NON SOLLE
ELSEVIER
Cancer Letters
journal homepage: www.elsevier.com/locate/canlet
CancerLetters
Expression of TBX2 promotes anchorage-independent growth and survival in the p53-negative SW13 adrenocortical carcinoma
Amin Ismail *, Andrew Bateman
Endocrine Research Laboratory, Division of Experimental Medicine, McGill University Health Center, Royal Victoria Hospital, Room L2.05, Montreal, QC, Canada H3A 1A1
ARTICLE INFO
Article history: Received 28 April 2008 Received in revised form 28 November 2008 Accepted 6 January 2009
Keywords: T-Box factor Growth factors
ABSTRACT
The transcriptional regulator TBX2 is genetically amplified in several cancers and has, in addition, important roles in development. In carcinogenesis, TBX2 regulates the cell cycle by suppressing the expression of cyclin-dependent kinase (CDK) inhibitors and destabilizes p53 by suppressing expression of ARF. In embryogenesis, however, TBX2 appears to act independently of the cell cycle or p53 and is regulated by growth factors. Tumorigenic functions of TBX2 that are independent of p53 or cell cycle regulation remain poorly under- stood. Here we used SW13 carcinoma cells which express inactive p53 and have no detect- able p16 or p21 CDK-inhibitors as a model to study these functions. Expression of TBX2 in SW13 cells had no effect on the cell cycle but promoted anchorage-independence and increased resistance to apoptotic stimuli including UV-irradiation, the cytotoxic drug doxo- rubicin and lethal endoplasmic-reticulum stress. This is a cell type-dependent effect as TBX2 overexpression in PANC1 pancreatic cancer cells which are p53-negative has no effect on colony formation or survival after irradiation. Mechanistically, in SW13 cells, TBX2 overexpression strongly reduced the activation of caspase 3, 8 and 9 following UV- irradiation but without altering the expression of the corresponding procaspases. There were, however, dramatic and specific decreases in the expression of procaspases 1 and 4. The expression of the inhibitor of apoptosis, cIAP2/BIRC3, increased in TBX2-overexpress- ing cells. TBX2 was upregulated in a PI3K-dependent manner by growth factors that are tumorigenic for SW13. Inhibition of Akt phosphorylation abrogates upregulation of TBX2 by FGF-4. Our findings identify TBX2 as a cell type-dependent survival factor under a p53-negative background, and are indicative of a potentially wider role for TBX2 in carci- nogenesis than hitherto described.
@ 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
TBX2 [1] is a member of the T-Box family of transcrip- tional regulators [2], with roles in embryonic development [2,3] and tumorigenesis [4-7]. It is often amplified in pan- creatic cancer [5], and in BRCA-1 and BRCA-2-related breast cancers [4,6,7]. TBX2 [8] and the closely related gene TBX3 [9] are highly expressed in a number of breast cancer cell
lines. TBX2 represses the transcription of two alternate products of the CDKN2A locus; p16ink4A, which is a cyclin- dependent kinase (CDK) inhibitor, and ARF [8,10], and, in addition, represses other CDK-inhibitors such as P15ink4b [8] and p21waf [11]. ARF inhibits mdm2-mediated degrada- tion of p53 [12] and therefore, by regulating ARF expres- sion, TBX2 indirectly decreases p53 stability (i.e. TBX21 → ARFĮ → mdm21(action) → p531). Thus, TBX2 and TBX3 may potentiate the mitogen-stimulated entry into the cell cycle by reducing the expression of CDK-inhibitors, while also reducing the apoptotic and growth inhibitory actions of p53 [13]. These properties are important in the bypass of senescence pathways in normal cells [8] and
* Corresponding author. Tel .: +1 514 934 1934x35239; fax: +1 514 843 1709.
some cancers including melanomas [14] and breast cancer [15]. A recent report shows that TBX2 confers resistance to cisplatin in fibroblasts [16] suggesting additional pathways by which TBX2 may regulate apoptosis.
In development, TBX2 expression is regulated by growth factors [17] and other extracellular signals, how- ever, in contrast to its proposed actions in carcinogenesis and the bypass of senescence, the developmental roles of TBX2 are often independent of either CDK-inhibitors or p53. Thus, there are no changes in the expression of CDK- inhibitors or ARF, or impaired p53 function in TBX2-/- mice, even though they die by day 14.5 p.c. [3]. Similarly TBX2, in conjunction with TBX3, regulates embryonic mammary gland formation [18] through mechanisms that are inde- pendent of ARF and p53 [18]. During cardiac development TBX2 suppresses the expression of the cell cycle gene Nmyc1 [19], and the ectopic expression of TBX2 and TBX3 decreases, rather than stimulates, cell division in the devel- oping heart [20]. TBX2 is clearly multifunctional given the disparity between its proposed roles in carcinogenesis and development, and may therefore support additional roles in cancer beyond its potential to influence the cell cy- cle or the stability of p53 [6]. Such functions would greatly extend the potential roles of TBX2 in carcinogenesis since in many cancer cells p53 is inactive, and mutations of CDK-inhibitors, particularly p16ink4a, are common. At pres- ent, however, there is limited experimental evidence sup- porting additional cell cycle or p53-independent actions for TBX2 in tumorigenesis.
In earlier experiments we noted a shared regulation of TBX2 in SW13 adrenal carcinoma cells by the secreted growth modulating protein progranulin and FGF-4 (unpub- lished). SW13 cells are anchorage-independent and tumor- igenic in mice only when they overexpress one of a small number of growth factors, such as progranulin [21], or a se- creted FGF such as FGF-4 [22], or pleiotrophin [23]. As we show below, SW13 cells have an inactive mutant p53, have undetectable levels of p16 and p21 cyclin-dependent ki- nase inhibitors and are reportedly resistant to growth sup- pression by Rb [24]. They therefore provide a very useful system to investigate those roles of TBX2 in cancer that are independent of its well-established actions on CDK- inhibitors or p53. Here we show that, in the p53-mutant SW13 cells, the elevated expression of TBX2 is cytoprotec- tive against a number of apoptotic challenges, it signifi- cantly dysregulates the activity and expression of several caspases, and upregulates the expression of the cIAP2/ BIRC3 caspase inhibitor. This cytoprotective effect of TBX2 seems to be cell type-dependent as overexpression of TBX2 in the p53-negative PANC1 cells showed no effect on colony formation or survival after stress.
