Download The transglutaminase 2 gene (TGM2), a potential

Carcinogenesis vol.29 no.3 pp.510–518, 2008
doi:10.1093/carcin/bgm280
Advance Access publication January 3, 2008
The transglutaminase 2 gene (TGM2), a potential molecular marker for
chemotherapeutic drug sensitivity, is epigenetically silenced in breast cancer
Lingbao Ai, Wan-Ju Kim, Berna Demircan, Lisa M.Dyer,
Kevin J.Bray, Ryan R. Skehan, Nicole A.Massoll1 and
Kevin D.Brown
Department of Biochemistry and Molecular Biology and University of Florida
Shands Cancer Center Program in Cancer Genetics, Epigenetics and Tumor
Virology and 1Department of Pathology, Immunology and Laboratory
Medicine, University of Florida College of Medicine, Box 100245,
Gainesville, FL 32610, USA
To whom correspondence should be addressed. Tel: 352 273 5458;
Fax: 352 392 1445;
Email: [email protected]
Tissue transglutaminase (TG2) is a ubiquitously expressed enzyme capable of catalyzing protein cross-links. TG2-dependent
cross-links are important in extracellular matrix integrity and it
has been proposed that this TG2 activity establishes a barrier to
tumor spread. Furthermore, TG2 controls sensitivity to the chemotherapeutic drug doxorubicin. Both doxorubicin sensitivity
and TG2 expression are highly variable in cultured human breast
cancer cell lines and inspection of the human gene (termed TGM2)
determined that a canonical CpG island exists within its 5# flank.
These features, when combined with its potential tumor suppressor activity, make TG2 an attractive candidate for epigenetic
silencing. Consistent with this, we observed that culturing
breast tumor cells with the DNA demethylating agent 5-aza-2#deoxycytidine (5-azadC) resulted in a robust increase in TG2
expression. Analysis of DNA harvested from cultured lines and
primary breast tumor samples indicated that TGM2 often displays aberrant hypermethylation and that there is a statistically
significant correlation between gene methylation and reduced expression. Finally, we observed that doxorubicin-resistant MCF-7/
ADR cells do not show TGM2 silencing but that doxorubicinsensitive MCF-7 cells do and that culturing MCF-7 cells on
5-azadC and subsequently restoring TG2 expression reduced sensitivity to doxorubicin. This work indicates that the TGM2 gene is
a target for epigenetic silencing in breast cancer and suggests that
this aberrant molecular event is a potential marker for chemotherapeutic drug sensitivity.
a lysine residue and the gamma-carboxyl group of a glutamine residue
(7). Functionally, transglutaminases cross-link proteins thus creating
proteinaceous structures resistant to proteolytic and mechanical degradation. Perhaps, the most enigmatic of the transglutaminases is type
2 transglutaminase, also referred to as TG2. TG2 is a ubiquitously
expressed protein comprised of a single 76 kD polypeptide and, like
other transglutaminases, exhibits Ca2þ-dependent protein crosslinking activity. However, evidence gathered indicates that TG2 is
a multifunctional protein. This feature seemingly stems, in part, from
the fact that this protein localizes to the cytosol and nucleus and can
be exported from the cell (8). In accordance with its transglutaminase
activity, extracellular pools of TG2 bind to and cross-link numerous
components of the extracellular matrix (ECM) such as fibronectin,
vitronectin, collagen, fibrin, laminin, osteonectin and osteopontin (9).
This has led to the view that TG2 cross-linking renders the ECM
resistant to mechanical and proteolytic degradation and stabilizes
cell–ECM interaction (10) and wound healing (11). TG2 has also been
shown to function as either a proapoptotic (12) or an antiapoptotic
(13–15) molecule. Another enigmatic feature of TG2 is that elevated
expression of this molecule renders cells resistant to the cytotoxic
effects of doxorubicin (16), although the mechanism for this drug
resistance remains controversial.
Based on published reports as well as our own observations that
doxorubicin sensitivity and TG2 expression are variable among cultured breast tumor cell lines and supported by an earlier study that
TG2 expression is influenced by cytosine methylation in a small panel
of leukemia lines (17), we hypothesized that the gene encoding TG2
(termed TGM2) could be a target for epigenetic silencing in breast
carcinoma. Consistent with this notion, we now report that the 5# flank
of the TGM2 gene is superimposed on a dense CpG island and the
culture of breast tumor lines that display low TG2 expression with the
global DNA demethylating agent 5-aza-2#-deoxycytidine (5-azadC)
dramatically increased expression of this gene. We further show that
analysis of genomic DNA (gDNA) from both cultured cells and primary breast tumors confirmed that this gene is silenced via aberrant
hypermethylation in this tumor type. Finally, we document that the
methylation status of TGM2 influences doxorubicin sensitivity in
doxorubicin-resistant and -sensitive MCF-7 cells establishing TGM2
methylation as a potential molecular marker for sensitivity to this
chemotherapeutic drug.
Introduction
Cancer-associated changes in cellular phenotype most often stem
from dysregulation of gene expression and/or protein function attributable to alterations within the genome (1). One prominent mechanism for these effects is de novo methylation and transcriptional
silencing of genes in cancer cells. Cytosine methylation leading to
gene silencing occurs at clustered 5#-CG-3# dinucleotides (termed
CpG islands) generally located within gene promoters and/or the 5#
flanking region (2). Aberrant CpG methylation events that arise during tumorigenesis are part of a more complex set of epigenetic events
that result in the transformation of chromatin from a transcriptionally
active to an inactive state (3). It is now widely accepted that epigenetic
gene silencing is either a causative or a correlative event in tumor
development (4–6).
Transglutaminases are a family of enzymes that catalyze the formation of a covalent bond between a free epsilon-amino group of
Abbreviations: 5-azadC, 5-aza-2#-deoxycytidine; BGS, bisulfite genomic sequencing; DCIS, ductal carcinoma in situ; ECM, extracellular matrix; gDNA,
genomic DNA; IBC, invasive breast carcinoma; M, methylated; MSP, methylation-specific polymerase chain reaction; PCR, polymerase chain reaction;
qPCR, quantitative polymerase chain reaction; RT–PCR, reverse transcription–
polymerase chain reaction; TG2, type 2 transglutaminase.
Materials and methods
Cell culture and drug treatment
The breast tumor and mammary epithelium-derived cell lines were maintained
as described previously (18). The doxorubicin-resistant MCF-7 derivative,
MCF-7/ADR cell line, was obtained from Dr A.R. Safa (Medical University
of South Carolina) and routinely cultured in complete medium containing 1 lg/ml
doxorubicin, however, was cultured without doxorubicin for 2 weeks prior to
experimentation.
