cancer: implications for field cancerization. Cancer Res 1996;56:2488 –92.
3. Mao L. Leukoplakia: molecular understanding of pre-malignant lesions and implications for clinical management. Mol Med 1997;442– 8.
4. Tsuji T, Duh FM, Latif F, et al. Cloning, mapping, expression, function, and mutation
analyses of the human ortholog of the hamster putative tumor suppressor gene doc-1.
J Biol Chem 1998;273:6704 –9.
5. Papadimitrakopoulou V, Izzo J, Lippman SM, et al. Frequent inactivation of
p16INK4a in oral premalignant lesions. Oncogene 1997;14:1799 – 803.
6. Alevizos I, Mahadevappa M, Zhang X, et al. Oral cancer in vivo gene expression
profiling assisted by laser capture microdissection and microarray analysis. Oncogene
2001;20:6196 –204.
7. Sherr CJ. Mammalian G1 cyclins. Cell 1993;73:1059 – 65.
8. Okamoto K, Beach D. Cyclin G is a transcriptional target of the p53 tumor suppressor
protein. EMBO J 1994;13:4816 –22.
9. Jensen MR, Audolfsson T, Keck CL, Zimonjic DB, Thorgeirsson SS. Gene structure
and chromosomal localization of mouse cyclin G2 (Ccng2). Gene (Amst) 1999;230:
171– 80.
10. Bates S, Rowan S, Vousden KH. Characterisation of human cyclin G1 and G2: DNA
damage inducible genes. Oncogene 1996;13:1103–9.
11. Horne MC, Donaldson KL, Goolsby GL, et al. Cyclin G2 is up-regulated during
growth inhibition and B cell antigen receptor-mediated cell cycle arrest. J Biol Chem
1997;272:12650 – 61.
8985
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2004 American Association for Cancer
Research.
DOWN-REGULATION OF CG2 IN HUMAN ORAL CANCER
12. Bennin DA, Arachchige Don AS, Brake T, et al. Cyclin G2 associates with protein
phosphatase 2A catalytic and regulatory B’ subunits in active complexes and induces
nuclear aberrations and a G1/S phase cell cycle arrest. J Biol Chem 2002;15:15.
13. Ito Y, Yoshida H, Uruno T, et al. Decreased expression of cyclin G2 is significantly
linked to the malignant transformation of papillary carcinoma of the thyroid. Anticancer Res 2003;23:2335– 8.
14. Frasor J, Danes JM, Komm B, et al. Profiling of estrogen up- and down-regulated
gene expression in human breast cancer cells: insights into gene networks and
pathways underlying estrogenic control of proliferation and cell phenotype. Endocrinology 2003;144:4562–74.
15. Martinez-Gac L, Marques M, Garcia Z, Campanero MR, Carrera AC. Control of cyclin
G2 mRNA expression by forkhead transcription factors: novel mechanism for cell cycle
control by phosphoinositide 3-kinase and forkhead. Mol Cell Biol 2004;24:2181–9.
16. Kim Y, Tsuji T, Elovic A, et al. Murine doc-1 cDNA cloning, sequencing and
expression in normal adult tissues. Int J Oral Biol 2002;26:87–92.
17. Horne MC, Goolsby GL, Donaldson KL, et al. Cyclin G1 and cyclin G2 comprise a
new family of cyclins with contrasting tissue-specific and cell cycle-regulated expression. J Biol Chem 1996;271:6050 – 61.
18. Ohyama H, Zhang X, Kohno Y, et al. Laser capture microdissection-generated target
sample for high-density oligonucleotide array hybridization. Biotechniques 2000;29:
530 – 6.
19. Bennin DA, Arachchige Don AS, Brake T, et al. Cyclin G2 associates with protein
phosphatase 2A catalytic and regulatory B’ subunits in active complexes and
induces nuclear aberrations and a G1/S phase cell cycle arrest. J Biol Chem
2002;277:27449 – 67.
20. Tian YL, Liu FR, Liu J, et al. Ectopic expression of cyclin G2 inhibits cell proliferation in HeLa cancer cell line [in Chinese]. Ai Zheng 2002;21:577– 81.
21. Dickson MA, Hahn WC, Ino Y, et al. Human keratinocytes that express hTERT and
also bypass a p16(INK4a)-enforced mechanism that limits life span become immortal
yet retain normal growth and differentiation characteristics. Mol Cell Biol 2000;20:
1436 – 47.
22. Todd R, Hinds PH, Munger K, et al. Cell cycle dysregulation in oral cancer. Crit Rev
Oral Biol 2002;13:51– 61.
23. Polverini PJ, Nor JE. Apoptosis and predisposition to oral cancer. Crit Rev Oral Biol
Med 1999;10:139 –52.
8986
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2004 American Association for Cancer
Research.
Cyclin G2 Dysregulation in Human Oral Cancer
Yong Kim, Satoru Shintani, Yohko Kohno, et al.
Cancer Res 2004;64:8980-8986.
Updated version
Cited articles
Citing articles
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/64/24/8980
This article cites 20 articles, 7 of which you can access for free at:
http://cancerres.aacrjournals.org/content/64/24/8980.full.html#ref-list-1
This article has been cited by 5 HighWire-hosted articles. Access the articles at:
/content/64/24/8980.full.html#related-urls
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2004 American Association for Cancer
Research.