2. Materials and methods
2.1. Cell culture and stable transfectants
SW13, PANC1 and MCF-7 cells (American Type Cell Col- lection, Manassas, VA) were maintained in DMEM (Invitro- gen Life Technologies, Inc., Burlington, ONT, Canada) supplemented with 10% FBS (HyClone, Logan, UT) at 37 ℃,
5% CO2. Six stable transfections were established: SW13/ pcDNA3.0, SW13/pgrn, SW13/FGF-4 and SW13/TBX2, PANC1/pcDNA3.1 and PANC1/TBX2 by transfecting empty vectors or vectors carrying full-length inserts of FGF-4 (American Type Cell Collection), TBX2 (gift, Dr. M. van Lohu- izen, Amsterdam, The Netherlands) or pgrn using Lipofect- amine transfection reagent (Invitrogen, Burlington, ONT, Canada) followed by selection using 800 µg/ml of Geneticin (G418) (Invitrogen). In order to avoid clonal variability, cells were pooled together as a population overexpressing the gene of interest. In each case, two independently transfected and selected populations were used and the phenotype was confirmed in each. Stable expression of the inserts was con- firmed by real time PCR and/or Western blot analysis.
2.2. Anchorage-independent and growth assays
Anchorage-independent cell growth in semi-solid agar was assayed as described [21] with 3000 cells plated per well in 24-well plates. Colonies were counted using an in- verted microscope. Proliferation in monolayer cultures was analyzed by plating 50,000 cells per well of a 6-well plate for 24 h in DMEM supplemented with 10% serum. After 24 h the medium was replaced either with serum free (SF) DMEM or DMEM with 5% serum, the cells were then incubated for 4 days, trypsinized and the number of viable cells counted in a hemocytometer using Trypan-blue stain- ing to exclude dead cells. Cytofluorimetric evidence for hypodiploid nuclei was obtained by propidium iodide staining as described previously [21].
2.3. Antibodies and Western blots
Antibodies against TBX2, p53, p14, p16, p21, CASP1, CASP3 (uncleaved), CASP4, FGF-4 and Actin were from San- ta Cruz Biotechnology (Santa Cruz, CA). Antibodies against CASP8, CASP9, CASP3 (cleaved), phosphorylated and non- phosphorylated Akt, p42/p44 MAPK and lysates from Jur- kat cells exposed to cytochrome c were from Cell Signaling Technology (Danvers, MA). Cells were seeded in 10 cm dishes at 2 × 106 cells in DMEM + 10%FBS, grown to 80% confluency, then washed with PBS, lysed with RIPA lysis buffer (1x PBS, 0.5% sodium deoxycholate, 0.1% SDS, 1% Nonidet P-40, 1% Aprotinin, 0.01% phenylmethylsulfonyl fluoride, and 1 mM sodium orthovanadate (Sigma Aldrich, Oakville, ONT, Canada), passed through a 23.5 G needle, and centrifuged at 15,000g for 10 min at 4 ℃. Whole cell lysate (50-100 µg) were electrophoresed in 10% SDS-PAGE, transferred onto nitrocellulose membrane and blotted as described [25].
2.4. Quantitative real time PCR (QRT-PCR)
QRT-PCR was performed as described elsewhere [26] using a Light Cycler FastStart DNA Master SYBR Green I kit (Roche Diagnostics GmbH, Mannheim, Germany). Melt- ing curve analysis confirmed the presence of a single prod- uct for every PCR primer used, and genomic contamination was excluded by amplification of a control sample without reverse transcription. Quantitation was calculated using the formula:
Relative expression ratio of target gene = (Etarget) ACPtarget(control - sample) (Eref) ACPref (control - sample) where E is the efficiency of the PCR reaction, ACP is the difference in crossing points and ref is the corresponding value for a reference gene (GAPDH or Actin), [27]. PCR primers (Sup- plementary Table 1) were synthesized by Invitrogen. All the primers were designed to anneal at 65 ℃. Each RT- PCR experiment was performed using 3 or 4 independent RNA extracts in duplicate using two independently trans- fected cell populations.
2.5. Detection of p53 mutation in SW13
Total RNA from SW13 was reverse transcribed and amplified by PCR using six sets of primers spanning p53 cDNA. The PCR products were sequenced in both ori- entations (Sheldon Biotechnology Center, McGill University).
2.6. P53 DNA-binding activity
P53 DNA-binding activity was determined using the Trans-Binding™ p53 assay kit (Panomics, Inc., Redwood City, CA). Active p53 in cell lysates binds to p53 consensus binding site oligonucleotides affixed on a 96-well plate, with the binding detected by a p53 ELISA quantitated col- orimetrically according to the manufacturer’s instructions. Results were plotted as extract concentration versus absor- bance at 450-nm.
2.7. Cell survival assays
For UV-irradiation 4.0 x 105 cells per well in 6-well plates were washed in PBS and exposed to a pulse of 150 J/m2 UV in a Stratalinker (Stratagen Inc., La Jolla, CA). Surviving cells were counted after 24 h in SF DMEM. For ER-stress apoptosis, the cells were exposed to either Thapsigargin (2 uM) for 72 h, or Tunicamycin (10 µg/ml) for 60 h (Sigma Aldrich) in SF DMEM, and surviving cells counted. Experiments were performed at least four times in duplicate. Doxorubicin hydrochloride and Caspase 1 and Caspase 4 inhibitors (Ac-YVAD-CHO and Ac-LEVD- CHO, respectively) and zVAD-fmk were from EMD Chem- ical Inc, San Diego, CA. Caspase 1 and 4 inhibition exper- iment was done by incubating cells in SFM containing 300 µM inhibitor(s) for 1 h prior to and after UV expo- sure. Recombinant TRAIL, FasL and anti 6X histidine were from R&D systems, Inc., Minneapolis, MN.
2.8. Caspase 4 colorimetric assay
Caspase 4 activity was detected with a colorimetric pNA-LEVD-cleavage kit (BioVision Research Products, Mountain View, CA). 10 million cells plated in 15 cm dishes were exposed to 2 uM of Thapsigargin or DMSO vehicle control for 72 h. Detached and adherent cells were lysed in the buffer provided, centrifuged at 13,000 rpm and the supernatant further clarified by ultracentrifugation at 80,000 rpm for 1 h at 4 ℃ using a bench top ultracentri- fuge; pNA-LEVD substrate cleavage was measured as absorbance at 405-nm.