5-azadC treatment was conducted as outlined previously (19). Briefly, cells
were seeded at low density (20% confluence) and cultured for 24 h. The next
day, 5-azadC was added to the medium to final concentration of 5 lM and cells
were cultured for an additional 24 h. The third day, medium was removed and
cells were washed with phosphate-buffered saline three times, then fresh
growth medium was added and cells were cultured for another 24 h. In total,
three of these 5-azadC treatment cycles were conducted prior to harvesting
cells. 5-azadC (Sigma, St Louis, MO) was freshly prepared as a 5 mM stock
solution in phosphate-buffered saline prior to use.
Sodium dodecyl sulfate–polyacrylamide gel electrophoresis and
immunoblotting
Sodium dodecyl sulfate–polyacrylamide gel electrophoresis and immunoblotting were performed using established protocols (20). TG2 mouse monoclonal
antibody (Ab-3) was purchased from Labvision Corp. (Fremount, CA), and
Ó The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
510
Epigenetic silencing of TGM2 in breast cancer
mouse monoclonal anti-tubulin (DM1A) was the generous gift of Dr D.W.
Cleveland (University of California, San Diego).
Reverse transcription–polymerase chain reaction
RNA extraction from cultured cells and tissues and first-strand cDNA synthesis
was conducted as outlined previously (18). The human TG2-specific primers
used are (F) 5#- TAAGAGATGCTGTGGAGGAG -3# and (R) 5#-CGAGCCCTGGTAGATAAA-3#. The thermocycling conditions used were 1 cycle
(95°C, 5 min); 35 cycles (94°C, 30 s; 58°C, 45 s; 72°C, 60 s) and final
extension at 72°C for 10 min. Human reduced form of guanosine adenine
dinucleotide phosphate (GAPDH) served as a control for RNA integrity and
was amplified using primers (F) 5#-ACCACAGTCCATGCCATCAC-3# and
(R) 5#-TCCACCACCCTGTTGCTGA-3#. Polymerase chain reaction (PCR)
products were analyzed by electrophoresis on 10% polyacrylamide gels followed by staining with ethidium bromide.
Quantitative polymerase chain reaction
SYBR green PCR master mix (Applied Biosystems, Foster City, CA) was used
for quantitative polymerase chain reaction (qPCR). In total, 2 ll of cDNA was
added to final 20 ll PCR reaction volume using above-outlined reverse transcription–polymerase chain reaction (RT–PCR) primer sets for both TG2 and
GAPDH. Thermocycling conditions used for qPCR were 1 cycle (95°C, 10
min) and 40 cycles (95°C, 15 s; 58°C, 30 s; 72°C, 30 s). Thermocycling was
conducted using a Bio-Rad MiniOpticon System (Bio-Rad Laboratories, Hercules, CA) and CT was calculated using supplied software. TG2 transcript
abundance relative to GAPDH transcript abundance (DCT) was calculated
and subsequently used to calculate changes (2DDCT ) in TG2 gene expression.
DNA extraction from human breast tumors and normal tissues
Fresh-frozen archival breast tumors were obtained from the University of
Florida Shands Cancer Center Molecular Tissue Bank. All specimens and
pertinent patient information were treated in accordance with policies of the
Institutional Review Board of the University of Florida Health Sciences Center.
Tumors analyzed in this study were examined by a surgical pathologist and
identified as invasive breast adenocarcinoma (stage II or III). Where indicated,
matched normal breast tissue was obtained from disease-free surgical margin.
gDNA was isolated from these samples, as well as cultured cell lines, using
a Qiagen blood and cell culture DNA kit (Qiagen Inc., Valencia, VA) and
stored at –20°C prior to use.
Archival blocks of formalin-fixed, paraffin-embedded breast tumors were
obtained from Department of Pathology, University of Florida Shands Hospital, in accordance with Institutional Review Board policy. All tumors were
confirmed by histological examination to contain both invasive breast carcinoma (IBC) and ductal carcinoma in situ (DCIS) lesions. For each tumor,
10–15 sections were cut and DNA extracted with EX-WAX DNA extraction
kit (Chemicon Inc., Temecula, CA) following the manufacturer’s protocol.
Sodium bisulfite DNA conversion and methylation-specific polymerase chain
reaction
DNA was bisulfite converted and methylation-specific polymerase chain reaction (MSP) reactions were conducted as outlined previously (18). MSP
primers specific for methylated (M) TGM2 gene are the forward primer
(MF) 5#-GGGTTATAAGTTAGCGTCGTTTTTC-3# and reverse primer
(MR) 5#-GAAACCC CTAAATAACGACTACGATA-3#. MSP primers specific
for unmethylated (U) TGM2 gene are the forward primer (UF) 5#-GGGTTATAAGTTAGTGTTGTTTTTTG-3# and the reverse primer (UR) 5#-AAAACCCCTAAATAACAACTACAATA-3#. PCR conditions used were 1 cycle
(95°C, 15 min); 38 cycles [94°C, 30 s; 60°C (for M primer set) or 56°C (for U
primer set), 30 s] and final extension at 72°C for 10 min. Human placental
gDNA (Biochain Institute, Hayward, CA) either unmodified or methylated
in vitro with SssI methylase (NEB, Ipswich, MA) was used as validation
controls.
Bisulfite genomic sequencing
Bisulfite genomic sequencing (BGS) was conducted as outlined previously
(21). A 362 bp segment (nucleotide # 294 to þ68, translational start site 5
þ1) was amplified from bisulfite-converted gDNA. The primer sequences used
for BGS are (UBF) 5#-TTGGGTTAGTTGTGTGTTTTTGTT-3# and (UBR)
5#-AATCCTAACCCCCAATACCCC-3#. Thermocycling conditions used
were: 1 cycle (95°C, 15 min); 40 cycles (94°C, 30 s; 55°C, 45 s; 72°C, 1 min)
and final extension at 72°C for 10 min. Amplicons were subcloned into the
plasmid pCR4-TOPO (Invitrogen, Carlsbad, CA) and recombinants identified
by EcoRI digestion and subsequently sequenced using vector-encoded T7 primer.
Immunohistochemistry
TG2 immunostaining was carried out using a previously established protocol
(18). Briefly, sections of formalin-fixed, paraffin-embedded human breast
tumors were sequentially incubated with mouse monoclonal TG2 antibody
(Ab-3), horseradish peroxidase-conjugated anti-mouse IgG, developed with
a buffered solution containing 3,3#-diamino-benzidine tetrahydrochloride
and counterstained with hematoxylin. As a non-specific antibody control, serial sections were incubated with purified mouse IgG (Sigma). Photomicrographs were taken on a Leica microscope equipped with a digital camera.
