The antagonism between MCT-1 and p53 affects the tumorigenic outcomes

Background MCT-1 oncoprotein accelerates p53 protein degradation via a proteosome pathway. Synergistic promotion of the xenograft tumorigenicity has been demonstrated in circumstance of p53 loss alongside MCT-1 overexpression. However, the molecular regulation between MCT-1 and p53 in tumor development remains ambiguous. We speculate that MCT-1 may counteract p53 through the diverse mechanisms that determine the tumorigenic outcomes. Results MCT-1 has now identified as a novel target gene of p53 transcriptional regulation. MCT-1 promoter region contains the response elements reactive with wild-type p53 but not mutant p53. Functional p53 suppresses MCT-1 promoter activity and MCT-1 mRNA stability. In a negative feedback regulation, constitutively expressed MCT-1 decreases p53 promoter function and p53 mRNA stability. The apoptotic events are also significantly prevented by oncogenic MCT-1 in a p53-dependent or a p53-independent fashion, according to the genotoxic mechanism. Moreover, oncogenic MCT-1 promotes the tumorigenicity in mice xenografts of p53-null and p53-positive lung cancer cells. In support of the tumor growth are irrepressible by p53 reactivation in vivo, the inhibitors of p53 (MDM2, Pirh2, and Cop1) are constantly stimulated by MCT-1 oncoprotein. Conclusions The oppositions between MCT-1 and p53 are firstly confirmed at multistage processes that include transcription control, mRNA metabolism, and protein expression. MCT-1 oncogenicity can overcome p53 function that persistently advances the tumor development.

The activity of p53 exerts paradoxically anti-apoptotic and pro-survival effects, which are essential for the development of an organism and may turn p53 into a tumor promoter. As a comprehensive guardian of genome integrity, p53 confers the survival-promoting advantages of cancer cells [18]. More substantial evidence have emerged that p53 protects cells from the genotoxin-induced apoptosis [19][20][21]. Though p53 induces Bax activation and apoptosis, relocating the p53 protein to mitochondria does not trigger tumor cell death, conversely grants apoptotic resistance to ionizing radiation [22]. Moreover, p53 reduces the oxidationinduced DNA damage and apoptosis [23][24][25]. Overall, p53 has its dark side that enhances the cell surviving mechanism and potentially inititates tumorigenicity. Exploration of p53 antagonists or p53 downstream targets which are implicated in tumorigenesis, is thus a very important task.
The regulations in opposition between p53 and MCT-1 have now been verified in vitro and in vivo. The wildtype p53 targeting the MCT-1 gene promoter could affect the presentation of MCT-1 mRAN and protein.
Reciprocally, MCT-1 depresses p53 gene promoter, mRNA stability, and protein function. Moreover, the reactivation of p53 cannot restrain the MCT-1 tumorigenic impacts on H1299 (p53 null) lung cancer cells xenografted mice and the stimulation of p53 repressors (MDM2, Pirh2, and Cop1). As well, the oncogenic MCT-1 persistently promotes the xenograft tumorigenicity of A549 (p53 wild-type) lung cancer cells. These data reveal that MCT-1 advances cellular malignancy and tumorigenic potency independent of p53 status.
By quantifying the overall MCT-1 mRNA levels with qRT-PCR analysis ( Figure 1C), MCT-1 mRNA levels in the p53 gene-restored H1299 cells (control + p53) were found to be decreased to 71% of that of sample without the p53 expression (control + vector). Consistent with decrease in cellular MCT-1 mRNA levels by p53, the levels of exogenic MCT-1 mRNA (MCT-1 + p53) (6.78) also dramatically reduced to 54% of that of the vector controls (MCT-1 + vector) (12.54) as well. These data demonstrate that the p53 reactivation can effectively repress MCT-1 gene presentation.