2.9. Pathway inhibition
SW13/FGF cells were treated with 0.09 mM of PI3 K sig- naling inhibitor LY 294002 or with 0.03 mM of MAPK sig- naling inhibitor U0126 for 24 h in SF DMEM. TBX2 mRNA expression was determined by QRT-PCR. The inhibition of phosphorylation of Akt or p44/p42 MAPK was confirmed by Western blot analysis.
2.10. Microarray analysis
SW13/Vector and SW13/TBX2 cells were grown in 15 cm dishes to about 80% confluency in 10% FBS, washed twice with PBS and left 24 h in serum free medium. Total RNA was extracted using TRIzol® reagent (Invitrogen Life Technologies, Inc., Burlington, ONT, Canada) and subjected to microarray analysis using human U133 chip (Affyme- trix, Santa Clara, CA). Microarray experiments were per- formed by Genome Quebec (McGill University, Montreal, QC, Canada) using RNA from two independently transfec- ted and selected SW13/vector and SW13/TBX2 cell popula- tions. Only genes regulated +2.5-fold in both independent populations were included, and genes that varied within the vector controls were excluded.
2.11. Cellular localization of TBX2 in SW13 cells
We used immunofluorescence to detect TBX2 in SW13/ pcDNA3.0 and SW13/TBX2 cells. Briefly, cells were fixed for 10 min on cover slips using PEM buffer containing 0.1 M PIPES (ICN Biochemicals, Cleveland, OH), 5 mM EGTA and 2 Mm MgCl2-6H2O. Cells were rinsed with PBS and permeabilized with cold ethanol for 5 min at -20 ℃, washed with PBS and TBX2 was detected using TBX2 pri- mary antibody (Santa Cruz Biotechnology) followed by Alexa 488 conjugate (Invitrogen). Nuclear staining was done using 4’,6-diamidino-2-phenylindole (DAPI) (Sigma Aldrich). The localization of TBX2 immunoreactivity was visualized using an inverted fluorescence microscope.
2.12. Statistical analysis
Statistics were done by ANOVA, and significance is re- ported as two-sided p.
3. Results
3.1. Overexpression of TBX2 confers anchorage-independence on SW13 cells
SW13 cells that overexpress TBX2 show an increase in TBX2 expres- sion compared to vector-transfected controls (Fig. 1A). Cell growth in monolayers, and the fraction of cells in S-phase (Fig. 1B) were unaffected by expression of TBX2, therefore excluding a functionally significant effect of TBX2 on the cell cycle in SW13 cells. TBX2 promoted anchorage-inde- pendent colony growth in soft agar, although the SW13/TBX2 colonies were smaller than those obtained by overexpressing FGF-4 (Fig. 1C). TBX2 reduced protein levels of ARF but not p53 (Fig. 1D). The CDK-inhib- itors p16ink4a and p21 were undetectable by Western blot analysis in SW13/vector control cells or SW13/TBX2 cells, but were detected in con- trol extracts of HeLa and MCF-7 cells, respectively (Fig. 1E). Therefore the biological effects of TBX2 on colony growth in SW13 cells cannot be due to repression of CDK-inhibitors. PANC1 cells overexpressing TBX2 (Fig. 1F, left panel) showed no enhancement of colony growth in soft agar (Fig. 1F, right panel).
A
B
180.0
Cnt
TBX2
Percent cell growth
160.0
10
140.0
9
TBX2
120.0
% S-phase
8
Actin
100.0
Cnt
7
80.0
TBX2
6
60.0
5
40.0
4
20.0
3
5
2
0.0
1
RER (TBX2)
4
SFM
5% FBS
Cnt
TBX2
3
Cnt
TBX2
2
G1
G1
1
0
Cnt
Tbx-2
S G2/M
S G2/M
C
Relative number of colonies
4
p<0.001
Relative size of colonies
3.5
3
Cnt
FGF-4
TBX2
2.5
Contact
FGF
TBX2
2
2
1.5
1
0. .5
0
0
Cnt
TBX2
TBX2
FGF-4
D
TBX2
Cnt
Relative expression (ARF)
1.20
Relative expression (p53)
1.1
1.00
p53
0.80
0.9
Actin
0.60
p<0.05
0.7
0.40
0.5
ARF
0.20
0.3
Actin
0.00
0.1
Cnt
TBX2
Cnt
TBX2
E
Hela
MCF7
SW13/TBX2
SW13
F
3
Number of colonies
200
RER (TBX2)
2.5
p16
2
150
Actin
1.5
100
1
p21
0.5
50
Actin
0
0
Cnt
TBX2
Cnt
TBX2
3.2. P53 is mutated and non-functional in SW13 cells
Since TBX2 may act in part through the regulation of p53 levels by repressing ARF, we investigated the mutational status of p53 and its re- sponse to irradiation stress in SW13 cells. P53 cDNA from SW13 cells is mutated in codon 193 from CAT to TAT, i.e. His(wt)-Tyr(mut) (data not shown). As is commonly observed in cells with mutationally inactivated p53, its levels are high in SW13 compared for example to MCF-7 cells, which have wild-type p53 (Fig. 2A, no UV control lanes). UV exposure did not increase P53 protein levels in vector control or TBX2-overexpress- ing SW13 cells whereas a marked increase occurred in MCF-7 cells
(Fig. 2A), moreover there was no detectable p53 DNA-binding activity either in wild-type SW13 cells or after exposure to UV, whereas UV-irra- diation increased p53 protein levels and DNA-binding activity in MCF-7 cells (Fig. 2B). The reported destabilization of p53 by TBX2 cannot there- fore account for the biological effects of TBX2 in SW13 cells.