Cell transfection
A full-length human TG2 cDNA clone was subcloned into the expression
vector pcDNA3-Myc as described previously (15). Cells were seeded into
6-well tissue culture plates and then transfected with 5 lg of plasmid with
Lipofectamine 2000 (Invitrogen) following the manufacturer’s instructions.
Cells were subsequently selected with 500 lg/ml G418 (Sigma) for 2 weeks.
Subsequently, individual colonies were picked, expanded and screened by
immunoblotting.
Cell viability assays
Cell viability was determined using MTT assays as described previously (22).
Briefly, cells were seeded into 96-well tissue culture plates at a density of 5 103
cells/well. The next day, cells were washed two times with phosphate-buffered
saline, fresh medium was added to the wells and indicated concentration of
doxorubicin (Sigma) was added. Forty-eight hours after drug addition, 10 ll
of 5 mg/ml (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide
(Sigma) was added to each well and incubated for 1 h at 37°C. The solution
was subsequently removed, 100 ll of dimethyl sulfoxide was added to each
well and the plate was read using a Bio-Rad microplate reader.
Statistical analysis
The results of cell viability (determined by MTT) and TG2 transcript abundance (determined by qPCR) measurements were expressed as calculated
means with standard deviation (SD). A two-tailed Student’s t-test was used
to compare measurements taken from two samples for statistical significance.
When comparing quantitative gene methylation results (obtained by determining the percent methylation of all 35 CpGs present in the BGS amplicon in five
individual clones) and gene expression (determined by qPCR), the nonparametric Spearman correlation coefficient (rs) was calculated. All statistical
analysis was conducted using software (ver15) developed by SPSS (Chicago,
IL). In all cases, a P value of ,0.05 was considered significant.
Results
Decreased TG2 expression in cultured breast cancer cells correlates
with TGM2 gene methylation
Several published reports indicate that TG2 expression is variable
among cultured human breast tumor lines (16,23,24). To examine this
in a large panel of breast tumor cell lines, we subjected extracts from
MDA-MB-231, MDA-MB-468, MCF-7, SK-BR-3, MDA-MB-435S,
BT-549 and T-47D cells to immunoblot analysis with anti-TG2
(Figure 1A). As controls, two human non-tumor mammary epithelial
cell lines, MCF-10A and 184B5, were included. This side-by-side
analysis readily detected expression of the 75 kD TG2 protein in
both control lines as well as extracts of MDA-MB-231 cells. In the
other breast tumor lines assayed (MDA-MB-468, MCF-7, SK-BR-3,
MDA-MB-435S, BT-549 and T-47D), we observed notably reduced
or undetectable expression of TG2.
Since TG2 expression has been linked to doxorubicin sensitivity,
we also conducted viability assays on this panel of cultured human
breast tumor lines treated with this drug. These results indicate that
most lines show similar levels of sensitivity to doxorubicin (Figure
1B). However, MDA-MB-231 cells, which show higher levels of TG2
expression, show significantly less sensitivity to doxorubicin than the
other cell lines tested confirming that sensitivity to this drug corresponds to TG2 expression.
Lu and Davies (17) observed that TG2 expression in human leukemia lines was controlled, in part, by DNA methylation. Therefore, we
hypothesized that the low level of TG2 expression we observed in
most cell lines stems from aberrant methylation of CpG dinucleotides
within the TG2 gene (termed TGM2). As an initial test of this hypothesis, these breast tumor cells were cultured in the presence or absence
of the DNA demethylating agent 5-azadC and RT–PCR assays to
examine TG2 expression conducted. Paralleling immunoblot results,
511
L.Ai et al.
Fig. 1. Culturing human breast cancer cells with the global DNA demethylating agent 5-azadC results in heightened TG2 expression. (A) Whole cell lysates were
formed from cultures of indicated human breast tumor (MDA-MB-231, MDA-MB-435S, MDA-MB-468, T-47D, BT-549, MCF-7 and SK-BR-3) and non-tumor
(MCF-10A and 184B5) cell lines. These lysates were subjected to immunoblot analysis with anti-TG2 (top) and tubulin (bottom) to confirm equivalent loading.
(B) Indicated cultured human breast tumor lines were treated with 0.25 lg/ml (shaded bars), 0.50 lg/ml (open bars) doxorubicin or mock-treated (black bars) for 48 h
and cell viability assessed by MTT assay (error bars 5 ±SD, n 5 8 for each measurement). Asterisks ( ) denote statistically significant (P , 0.01, Student’s t-test)
differences in viability compared with MDA-MB-231 cells at indicated drug dose. (C) Breast tumor lines were cultured in the presence or absence of 5-azadC.
Following this, cells were harvested, total RNA was isolated and cDNA was synthesized. PCR reactions were then conducted using primers designed to amplify the
TG2 (top) and GAPDH (bottom) transcripts. (D) qPCR was performed and relative TG2 expression in cultures of cells either treated or untreated with 5-azadC.
Graphed is the fold increase in relative TG2 transcript abundance in various breast tumor lines in response to 5-azadC treatment (error bars 5 ±SD, n 5 4).
we observed that MDA-MB-231 cells displayed expression of the
TG2 transcript and that the relative abundance of this transcript was
unaltered by culture on 5-azadC (Figure 1C). In contrast, TG2 transcript abundance in the other six breast tumor lines was either very
low or subdetectable in untreated cells; however, gene expression was
notably increased following culture on 5-azadC.
To quantify changes in gene expression, qPCR was performed.
Consistent with our RT–PCR results, we measured an insignificant
(1.1-fold, ±0.2 SD, n 5 4) increase in TG2 mRNA in 5-azadC-treated
MDA-MB-231 cells (Figure 1D). Conversely, culturing MDA-MB-468
512
cells with 5-azadC resulted in an 18.9-fold (±3.4 SD, n 5 4) increase
in TG2 transcript abundance compared with untreated cultures of this
cell line (Figure 1D). Similarly, statistically significant (P , 0.01,
Student’s t-test) increases in TG2 transcripts were measured in T-47D
(10.6-fold, ±2.1 SD, n 5 4), MCF-7 (22.9-fold, ±4.3 SD, n 5 4),
BT-549 (142.0-fold, ±8.9 SD, n 5 4), MDA-MB-435S (90.5-fold,
±7.8 SD, n 5 4) and SK-BR-3 (37.0-fold, ±4.5 SD, n 5 4) cells.