3.3. TBX2 confers resistance to irradiation and doxorubicin
TBX2 increased the 24 h cell survival when SW13 cells were ex- posed to 150 J/m2 UV-irradiation (Fig. 2C, left panel). In addition, FACS analysis of propidium iodide stained cells 6 h post UV exposure
A
B
0.6
SW13 noUV
SW13
MCF-7
Absorbance (450-nm)
MCF-7 noUV
0.5
SW13, UV/6hr
Post UV
MCF-7, UV/6hr
Post UV
0.4
No UV
4 hrs
24 hrs
0.3
Cnt
TBX2
Cnt
TBX2
TBX2
No UV
4 hrs
24 hrs
0.2
Cnt
0.1
p53
p53
0
Actin
Actin
-0.1
2.5
5
10
20
Lysate (ug protein)
C
CNT
Percent cell survival
D
120
p<0.01
*
100
Cnt
Percent cell survival
TBX2
120
80
100
Cnt
TBX2
80
TBX2
60
60
40
40
p<0.05
*
20
20
.
-
0
0
SW13
PANC1
0.2 mg/l
0.5 mg/l
1.0 mg/l
6 hrs post UV
2.0 mg/l
4.0 mg/l
Doxorubicin
E
Post UV
No UV
6 hrs
9 hrs
F
SW13 Cells
Cnt
TBX2
Cnt
TBX2
Cnt
TBX2
UV
-
+
+
kDa
ZVAD-fmk - -
+
Pro-CASP3
32
kDa
19
Pro-CASP3
32
*
Mature CASP3
19
17
Mature CASP3
Pro-CASP9
17
Mature CASP9
36
100
p<0.001
Mature CASP8
43
Percent survival
90
E
80
p<0.005
Actin
43
70
60
PANC1 Cells
Cnt
TBX2
Cnt
TBX2
Jurkat
50
40
30
20
Pro-CASP3
10
0
Mature CASP3
Vehicle
40uM
90uM
9hrs post UV
+
+
-
-
27hrs post UV
zVAD-fmk
-
-
+
+
demonstrated fewer hypodiploid cells in the SW13/TBX2 cultures than did the SW13/vector controls (Fig. 2C, right panel). There was
no increase in cell survival in PANC1 cells overexpressing TBX2 (Fig. 2C, left panel). To assess whether the cytoprotective effect of
TBX2 for SW13 is specific for irradiation mediated cell death, or is more general, we tested the effects TBX2 expression on the response to the chemotherapeutic drug doxorubicin. SW13/TBX2 cells were more resistant to cell death induced by doxorubicin (Fig. 2D) than were control cells.
SW13 control cells exposed to 50 ng/ml and 100 ng/ml of TRAIL and up to 1 µg/ml of FasL in the presence of 10 µg/ml of anti 6x- histidine as an activating cross linker showed no sign of apoptosis after 24 h of exposure to the recombinant proteins (not shown). This precluded the possibility of investigating the role of TBX2 in inhibiting receptor-medi- ated apoptosis in this system. We confirmed the apoptotic activity of these reagents in another cell line (not shown).
3.4. Regulation of caspase activity by TBX2 overexpression
To investigate the mechanism of cytoprotection against UV-irradia- tion provided by TBX2 expression we measured the activation of caspas- es. In SW13 cells, overexpression of TBX2 inhibited the activating cleavage of the effector caspase 3 from the precursor to the 17 kDa en- zyme 6 h and 9 h following exposure to UV-irradiation (Fig. 2E). Simi- larly the cleavage activation of the initiator caspases 8 and 9 was decreased in SW13/TBX2 cells relative to empty vector controls 6 h and 9 h post-irradiation (Fig. 2E). PANC1 showed enhanced cell death (Fig. 2C) but no caspase 3 activation 9 and 27 h following UV-irradiation (Fig. 2E, lower panel). In order to confirm that the reduction of the cas- pase-dependent apoptosis is the cause for the enhancement survival manifested in SW13/TBX2 cells, we used the general caspase inhibitor zVAD-fmk to block apoptosis immediately after UV exposure. Cleavage of caspase 3 occured after UV exposure and in the presence of zVAD- fmk, but only to an intermediate cleavage product (Fig. 2F, upper panel) which has been reported by others to be inactive [28]. Indeed zVAD-fmk was protective against apoptosis as it enhanced cell survival after UV- irradiation (Fig. 2F, lower panel).
3.5. Regulation of caspase expression by TBX2 overexpression
Expression of caspase mRNAs were investigated by QRT-PCR. Tran- scripts for caspases 1 and 4 mRNAs decreased significantly in SW13/ TBX2, but expression of other caspases including 3, 8 and 9 was unaf- fected (Fig. 3A). The protein levels of caspases 1 and 4 were lower in unstimulated SW13/TBX2 compared to control cells, while caspase 3 pro- tein levels remained unchanged (Fig. 3B). In control SW-13 cells, the spe- cific inhibition of caspase 4, but not caspase 1, provided partial protection against cell death in UV-mediated apoptosis (Fig. 3C).
3.6. TBX2 confers resistance to ER-stress mediated cell death
Caspase 4 has recently been implicated in endoplasmic-reticulum (ER) stress-related cell death [29]. To confirm that the decreased expres- sion of caspase 4 is associated with decreased activity, we stimulated ER- stress mediated cell death in SW13/TBX2 cells. Survival was greater in SW13/TBX2 cells than SW13/vector cells when ER-stress was provoked by tunicamycin or thapsigargin (Fig. 4A). Caspase 4 activity, as assessed by a colorimetric substrate conversion assay of pNA-LEVD, was lower in SW13/TBX2 than control cells after thapsigargin-stimulated ER-stress (Fig. 4B). TBX2 had no effect on mRNA expression of other ER-stress-re- lated genes, ATF-6 or PERK (Fig. 4C).
3.7. Cellular localization of TBX2 transgene in transfected SW13 cells
Supplementary Fig. 1 shows cellular localization of TBX2 transgene in control and TBX2-transfected SW13 cells. When comparing DAPI detec- tion of nuclei to Alexa detection of TBX2, it is clear that the transgene is localized in the nucleus.
3.8. Gene expression regulated by TBX2
Given that TBX2 is a transcriptional regulator, we investigated its abil- ity to regulate gene expression when expressed in SW13 cells. To mini- mize false positives due to non-specific idiosyncrasies of a particular cell population, two independently transfected and selected cell popula- tions were screened. Only genes that were regulated by 2.5-fold com- pared to the SW13/vector controls in two independently selected
A
P<0.001
2.5
2
P<0.001
RER RER
1.5
1
0.5
0
Cnt
CASP1
CASP3
CASP4
CASP7
CASP8
CASP9
B
Cnt
TBX2
Precursor
CASP1
Cleaved (p20)
CASP1
Precursor
CASP4
Precursor
CASP3
C
Relative cell survival
2
p<0.05
p<0.05
1.5
1
0.5
0
T
T
1
CASP1 Inhibitor
-
+
-
+
CASP4 Inhibitor
-
-
+
+
populations are listed in Supplementary Table 2. The TBX2-mediated reg- ulation of several genes including the apoptosis inhibitor cIAP2/BIRC3 was confirmed by real time PCR (Supplementary Fig. 2).
3.9. Growth factor expression stimulates TBX2 expression
At present the carcinogenic role of TBX2 has been confined to cancer cells in which the TBX2 gene is amplified. Since TBX2 is not, to our knowl- edge, amplified in SW13 cells, or in many other cancers, it was important to confirm our earlier observation that TBX2 expression levels were regu- lated by non genomic stimuli that are known to increase the malignancy of SW13 cells. Overexpression of progranulin or FGF-4 in SW13 cells was confirmed by Western blot analysis (Fig. 5A) and both growth factors en- hanced anchorage-independent growth as expected, with FGF-4 yielding more colonies than progranulin (Fig. 5B). The stimulation of TBX2 expres- sion by growth factors was confirmed by quantitative real time PCR (Fig. 5C and D) and Western blot (Fig. 5D, right panel), with FGF-4 being a greater stimulus for TBX2 expression than progranulin. Given the stron- ger response of TBX2 to FGF-4, subsequent experiments were performed
A
B
6
Relative cell survival
Relative cleavage of pNA-LEVD
120
5
p<0.01
100
4
80
p<0.05
TM
3
60
p<0.01
TG
2
40
1
20
0
Cnt
TBX2
0
Cnt
TBX2
C
1.6
1.4
1.2
RER
1
0.8
0.6
0.4
0.2
0
Control
ATF6
PERK
using the SW13/FGF-4 cells. The PI3K inhibitor LY 294002 reduced the expression of TBX2 in SW13/FGF-4 (Fig. 6A) whereas the MAP kinase (MEK) inhibitor U0126 increased the expression of TBX2. Phosphorylation of p44/42 MAPK and the PI3 K target Akt were both enhanced by expres- sion of FGF-4 (Fig. 6B). LY 294002 blocked phosphorylation of Akt appro- priately and had no effect on p44/42 MAPK, while U0126 blocked phosphorylation of p44/42 MAPK as expected, but increased phosphory- lation of Akt (Fig. 6B). Thus, the expression of TBX2 is sensitive to tumor- promoting growth factors and is downstream of the PI3-K pathway.
4. Discussion
The possibility that TBX2 regulates tumorigenesis through pathways that are independent of p53 and the cell cycle has been suggested [6], but the mechanisms through which it might do so are poorly understood. Losses of func- tion of p53 or CDK-inhibitors are the most commonly occurring oncogenic lesions and it is important to establish what effects, if any, TBX2 might have in cells carrying these lesions. SW13 cells provide a model to address this ques- tion. These cells are poorly transformed unless stimulated by a limited range of growth factors that include FGFs and progranulin. In addition, as we showed here, the p53 gene in SW13 cells is mutated within the DNA-binding domain at codon 193. In a global mutational analysis of p53 this mutation results in greatly impaired function [30], [data stored in the IARC TP53 Mutation Database (http://www- p53.iarc.fr/index.html)], and here we demonstrated that there is no p53 DNA-binding activity in SW13 cells before or after stimulation with UV-irradiation (Fig. 2B). Further-
more the increased expression of TBX2 does not act on the cell cycle in SW13 cells since it had no effect on either cell number or the percent of cells in S-phase (Fig. 1B), more- over, the CDK-inhibitors, p16ink4a and p21waf, that are re- pressed by TBX2 in other cell lines, were undetectable even in the parental cell line. The decrease in ARF expres- sion in SW13/TBX2 cells compared to SW13/vector cells confirms, however, that as expected, TBX2 targets the CDKN2A locus in SW13. Mdm2 was undetectable in SW13 cells (not shown), however ARF associates with pro- teins other than mdm2 [31], and this may provide further potential routes for TBX2 to influence tumorigenesis in p53-mutant cancer cells.
Notwithstanding the mutation of p53, or the absence of an effect on the cell cycle, the overexpression of TBX2 in SW13 results in important phenotypic outcomes. SW13/ TBX2 cells are more anchorage-independent than SW13/ vector cells. The ability to grow unattached to a solid sub- stratum is an important step in the development of malig- nancy [32]. The small size of the colonies, together with the lack of enhanced proliferation in monolayer growth, sug- gested that the anchorage-independence of SW13/TBX2 cells was due to decreased apoptosis, allowing the colonies to survive, but grow only slowly. Consistent with the hypothesis of a cytoprotective role in SW13 cells, TBX2 expression reduced apoptosis provoked by UV or doxorubi- cin (Fig. 2C and D) or ER-stress (Fig. 4A). This may have clinical implications by conferring increased resistance to radiation or chemotherapy. In accordance with this data,
A
B
SW13
Relative number of colonies
14
p<0.05
Vector
FGF-4
12
Pgrn
10
8
Pgrn
6
p<0.05
FGF-4
4
Actin
2
0
Cnt
Pgrn
FGF-4
C
16.0-
30.0-
15.0-
28.0-
14.0
GAPDH
20.0-
13.0-
24.0-
Fluorescence (F1)
TBX2
Fluorescence (F1)
12.0-
22.0-
11.0-
20.0
10.0-
18.0-
9.0-
0.0-
14.0
7.0-
12.0
6.0-
100-
FGF-4
5.0
FGF-4
80-
Pgrn
4.0-
Pgrn
3.0-
40-
Control
2.0-
Control
20-
1.0-
0.0
0.0-
20
-1.0
0
a
.
.