To directly examine TGM2 gene hypermethylation, we first inspected the architecture of the human TGM2 gene using public
genome sequence databases. We observed that the TGM2 gene maps
Epigenetic silencing of TGM2 in breast cancer
to chromosome 20.q11.23, consistent with the location of this gene
as previously determined by fluorescence in situ hybridization (25).
Inspection of the 5# flanking region of the TGM2 gene showed this
region to be unusually CG rich (Figure 2A). As originally described
by Gardinier-Garden and Frommer (26), a 316 bp segment of this
region of the genome (bp –284 to þ32, translational start site 5 þ1)
that contains 35 CpG dinucleotides meets the criteria of a canonical
CpG island. This CpG island overlaps both the transcriptional and
the translational start sites of the TGM2 gene. Aberrant cytosine
methylation within CpG islands that encompass both the 5# flanking
region and transcriptional start site of a silenced gene is commonly
observed in cancer cell genomes (27,28); thus, the architecture
of TGM2 marks this gene as a potential candidate for epigenetic
silencing.
To directly assess CpG methylation in human breast cancer cell
lines, we developed a set of oligonucleotide primers useful in MSP
(29) and BGS (30) analyses (Figure 2A). For MSP, oligonucleotide
primers were designed to amplify a segment of the TGM2 CpG island
existing in either an M or a U state based on the modification status of
CpG dinucleotides following sodium bisulfite modification. The primers designed to amplify U TGM2 gene sequence yield a 183 bp
amplicon spanning from nucleotide # 133 to þ50, and the M primer
set yield a 182 bp amplicon spanning from nucleotide # 132 to þ50.
Amplification of the TGM2 gene from bisulfite-treated human
Fig. 2. The TGM2 gene is hypermethylated in cultured human breast tumor cells. (A) The 5# flank of the human TGM2 gene is shown. Illustrated are the GþC
percentage, the boundaries of the CpG island, the transcriptional start (TSS) and translational start (ATG) sites of the TGM2 gene. Additionally, the location of
primers used for MSP and BGS are illustrated. (B) MSP primers were designed to amplify a segment of the TGM2 gene in an M-specific or U-specific manner.
M-specific (top) and U-specific (bottom) primers were used in PCR reactions containing sodium bisulfite-converted human placental DNA (lane 1), bisulfiteconverted placental DNA methylated in vitro using SssI methylase (lane 2) and placental DNA that was not bisulfite treated prior to use in MSP reactions (lane 3).
(C) TG2 and GAPDH transcripts were amplified by RT–PCR from total RNA isolated from the non-tumor mammary epithelial cell lines MCF-10A and 184B5
(upper panels). Alternatively, gDNA from these cells was bisulfite treated and subjected to MSP analysis to assess the methylation status of TGM2 (lower panels).
(D) gDNA harvested from indicated breast tumor cell lines was bisulfite modified and subjected to MSP analysis using M-specific (top) and U-specific (bottom)
primers. (E) BGS analysis was performed on the indicated cell lines. Indicated are the position of methylated CpG dinucleotides (filled circles), unmethylated
CpGs (open circles), the overall percentage of methylated CpGs in the sequenced clones derived from each indicated cell line and the TG2 expression relative to
MCF10A cells.
513
L.Ai et al.
placental gDNA was observed using the U-specific primer set but not
with the M-specific primer set (Figure 2B). Conversely, when this
gDNA was methylated in vitro using SssI methylase and subsequently
subjected to bisulfite modification prior to PCR, we observed amplification of the M-specific amplicon but not the U-specific product, and
amplification with either primer set was strictly dependent on bisulfite
modification of the target DNA. Sequence analysis of both the U and
the M amplicons indicated that these primer sets specifically amplify
the targeted portion of the TGM2 gene (data not shown). These validation experiments indicate that our MSP primers specifically amplify bisulfite-treated human gDNA in a manner that clearly ascertains
the methylation status of the TGM2 gene.
RT–PCR and MSP analysis were subsequently conducted on
184B5 and MCF-10A cells (Figure 2C). RT–PCR analysis of these
two normal (non-tumor) control lines indicated that each expresses
detectable levels of TGM2 transcript. Additionally, MSP analysis determined that the TGM2 gene is unmethylated in both these cell lines.
gDNA was harvested from our panel of breast cancer cell lines
cultured either in the absence or in the presence of 5-azadC and used
in MSP assays. We observed amplification of the U-specific, but not
the M-specific, amplicon in MDA-MB-231 DNA, indicating that the
TGM2 gene is unmethylated in this cell line (Figure 2D). Analysis of
T-47D and MBA-MD-468 cells not cultured on 5-azadC showed amplification with both U- and M-specific amplicons indicating the presence of both M and U DNA in these samples. In non-treated cultures
of MCF-7, BT-549, MDA-MB-435S and SK-BR-3, we observed exclusive amplification with the M-specific primer set. MSP analysis
conducted on these lines after treatment with 5-azadC indicated a clear
increase in the amplification of the U-specific product, consistent with
demethylation of the TGM2 gene.
To quantitatively assess TGM2 methylation, bisulfite-modified gDNA
from these breast cancer lines was analyzed by BGS. This entailed
sequencing amplicons obtained using primers designed to regions of
the TGM2 gene within the CpG island (see Figure 2A) and devoid of
CpG dinucleotides; thus, amplification of bisulfite-modified gDNA proceeds in a manner unbiased by gene methylation status. Following PCR,
the 362 bp BGS amplicon was subsequently subcloned into a vector,
recombinants were identified, multiple clones were sequenced and the
methylation status of the 35 CpG dinucleotides within this region of the
genome was determined (30). Results from BGS analysis of MDAMB-231, MCF-7, MDA-MB-468, BT-549, MDA-MB-435S and SKBR-3 (with and without culture on 5-azadC) are shown (Figure 2E).
MDA-MB-231 cells, which contain a U TGM2 gene as judged by MSP,
showed very limited CpG methylation (0.6% of the CpG analyzed) by
BGS. Alternatively, BT-549 (70.8%), MCF-7 (33.1%), MDA-MB-435S
(73.1%), MDA-MB-468 (25.1%) and untreated SK-BR-3 (37.1%) displayed dense methylation of the CpG island of the TGM2 gene. Further,
sparse methylation of the TGM2 gene was observed when gDNA was
harvested and analyzed from SK-BR-3 cells cultured in the presence of
5-azadC (5.7%). When gDNA harvested from the control MCF-10A and
184B5 lines was analyzed by BGS, we did not observe any methylation
within the CpG island of the TGM2 gene in these cells (see supplementary Figure S1 available at Carcinogenesis online).