8
&-
S
2
8-
8-
%
00
4
2
0
a
0
O
2
a
8
8-
13-
4
e
38
4)
2
À
Cycle Number
Cycle Number
D
6
M
RER TBX2
5
-
4
Cnt
Pgrn
FGF-4
3
-
2
TBX2
1
C
0
Actin
L
U
U
Cnt
Pgrn
FGF-4
2
a recent report showed a link between ectopic expression of TBX2 and increased resistance to the chemotherapeutic agent cisplatin in transformed human lung fibroblasts [16]. Another report demonstrated a UV-mediated phos- phorylation and activation of TBX2 by p38 stress regulated kinase in MCF-7 breast cancer cells [33]. The cytoprotec- tive effect of TBX2 seems to be cell type-dependent as there was no increase in anchorage-independent growth or cell survival after UV exposure in the p53-mutant PANC1/TBX2 cells (Figs. 1F and 2C). Since the cytoprotec- tive effect of TBX2 in SW13 cells following irradiation must be independent of p53 (Fig. 2A and B), we investigated other mechanisms through which apoptosis might be impaired.
Overexpression of TBX2 dramatically dysregulated the caspase 3 response to UV, since the protolytic conversion of the inactive precursor to activated cleaved caspase 3 was greatly impaired in UV irradiated SW13/TBX2 cells even though TBX2 expression has no effect on pro-caspase 3 levels, (Fig. 2E). In control experiments in which caspase activity was inhibited using zVAD-fmk there was a corre- sponding resistance to UV-mediated apoptosis, confirming the relevance of the caspase effect in enhancing survival in SW13 cells (Fig. 2F, lower panel). Caspase 3 is cleaved
downstream from initiator caspases such as caspase 2, 8, 9 and 10. Caspase 8 mediates receptor-activated apoptotic events, whereas caspase 9 is activated by the mitochon- drial apoptosome. The activation of both caspases 8 and 9 were decreased following UV-irradiation in the SW13/ TBX2 cells relative to the vector control cells demonstrat- ing that TBX2 expression influenced multiple caspase pathways. In PANC1 cells, however, caspase 3 appears not to be activated in the UV-mediated apoptosis (Fig. 2E, lower panel) and TBX2 has no cytoprotective effect on these cells.
Although the mRNA levels of caspase 3, 8 and 9 are not affected by TBX2 expression, other caspases, such as casp- ases 1 and 4 exhibited highly specific decreases in mRNA and protein levels (Fig. 3A and B). In parental SW13 cells, a small, but significant, reduction in sensitivity to UV-irra- diation followed the chemical inhibition of caspase 4, but not caspase 1 (Fig. 3C), suggests that the reduction in cas- pase 4 expression contributes to the radiation-protective effects of TBX2. Caspases 1 and 4 are part of a structurally related caspase subfamily. Caspase 1 processes interleu- kin-1, hence its alternative name Interleukin-16 cleavage enzyme (ICE), but it also promotes programmed cell death [34-36]. Caspase 4 is associated with ER-stress mediated
A
12
p<0.05
Fold induction of TBX2
10
8
6
4
p<0.05
2
0
FGF-4
LY 294002
U0126
B
Cnt
FGF-4
LY294002
U0126
Phospho-Akt (Ser-473)
Phospho-p44/p42 MAPK
(Thr202/Tyr204)
Akt
p44/p42 MAPK =
cell death [29], which may contribute to tumor progression [37]. Microarray analyses and QRT-PCR revealed no changes in transcript levels for other genes that regulate the ER-stress response such as ARF6 and PERK [37]. This demonstrates a specific targeting of caspase 4 rather than a global down regulation of ER-stress genes.
SW13/TBX2 cells were resistant to the apoptotic effects of two drugs that stimulate ER-stress mediated apoptosis; tunicamycin, which initiates an unfolded-protein response [38], and thapsigargin, which deregulates cellular calcium balance [39]. The insensitivity to ER-stress mediated apop- tosis is probably due mainly to the impaired activity of the major effector caspases such as casapse 3. However, con- sistent with the decrease in caspase 4 mRNA levels, its bio- activity following treatment with thapsigargin was ~60% lower in SW13/TBX2 than in vector controls.
The dysregulation of the caspase system reported here, together with increased resistance of fibroblasts that over- express TBX2 to cisplatin [16] and the decrease in apopto- sis through p53-based pathways reported elsewhere for the closely related protein TBX3 [13], suggest that TBX2 and related proteins may influence apoptosis at several distinct levels depending on the molecular background of the cell.
Given that TBX2 is a transcriptional regulator we inves- tigated the global transcriptional response to TBX2 overex- pression in SW13 and whether the transgene was appropriately localized in the nucleus. Microarray analysis and real time RT-PCR showed significant upregulation of a
member of the IAP (Inhibitors of apoptosis) family pro- teins, cIAP2/BIRC3, in SW13/TBX2 cells compared to con- trol cells (see Supplementary Fig. 2). cIAP2 is a negative regulator of caspase 9 activation [40] and may contribute to the impaired activation of the caspase system in SW13/TBX2 cells. Many of the other genes whose expres- sion is reliably altered in SW13/TBX2 cells including TGFß2, osteopontin, the transcription suppressor BHLHB2 and the notch signaling target HEY2 (see Supplementary material) are involved in cell signaling and together these genes may enable elevated TBX2 to contribute to tumorigenesis through a number of distinct routes in addition to the ef- fects on the caspases.
TBX2 has been invoked as a tumorigenic gene in cells in which it is genetically amplified, however this repre- sents only a relatively small subset of cancers, and its role in cancers in which it is not amplified is unclear. The expression of TBX2 during development is regulated by growth factors such as FGFs [17], and here we dem- onstrate similar growth factor regulation of TBX2 in adult cancer cells. Since growth factor signaling pathways are often overactivated in cancer, this provides an additional mechanism for TBX2 to contribute to carcinogenesis in tumors in which it is not genetically amplified. The ele- vated expression of TBX2 in SW13/FGF-4 cells is PI3K- dependent (Fig. 6). Unexpectedly, however, when the ERK pathway was inhibited by U0126, TBX2 expression increased. The stimulatory effect of ERK inhibition on TBX2 expression most likely results from negative cross-talk between the ERK to the PI3K pathway [41,42], since U0126 also increased the phosphorylation of the PI3K substrate Akt (Fig. 6). Although the major growth factor signaling pathways are multifunctional, the PI3K pathway plays significant roles in the suppres- sion of apoptosis, and therefore the regulation of TBX2 by the PI3K pathway is consistent with the cytoprotective role for TBX2 reported here.