To compare TG2 expression with TGM2 gene methylation, we
calculated the relative TG2 transcript abundance (as measured by
qPCR) using TG2 expression in the non-tumor cell line MCF-10A as
the reference. This revealed that the breast tumor line MDA-MB-231,
which does not show aberrant CpG methylation within the TGM2 gene,
expresses 2-fold higher levels of TG2 than MCF-10A (Figure 3D).
Conversely, other breast cancer lines displaying aberrant methylation
of the TGM2 gene showed markedly reduced TG2 expression compared with MCF-10A. Analysis of this set of breast tumor lines indicated that the correlation between reduced expression and increased
TGM2 methylation (as scored by BGS) is statistically significant
[Spearman correlation coefficient (rs) 5 0.91, n 5 7, P , 0.01].
Thus, based on several molecular analyses of cultured breast cancer
cell lines, we conclude that the TGM2 gene is a target for CpG hypermethylation and that reduced expression of TG2 correlates with this
aberrant epigenetic event.
514
Aberrant methylation of the TGM2 gene is common in primary human
breast tumors
Nine fresh-frozen, locally IBC tumors (stage II or III) and normal
breast tissue from the same patient were obtained from the University
of Florida Shands Cancer Center Molecular Tissue Bank. gDNA was
extracted from these samples, bisulfite modified and used in MSP
reactions. Representative data from four of these samples, designated
BrCa-11, 12, 13 and 14, are shown (Figure 3A). In all normal breast
tissue samples, we observed exclusive amplification of the U-specific
amplicon indicating that the TGM2 gene is not methylated in normal
mammary tissue. In tumor BrCa-13, we only observed exclusive amplification of the U-specific amplicon, indicating that the TGM2 gene
is not methylated in this tumor sample. Conversely, in tumors BrCa11, 12 and 14, amplification of the M-specific amplicon was detected
indicating that CpG methylation of the TGM2 gene also occurs in
primary breast tumors. In total, MSP analysis indicated aberrant
TGM2 CpG methylation in four of these frozen tumor samples,
whereas the remaining five showed no evidence of CpG methylation
(see supplementary Table S2 available at Carcinogenesis online).
Total RNA was also harvested from these frozen tumor/normal
tissue samples, cDNA was synthesized and qPCR reactions were performed to measure relative TG2 and GAPDH transcript abundance.
Again, representative data shown for tumors BrCa-11 through 14 indicate that in tumors BrCa-11, 12 and 14, relative TG2 transcript
abundance was significantly reduced (P , 0.01) compared with
patient-matched normal tissue samples (Figure 3B). In contrast, TG2
expression was roughly equivalent in tumor and matched normal
tissues in sample BrCa-13.
Again, a statistical approach was used to compare TG2 expression
with TGM2 gene methylation in this panel of nine frozen matched
tumor/normal breast samples. Analysis of this set of breast tumors indicated a statistically significant correlation between reduced expression and increased CpG methylation (rs 5 0.85, n 5 9, P , 0.01).
The results obtained from this qPCR and TGM2 methylation analysis
of matched frozen and normal breast samples indicate that, similar to
what we have documented in cultured cell lines, reduced TG2 expression correlates with aberrant methylation of the TGM2 gene in breast
tumors.
Figure 3C shows representative MSP data obtained from 11 additional IBC tumor samples for which no matched normal samples were
available. In tumor samples BrCa-02, 04, 16, 17 and 21, we observed
amplification with only the U-specific primer set indicating a lack of
TGM2 gene hypermethylation in these tumor samples. Tumor samples
BrCa-06, 09 and 22 only showed amplification with the M-specific
primer set, whereas BrCa-18 and 20 show amplification with both Uand M-specific primer sets. Of the 45 IBC tumors analyzed during this
study, we observed that 20 (44.4%) displayed aberrant methylation of
the TGM2 gene as judged by MSP (supplementary Table S2 available
at Carcinogenesis online), leading us to conclude that hypermethylation of the TGM2 gene is a frequent event in invasive breast tumors.
BGS was again conducted to examine TGM2 gene methylation in
a detailed manner. In the normal breast tissue from patient BrCa-14,
we observed limited methylation (2.3%) of the CpG island associated
with the TGM2 gene (Figure 3D). Similarly, we observed no methylation of the TGM2 gene in tumor samples BrCa-04 and 13. These
results are in accordance with MSP results observed on these tumors
and this normal breast tissue sample. In tumor BrCa-06, 09 and 14, we
detected extensive CpG methylation clustered in close proximity to the
translational start site of the TGM2 gene (70.8%, 38.3% and 37.7%,
respectively). Of note, diminished CpG methylation was detected in
the region of the CpG island investigated that lies farthest upstream
from the transcriptional start site of TGM2. Nevertheless, the results of
this experiment clearly indicate that the CpG island present at the 5#
flank of the human TGM2 gene is subject to dense methylation in
breast tumors. Moreover, since dense TGM2 methylation is not observed in normal breast tissues such as BrCa-14N (Figure 3D) and
other normal tissues examined (data not shown), TGM2 hypermethylation in tumor tissues is an aberrant, disease-associated event.
Epigenetic silencing of TGM2 in breast cancer
Fig. 3. Hypermethylation of the TGM2 gene occurs in primary breast tumors and correlates with reduced TG2 expression. (A) gDNA isolated from the indicated
patient-matched normal breast tissue (N) and tumor tissue (T) was bisulfite converted and subjected to MSP analysis to assess the methylation status of TGM2
using M-specific and U-specific primer sets (top). Alternatively, TG2 and GAPDH transcripts were amplified by RT–PCR from total RNA isolated from these
matched normal and breast tumor tissue samples (bottom). (B) TG2 and GAPDH transcript abundance was measured by qPCR. Graphed is the relative TG2
expression (as a function of GAPDH transcript abundance) in patient-matched normal and tumor tissues (error bars 5 ±SD, n 5 4). Asterisks ( ) denote
statistically significant (P , 0.01, Student’s t-test) differences in TG2 transcript abundance from matched normal tissue sample. (C) gDNA isolated from various
breast tumor samples was subjected to MSP analysis using M-specific (top) and U-specific (bottom) primer sets. (D) Bisulfite-converted gDNA from indicated
breast tumors or normal tissue was subjected to BGS analysis. Methylated CpG dinucleotides are indicated by the filled circles, unmethylated CpGs indicated by
open circles and the percentage of methylated CpGs in each sample is indicated.