In conclusion, although research into the tumorigenic potential of TBX2 has focused mainly on the regulation of the cell cycle and other nuclear regulatory events, here we have identified new mechanisms whereby TBX2 expression may enhance tumor progression that are inde- pendent of either the cell cycle or p53. These included in- creased anchorage-independence and protection against apoptotic insults such as irradiation, doxorubicin and ER- stress. TBX2 overexpression caused cytoprotection by greatly dysregulating the casapse system. It reduced the activation of the caspase 3, 8 and 9, and inhibited the expression of caspases 1 and 4. Thus, at least in the adrenal cancer model used here, the expression of TBX2 in response to growth factors, and its ability to regulate the trans- formed phenotype independently of cell cycle regulation or the p53 tumor suppressor, parallels its activities in embryogenesis, and significantly extends the possible con- tributions that TBX2 may make to cancer development.
Conflicts of interest statement
The authors declare no conflict of interests associated with the present manuscript.
Acknowledgements
Supported by the Canadian Institute of Health Research (CIHR) and the Terry Fox Foundation Research Grant from the Canadian National Institute of Cancer (NCIC) to AB. The authors thank Dr. Luis Fernando Congote for critical read- ing of the manuscript and Huishi Toh for help with real time PCR and Western blots.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in the online version, at doi:10.1016/ j.canlet.2009.01.006.
References
[1] R.J. Bollag, Z. Siegfried, J.A. Cebra-Thomas, N. Garvey, E.M. Davison, L.M. Silver, An ancient family of embryonically expressed mouse genes sharing a conserved protein motif with the T locus, Nat. Genet. 7 (1994) 383-389.
[2] V.E. Papaioannou, L.M. Silver, The T-box gene family, Bioessays 20 (1998) 9-19.
[3] Z. Harrelson, R.G. Kelly, S.N. Goldin, J.J. Gibson-Brown, R.J. Bollag, L.M. Silver, V.E. Papaioannou, Tbx2 is essential for patterning the atrioventricular canal and for morphogenesis of the outflow tract during heart development, Development 131 (2004) 5041-5052.
[4] M. Barlund, O. Monni, J. Kononen, R. Cornelison, J. Torhorst, G. Sauter, O .- P. Kallioniemi, A. Kallioniemi, Multiple genes at 17q23 undergo amplification and overexpression in breast cancer, Cancer Res. 60 (2000) 5340-5344.
[5] E.H. Mahlamaki, M. Barlund, M. Tanner, L. Gorunova, M. Hoglund, R. Karhu, A. Kallioniemi, Frequent amplification of 8q24, 11q, 17q, and 20q-specific genes in pancreatic cancer, Genes Chromos Cancer 35 (2002) 353-358.
[6] M. Rowley, E. Grothey, F.J. Couch, The role of Tbx2 and Tbx3 in mammary development and tumorigenesis, J. Mammary Gland Biol. Neoplasia 9 (2004) 109-118.
[7] C.S. Sinclair, C. Adem, A. Naderi, C.L. Soderberg, M. Johnson, K. Wu, L. Wadum, V.L. Couch, T.A. Sellers, D. Schaid, J. Slezak, Z. Fredericksen, J.N. Ingle, L. Hartmann, R.B. Jenkins, F.J. Couch, TBX2 is preferentially amplified in BRCA1- and BRCA2-related breast tumors, Cancer Res. 62 (2002) 3587-3591.
[8] J.J. Jacobs, P. Keblusek, E. Robanus-Maandag, P. Kristel, M. Lingbeek, P.M. Nederlof, T. van Welsem, M.J. van de Vijver, E.Y. Koh, G.Q. Daley, M. van Lohuizen, Senescence bypass screen identifies TBX2, which represses Cdkn2a (p19(ARF)) and is amplified in a subset of human breast cancers, Nat. Genet. 26 (2000) 291-299.
[9] W. Fan, X. Huang, C. Chen, J. Gray, T. Huang, TBX3 and its isoform TBX3+2a are functionally distinctive in inhibition of senescence and are overexpressed in a subset of breast cancer cell lines, Cancer Res. 64 (2004) 5132-5139.
[10] M.E. Lingbeek, J.J. Jacobs, M. van Lohuizen, The T-box repressors TBX2 and TBX3 specifically regulate the tumor suppressor gene p14ARF via a variant T-site in the initiator, J. Biol. Chem. 277 (2002) 26120-26127.
[11] S. Prince, S. Carreira, K.W. Vance, A. Abrahams, C.R. Goding, Tbx2 directly represses the expression of the p21(WAF1) cyclin- dependent kinase inhibitor, Cancer Res. 64 (2004) 1669-1674.
[12] C.J. Sherr, The INK4a/ARF network in tumour suppression, Nat. Rev. Mol. Cell Biol. 2 (2001) 731-737.
[13] H. Carlson, S. Ota, Y. Song, Y. Chen, P.J. Hurlin, Tbx3 impinges on the p53 pathway to suppress apoptosis, facilitate cell transformation and block myogenic differentiation, Oncogene 21 (2002) 3827- 3835.
[14] K.W. Vance, S. Carreira, G. Brosch, C.R. Goding, Tbx2 is overexpressed and plays an important role in maintaining proliferation and suppression of senescence in melanomas, Cancer Res. 65 (2005) 2260-2268.
[15] K.M. Dobrzycka, K. Kang, S. Jiang, R. Meyer, P.H. Rao, A.V. Lee, S. Oesterreich, Disruption of scaffold attachment factor B1 leads to TBX2 up-regulation, lack of p19ARF induction, lack of senescence, and cell immortalization, Cancer Res. 66 (2006) 7859-7863.
[16] E. Davis, H. Teng, B. Bilican, M.I. Parker, B. Liu, S. Carriera, C.R. Goding, S. Prince, Ectopic Tbx2 expression results in polyploidy and cisplatin resistance, Oncogene (2007).
[17] N. Firnberg, A. Neubuser, FGF signaling regulates expression of Tbx2, Erm, Pea3, and Pax3 in the early nasal region, Dev. Biol. 247 (2002) 237-250.
[18] L.A. Jerome-Majewska, G.P. Jenkins, E. Ernstoff, F. Zindy, C.J. Sherr, V.E. Papaioannou, Tbx3, the ulnar-mammary syndrome gene, and Tbx2 interact in mammary gland development through a p19Arf/ p53-independent pathway, Dev. Dyn. 234 (2005) 922-933.