We next examined TG2 expression in 10 different archival breast
tumor samples by immunohistochemistry and, in general, found tumors to have three types of staining patterns. In some patient samples,
such as BrCa-101, we observed prominent staining of normal ductal
epithelium, DCIS and regions of IBC (Figure 4A). When this tumor
was incubated with non-specific purified mouse IgG, we observed no
staining (Figure 4A), confirming the specificity of the staining using
anti-TG2. In other tumors, such as BrCa-104, strong staining was
observed in normal mammary ductal epithelium; however, moderate
staining was evident in both DCIS and IBC lesions (Figure 4B). Yet
other tumors, such as BrCa-109, displayed sharply reduced staining in
both DCIS and IBC lesions (Figure 4C).
Following immunohistochemistry analysis, gDNA was extracted
from each of these tumor specimens, bisulfite modified and used in
MSP assays. MSP analysis of DNA extracted from tumor BrCa-101
indicated that this tumor contains U TGM2 and no M-specific PCR
amplicon was detected (Figure 4A). MSP analysis of tumor BrCa-104
showed amplification of both U- and M-specific TGM2 amplicons
indicating some level of TG2 silencing in this tumor (Figure 4B).
MSP analysis of DNA extracted from tumor BrCa-109 displayed
strong amplification of the TGM2 M-specific amplicon and relatively
weak amplification of the U-specific amplicon (Figure 4C). Of note,
in tumors such as BrCa-109 that display hallmarks consistent with
epigenetic silencing of TG2 expression, we have observed regions of
the tumor that express abundant TG2 as judged by immunohistochemistry (data not shown). This suggests heterogeneous TG2 silencing
within individual breast tumors. Nevertheless, our findings indicate
that methylation of the TGM2 gene correlates with reduced expression
of TG2 in situ and that this aberrant molecular event can occur at
a relatively early stage in the process of breast tumor progression.
Epigenetic silencing of the TGM2 gene results in heightened
doxorubicin sensitivity in cultured breast tumor lines
Others have documented that derivatives of MCF-7 cells that demonstrate doxorubicin resistance (designated MCF-7/ADR) express higher levels of TG2 than parental MCF-7 cells that retain sensitivity to
this drug (16). Based on our studies, we hypothesized that this difference in TG2 expression was attributable to differences in the methylation status of the TGM2 gene in these cells. To test this, MSP was
conducted on MCF-7, MCF-7/ADR and MCF-7 cells treated with
5-azadC. This analysis indicated, although MCF-7 shows TGM2
hypermethylation that can be partially reversed by 5-azadC, that
MCF-7/ADR cells do not display methylation of TGM2 (Figure 5A).
Furthermore, whereas BGS analysis confirmed that the TGM2 gene is
515
L.Ai et al.
Fig. 4. Silencing of TGM2 occurs in both invasive and non-invasive breast tumor lesions. (A) Tumor BrCa-101 was sectioned and subjected to
immunohistochemistry with anti-TG2 (left). Sections were either incubated with anti-TG2 (top row) or with non-specific mouse IgG (bottom row). Normal
mammary ducts, DCIS and IBC lesions present in this tissue sample are shown. Alternatively, gDNA was isolated from this tumor and subjected to MSP analysis
(right). Tumor BrCa-104 (B) and BrCa-109 (C) were analyzed as outlined in A except that non-specific IgG control is omitted.
densely methylated in MCF-7 (see Figure 2E), this gene was significantly demethylated in MCF-7 treated with 5-azadC and very low
levels of methylation were scored in MCF-7/ADR cells (see supplementary Figure S1 available at Carcinogenesis online). Immunoblotting showed that TG2 expression is markedly lower in MCF-7 cells
compared with MCF-7/ADR and that 5-azadC treatment clearly raises
the TG2 expression level in MCF-7 cells but not to the level observed
in MCF-7/ADR cells or MCF-7 cells stably transfected with a recombinant human TG2 expression construct (Figure 5B).
Next, the effect of 5-azadC treatment on the sensitivity of MCF-7 to
doxorubicin was examined. As expected, we observed that MCF-7/
ADR cells show significantly less sensitivity to this drug than MCF-7
(Figure 5C) as judged by MTT viability assays 48 h after drug addition. We also observed that 5-azadC treatment of MCF-7 cells resulted
in a statistically significant (P , 0.01, Student’s t-test) increase in
doxorubicin resistance; however, 5-azadC-induced resistance was
lower than that exhibited by MCF-7/ADR cells. This partial complementation of the doxorubicin-sensitive phenotype of MCF-7 cells
may perhaps be due to the lower level of TG2 expression exhibited
by the 5-azadC-treated MCF-7 compared with MCF-7/ADR cells,
suggesting that the level of TG2 expression quantitatively influences
doxorubicin sensitivity. In addition to these experiments, and consistent with the findings of others (23), we observed that expression of
recombinant human TG2 in MCF-7 cells also reduces sensitivity to
doxorubicin.
Discussion
We have documented that cultured human breast tumor cell lines
often show reduced expression of the TG2 protein. TG2 expression
is restored by culturing these cells in the presence of the global de-
516
methylating agent 5-azadC, consistent with DNA methylation controlling expression of the TGM2 gene. Using a variety of approaches,
we confirmed that the CpG island present at the 5# flank of the human
TGM2 gene is subject to dense methylation in both cultured cells and
primary breast tumors. Moreover, this aberrant CpG methylation is
associated with reduced TG2 expression. The frequency at which this
aberrant event was encountered in primary breast tumors (44%)
indicates that epigenetic silencing of the TGM2 gene is a common
event in this tumor type.
Owing to the multifunctional nature of the TG2 protein, there are
several ways in which loss of TG2 expression through epigenetic
silencing could impact the process of tumorigenesis and, indeed, several lines of evidence implicate TG2 as a tumor suppressor (31–34).
One such way is through suppression of its proposed proapoptotic
function given the prominent role that dysregulation of apoptotic
mechanisms play in tumorigenesis by allowing cells to extend life
span (35). Rodolfo et al. (12) observed that TG2 possesses an eight
amino acid region sharing .70% of identity with the BH3 domain of
the Bcl-2 family, thus placing TG2 in a growing family of proapoptotic BH3-only proteins (36). In support of this assertion, cell-permeable
TG2-BH3 peptide was observed to induce a conformational change
and translocation of Bax to the mitochondria and promote release of
cytochrome c. In addition, TG2-induced Bax polymerization may
occur when the cytosolic level of Ca2þ is increased during apoptosis
resulting in stabilization of the apoptosis-promoting conformation of
Bax and prolonging cytochrome c release. In this view, TG2 would
apparently act both as an upstream regulator of the apoptotic response
and as an intensifier of the process by prolonging Bax activity (37).
Another mechanism by which TG2 silencing may contribute to
tumor development is through its proposed function as a barrier to
tumor spread by promoting ECM cross-linking, rendering the ECM
Epigenetic silencing of TGM2 in breast cancer
doxorubicin-sensitive MCF-7 line. We extended this study by documenting that MCF-7 cells have hypermethylated the TGM2 gene and
hence silenced TG2 expression, whereas MCF-7/ADR cells have not
silenced this gene. This finding argues that the difference in doxorubicin sensitivity between MCF-7 and MCF-7/ADR cells does not stem
from overexpression of TG2 in MCF-7/ADR per se, but rather an
aberrant molecular event within the MCF-7 line resulting in abnormally low TG2 expression. Taken together, the conclusion drawn is
that epigenetic silencing of TG2 results in a doxorubicin-sensitive
phenotype in MCF-7 cells.
Various mechanistic models have been proposed to explain the
molecular mechanism by which TG2 influences sensitivity to doxorubicin (23,24,42,43). Although the exact mechanism(s) by which
TG2 controls sensitivity to this drug remains an active topic for investigation, it is clear that TG2 expression status exerts control over
the cytotoxic response to doxorubicin. Since doxorubicin (adriamycin) is a common, and effective, drug for treating breast cancer
as either a neoadjuvant and/or an adjuvant therapeutic, our findings
have potential clinical relevance given that doxorubicin has welldocumented adverse cardiac side-effects (44) and that anthracycline
resistance is commonly encountered in breast tumors (45).
Fig. 5. Epigenetic silencing of TGM2 corresponds to doxorubicin sensitivity.
(A) gDNA was isolated from MCF-7 cells (lane 1), MCF-7 treated with
5-azadC (lane 2) and doxorubicin-resistant MCF-7/ADR (lane 3) cells, and
subjected to MSP analysis using M-specific (top) and U-specific (bottom)
primers to assess the methylation status of TGM2. (B) TG2 expression in
MCF-7 (lane 1), MCF-7 treated with 5-azadC (lane 2), MCF-7 stably
expressing recombinant human TG2 (lane 3) and MCF-7/ADR cells (lane 4)
was assessed by immunoblotting with anti-TG2 (top) or anti-tubulin
(bottom). (C) MCF-7, MCF-7 treated with 5-azadC, MCF-7 stably
expressing recombinant human TG2 and MCF-7/ADR cells were either
mock-treated (filled bars), treated with 0.25 lg/ml (shaded bars) or 0.50
lg/ml (open bars) doxorubicin. Forty-eight hours after drug treatment, cell
viability was measured using MTT (error bars 5 ±SD, n 5 8 for each
measurement). Asterisks ( ) denote statistically significant (P , 0.01,
Student’s t-test) differences in viability compared with MCF-7 cells at
indicated drug dose.
Supplementary material
Supplementary material can be found at http://carcin.oxfordjournals.
org/
Funding
National Institutes of Health (R01 CA102289 and R21 CA102220) to
K.D.B.
Acknowledgements
The authors thank Drs Michael Kladde and Mandy Robertson for helpful
comments on this manuscript.
Conflict of Interest Statement: None declared.
resistant to mechanical and proteolytic degradation and stabilizing
cell–ECM interaction (10). Maintenance of ECM integrity has a distinctly important role in breast cancer since the majority of breast
cancers are adenocarcinomas that develop from the ductal epithelium
of the mammary gland. Although cells that comprise a DCIS lesion
are tumor cells and free to infiltrate ductal structure, this type of lesion
is physically constrained, in part, by the basement membrane surrounding the duct. DCIS lesions that acquire an invasive capability
can break through this barrier and invade surrounding stroma resulting
in a more serious invasive tumor. Our observation that TG2 expression
can be diminished in DCIS lesions suggests that reduced expression of
this gene may help facilitate breast tumor invasiveness.
TG2 also facilitates the incorporation of transforming growth factor-b
into the ECM (38) and loss of extracellular TG2 activity could potentially negatively impact transforming growth factor-b function as
a tumor suppressor (39). More recently, the atypical G protein-coupled
receptor and metastasis suppressor, GPR56, was shown to bind TG2
suggesting yet another mechanism by which loss of TG2 would favor
tumor progression (33). Furthermore, ECM stiffness, which would potentially be influenced by TG2-induced cross-linking, is gaining acceptance as a key regulator of cell behavior and may play an integral role in
the process of tumorigenesis and invasion (40,41). Thus, although the
exact mechanism(s) by which TG2 functions to limit tumorigenesis
remains unclear, it is reasonable to conclude that epigenetic silencing
of TGM2 would contribute to the carcinogenic process by severely
reducing its ability to suppress tumor growth and/or metastasis.
Consistent with the observations originally made by Mehta (16),
we have documented that doxorubicin-resistant MCF-7/ADR cells
display heightened TG2 expression when compared with the parental,
References
1. Weinberg,R.A. (1996) How cancer arises. Sci. Am., 275, 62–70.
2. Bird,A.P. et al. (1999) Methylation-induced repression—belts, braces, and
chromatin. Cell, 99, 451–454.
3. Wolffe,A.P. (2001) Chromatin remodeling: why it is important in cancer.
Oncogene, 20, 2988–2990.
4. Esteller,M. (2003) Cancer epigenetics: DNA methylation and chromatin
alterations in human cancer. Adv. Exp. Med. Biol., 532, 39–49.
5. Jones,P.A. et al. (2002) The fundamental role of epigenetic events in cancer. Nat. Rev. Genet., 3, 415–428.
6. Robertson,K.D. (2005) DNA methylation and human disease. Nat. Rev.
Genet., 6, 597–610.
7. Lorand,L. et al. (1984) Transglutaminases. Mol. Cell. Biochem., 58, 9–35.
8. Chen,J.S. et al. (1999) Tissue transglutaminase: an enzyme with a split
personality. Int. J. Biochem. Cell. Biol., 31, 817–836.
9. Aeschlimann,D. et al. (2000) Protein crosslinking in assembly and remodelling of extracellular matrices: the role of transglutaminases. Connect.
Tissue Res., 41, 1–27.