[19] C.L. Cai, W. Zhou, L. Yang, L. Bu, Y. Qyang, X. Zhang, X. Li, M.G. Rosenfeld, J. Chen, S. Evans, T-box genes coordinate regional rates of proliferation and regional specification during cardiogenesis, Development 132 (2005) 2475-2487.
[20] I. Ribeiro, Y. Kawakami, D. Buscher, A. Raya, J. Rodriguez-Leon, M. Morita, C. Rodriguez Esteban, J.C. Izpisua Belmonte, Tbx2 and Tbx3 Regulate the dynamics of cell proliferation during heart remodeling, PLoS ONE 2 (2007) e398.
[21] Z. He, A. Bateman, Progranulin gene expression regulates epithelial cell growth and promotes tumor growth in vivo, Cancer Res. 59 (1999) 3222-3229.
[22] A. Wellstein, R. Lupu, G. Zugmaier, S.L. Flamm, A.L. Cheville, P. Delli Bovi, C. Basilico, M.E. Lippman, F.G. Kern, Autocrine growth stimulation by secreted Kaposi fibroblast growth factor but not by endogenous basic fibroblast growth factor, Cell Growth Differ. 1 (1990) 63-71.
[23] W. Fang, N. Hartmann, D.T. Chow, A.T. Riegel, A. Wellstein, Pleiotrophin stimulates fibroblasts and endothelial and epithelial cells and is expressed in human cancer, J. Biol. Chem. 267 (1992) 25889-25897.
[24] A.K. Wong, F. Shanahan, Y. Chen, L. Lian, P. Ha, K. Hendricks, S. Ghaffari, D. Iliev, B. Penn, A.M. Woodland, R. Smith, G. Salada, A. Carillo, K. Laity, J. Gupte, B. Swedlund, S.V. Tavtigian, D.H. Teng, E. Lees, BRG1, a component of the SWI-SNF complex, is mutated in multiple human tumor cell lines, Cancer Res. 60 (2000) 6171-6177.
[25] Z. He, A. Ismail, L. Kriazhev, G. Sadvakassova, A. Bateman, Progranulin (PC-cell-derived growth factor/acrogranin) regulates invasion and cell survival, Cancer Res. 62 (2002) 5590-5596.
[26] C.H. Ong, Z. He, L. Kriazhev, X. Shan, R.G. Palfree, A. Bateman, Regulation of progranulin expression in myeloid cells, Am. J. Physiol. Regul. Integr. Comp. Physiol. 291 (2006) R1602-R1612.
[27] M.W. Pfaffl, A new mathematical model for relative quantification in real-time RT-PCR, Nucl. Acids Res. 29 (2001) e45.
[28] R. Gerhard, S. Nottrott, J. Schoentaube, H. Tatge, A. Olling, I. Just, Glucosylation of Rho GTPases by Clostridium difficile toxin A triggers apoptosis in intestinal epithelial cells, J. Med. Microbiol. 57 (2008) 765-770.
[29] J. Hitomi, T. Katayama, Y. Eguchi, T. Kudo, M. Taniguchi, Y. Koyama, T. Manabe, S. Yamagishi, Y. Bando, K. Imaizumi, Y. Tsujimoto, M. Tohyama, Involvement of caspase-4 in endoplasmic reticulum stress-induced apoptosis and abeta-induced cell death, J. Cell Biol. 165 (2004) 347-356.
[30] S. Kato, S.Y. Han, W. Liu, K. Otsuka, H. Shibata, R. Kanamaru, C. Ishioka, Understanding the function-structure and function- mutation relationships of p53 tumor suppressor protein by high- resolution missense mutation analysis, Proc. Natl. Acad. Sci. USA 100 (2003) 8424-8429.
[31] C.J. Sherr, Divorcing ARF and p53: an unsettled case, Nat. Rev. Cancer 6 (2006) 663-673.
[32] D.N. Carney, A.F. Gazdar, J.D. Minna, Positive correlation between histological tumor involvement and generation of tumor cell colonies in agarose in specimens taken directly from patients with small-cell carcinoma of the lung, Cancer Res. 40 (1980) 1820-1823.
[33] A. Abrahams, S. Mowla, M.I. Parker, C.R. Goding, S. Prince, UV- mediated regulation of the anti-senescence factor TBX2, J. Biol. Chem. (2007).
[34] M. Los, M. Van de Craen, L.C. Penning, H. Schenk, M. Westendorp, P.A. Baeuerle, W. Droge, P.H. Krammer, W. Fiers, K. Schulze-Osthoff, Requirement of an ICE/CED-3 protease for Fas/APO-1-mediated apoptosis, Nature 375 (1995) 81-83.
[35] H. Pei, C. Li, Y. Adereth, T. Hsu, D.K. Watson, R. Li, Caspase-1 is a direct target gene of ETS1 and plays a role in ETS1-induced apoptosis, Cancer Res. 65 (2005) 7205-7213.
[36] R.N. Winter, J.G. Rhee, N. Kyprianou, Caspase-1 enhances the apoptotic response of prostate cancer cells to ionizing radiation, Anticancer Res. 24 (2004) 1377-1386.
[37] Y. Ma, L.M. Hendershot, The role of the unfolded protein response in tumour development: friend or foe?, Nat Rev. Cancer 4 (2004) 966- 977.
[38] T. Shinohara, K. Ikesugi, M.L. Mulhern, Cataracts: role of the unfolded protein response, Med. Hypotheses 66 (2006) 365-370.
[39] G.E. Kass, S. Orrenius, Calcium signaling and cytotoxicity, Environ. Health Persp. 107 (Suppl 1) (1999) 25-35.
[40] I. Budihardjo, H. Oliver, M. Lutter, X. Luo, X. Wang, Biochemical pathways of caspase activation during apoptosis, Annu. Rev. Cell Dev. Biol. 15 (1999) 269-290.
[41] V.P. Eswarakumar, I. Lax, J. Schlessinger, Cellular signaling by fibroblast growth factor receptors, Cytok Growth Factor Rev. 16 (2005) 139-149.
[42] H. Hayashi, Y. Tsuchiya, K. Nakayama, T. Satoh, E. Nishida, Down-regulation of the PI3-kinase/Akt pathway by ERK MAP kinase in growth factor signaling, Genes Cells 13 (2008) 941- 947.