10. Zemskov,E.A. et al. (2006) The role of tissue transglutaminase in cellmatrix interactions. Front. Biosci., 11, 1057–1076.
11. Telci,D. et al. (2006) Tissue transglutaminase (TG2)—a wound response
enzyme. Front. Biosci., 11, 867–882.
12. Rodolfo,C. et al. (2004) Tissue transglutaminase is a multifunctional BH3only protein. J. Biol. Chem., 279, 54783–54792.
13. Boehm,J.E. et al. (2002) Tissue transglutaminase protects against apoptosis
by modifying the tumor suppressor protein p110 Rb. J. Biol. Chem., 277,
20127–20130.
14. Milakovic,T. et al. (2004) Intracellular localization and activity state of
tissue transglutaminase differentially impacts cell death. J. Biol. Chem.,
279, 8715–8722.
517
L.Ai et al.
15. Yamaguchi,H. et al. (2006) Tissue transglutaminase serves as an inhibitor
of apoptosis by cross-linking caspase 3 in thapsigargin-treated cells. Mol.
Cell Biol., 26, 569–579.
16. Mehta,K. (1994) High levels of transglutaminase expression in doxorubicin-resistant human breast carcinoma cells. Int. J. Cancer, 58, 400–406.
17. Lu,S. et al. (1997) Regulation of the expression of the tissue transglutaminase gene by DNA methylation. Proc. Natl Acad. Sci. USA, 94, 4692–4697.
18. Ai,L. et al. (2006) Epigenetic silencing of the tumor suppressor cystatin M
occurs during breast cancer progression. Cancer Res., 66, 7899–7909.
19. Kim,W.J. et al. (2002) Aberrant methylation of the ATM promoter correlates with increased radiosensitivity in a human colorectal tumor cell line.
Oncogene, 21, 3864–3871.
20. Adamson,A.W. et al. (2005) Methylator-induced, mismatch repair-dependent G2 arrest is activated through Chk1 and Chk2. Mol. Biol. Cell, 16,
1513–1526.
21. Vo,Q.N. et al. (2004) The ATM gene is a target for epigenetic silencing in
locally advanced breast cancer. Oncogene, 23, 9432–9437.
22. Kim,W.J. et al. (2005) The monofunctional alkylating agent N-methyl-N#nitro-N-nitrosoguanidine triggers apoptosis through p53-dependent and
-independent pathways. Toxicol. Appl. Pharmacol., 202, 84–98.
23. Herman,J.F. et al. (2006) Implications of increased tissue transglutaminase
(TG2) expression in drug-resistant breast cancer (MCF-7) cells. Oncogene,
25, 3049–3058.
24. Kim,D.S. et al. (2006) Reversal of drug resistance in breast cancer cells by
transglutaminase 2 inhibition and nuclear factor-kappaB inactivation. Cancer Res., 66, 10936–10943.
25. Gentile,V. et al. (1994) The human tissue transglutaminase gene maps on
chromosome 20q12 by in situ fluorescence hybridization. Genomics, 20,
295–297.
26. Gardiner-Garden,M. et al. (1987) CpG islands in vertebrate genomes.
J. Mol. Biol., 196, 261–282.
27. Feltus,F.A. et al. (2003) Predicting aberrant CpG island methylation. Proc.
Natl Acad. Sci. USA, 100, 12253–12258.
28. Ushijima,T. (2005) Detection and interpretation of altered methylation
patterns in cancer cells. Nat. Rev. Cancer, 5, 223–231.
29. Herman,J.G. et al. (1996) Methylation-specific PCR: a novel PCR assay for
methylation status of CpG islands. Proc. Natl Acad. Sci. USA, 93, 9821–9826.
30. Frommer,M. et al. (1992) A genomic sequencing protocol that yields a positive display of 5-methylcytosine residues in individual DNA strands. Proc.
Natl Acad. Sci. USA, 89, 1827–1831.
518
31. Haroon,Z.A. et al. (1999) Tissue transglutaminase is expressed as a host
response to tumor invasion and inhibits tumor growth. Lab. Invest., 79,
1679–1686.
32. Johnson,T.S. et al. (1994) Transfection of tissue transglutaminase into
a highly malignant hamster fibrosarcoma leads to a reduced incidence of
primary tumour growth. Oncogene, 9, 2935–2942.
33. Xu,L. et al. (2006) GPR56, an atypical G protein-coupled receptor, binds
tissue transglutaminase, TG2, and inhibits melanoma tumor growth and
metastasis. Proc. Natl Acad. Sci. USA, 103, 9023–9028.
34. Jones,R.A. et al. (2006) Matrix changes induced by transglutaminase 2 lead
to inhibition of angiogenesis and tumor growth. Cell Death Differ., 13,
1442–1453.
35. Reed,J.C. (1999) Dysregulation of apoptosis in cancer. J. Clin. Oncol., 17,
2941–2953.
36. Willis,S.N. et al. (2005) Life in the balance: how BH3-only proteins induce
apoptosis. Curr. Opin. Cell Biol., 17, 617–625.
37. Fesus,L. et al. (2005) Transglutaminase 2 in the balance of cell death and
survival. FEBS Lett., 579, 3297–3302.
38. Verderio,E. et al. (1999) Regulation of cell surface tissue transglutaminase:
effects on matrix storage of latent transforming growth factor-beta binding
protein-1. J. Histochem. Cytochem., 47, 1417–1432.
39. Jakowlew,S.B. (2006) Transforming growth factor-beta in cancer and metastasis. Cancer Metastasis Rev., 25, 435–457.
40. Discher,D.E. et al. (2005) Tissue cells feel and respond to the stiffness of
their substrate. Science, 310, 1139–1143.
41. Huang,S. et al. (2005) Cell tension, matrix mechanics, and cancer development. Cancer Cell, 8, 175–176.
42. Fok,J.Y. et al. (2006) Implications of tissue transglutaminase expression in
malignant melanoma. Mol. Cancer Ther., 5, 1493–1503.
43. Mann,A.P. et al. (2006) Overexpression of tissue transglutaminase leads to
constitutive activation of nuclear factor-{kappa}B in cancer cells: delineation of a novel pathway. Cancer Res., 66, 8788–8795.
44. Perez,E.A. (2001) Doxorubicin and paclitaxel in the treatment of advanced
breast cancer: efficacy and cardiac considerations. Cancer Invest., 19,
155–164.
45. Hortobagyi,G.N. (1995) Management of breast cancer: status and future
trends. Semin. Oncol., 22, 101–107.
Received July 21, 2007; revised November 14, 2007;
accepted December 1, 2007