Further analysis was examined whether p53 reduction conversely improved MCT-1 expression in normal breast epithelial MCF-10A cells with wild-type p53 and regular genetic background ( Figure 1D). MCT-1 protein levels were detected by the specific Ab against a synthetic peptide (a.a. 72-88) of MCT-1 polypeptide. Following etoposide (ETO) genotoxin treatment for 4 h, cellular p53 was accumulated and functionally activated. MCT-1 amount was found to be a 2-fold increase after p53 silencing (p53 shRNA) relative to the non-silence control (MOCK) (lanes 1 vs. 3). Moreover, ectopic MCT-1 protein (V5-tag) was recognized by the V5-epitope Ab that showed a 1.62-fold elevation after p53 knockdown comparative to the non-silence group (lanes 2 vs. 4). These validate that p53 presence actually counteracts MCT-1 protein production.
The unexpected data indicated that intrinsic MCT-1 protein was dramatically inhibited after ectopic MCT-1 expression ( Figure 1E). As well, the autoregulation of MCT-1 gene presentation was identified as the pGL3-MCT-1 promoter activity was diminished by ectopic MCT-1 to 67% of that of control group. Even so, the entire MCT-1 protein amounts (ectopic plus intrinsic) still promoted to a 2.2-fold induction as compared with control group. These data establish for the first time that MCT-1 controls itself, via a feedback loop involving the promoter downregualtion.
Interaction of p53 with the MCT-1 promoter region MCT-1 promoter region was searching for the consensus p53-binding element, 5'-RRRC(A/T)(T/A)GYYY-3', using the NTI vector program. Six potential p53-binding sites were identified at the MCT-1 promoter region that located between nucleotides (nt.) -1301 and -801 ( Figure 2A). ChIP analysis was studied whether the MCT-1 promoter DNA associated with the activated p53 protein under ETO genotoxic stress in MCF-10A cells ( Figure 2B). Immune complex of p53 antibody (p53 Ab-IP) were   The half-life (t 1/2 ) of MCT-1 mRNA was analyzed by qRT-PCR after actinomycin D (5 μg/ml) treatment for the indicated time. Quantitative data was acquired as normalized with 18S rRNA levels. Cellular MCT-1 mRNA turnover was greater induced by p53 expression (control + p53) than p53 null condition (control). The exogenic MCT-1 mRNA decayed quickly in p53 renovation (MCT-1 + p53) versus (vs.) without p53 expression (MCT-1). (C) As determined with qRT-PCR analysis, the p53 knock-in H1299 moderately depressed intrinsic MCT-1 mRNA levels (control vs. control + p53). As well, the exogenic MCT-1 mRNA levels were significantly depressed by p53 ( PCR-amplified with the primers (Additional File 2) for the MCT-1 promoter region that correspondingly produced DNA fragment sizes of 166, 173 and 199 base pair (bp). Conversely, the nonspecific (NS) site located at the MCT-1 coding region (nt. 1~149) was undetectable in the p53 Ab-IP complex. The RNA polymerase II specifically recognized GAPDH gene as a positive control, but no DNA associated with normal IgG. Using a qPCR analysis to quantify ChIP results, the 166 and 199 bp locations exhibited higher associations with p53 than that of the 173 bp region (p < 0.002) ( Figure 2C), indicating their differential connections with p53 protein. Furthermore, the parental (p53 null) or the p53 gene restored H1299 cells (p53-reconstituted) were conducted with ChIP studies ( Figure 2D). The MCT-1 promoter DNA was only detectable in the p53-IP complex of the p53-reconstituted sample but not in the parental group. Though no p53 repressor element (p53TRE) is noticed in the MCT-1 promoter region, the relation of p53 protein with MCT-1 promoter may obstruct MCT-1 gene transcription. Electrophoretic-mobility shift assay (EMSA) was investigated whether p53 protein closely interacted with the MCT-1 promoter sites. Biotin-labeled probes (166, 173, and 199 bp), covering the promoter region from -1301 to -801 (Figure 2A), were prepared and incubated with nuclear extracts (NE) from MCF-10A cells after etoposide (ETO) treatment that stimulated and stabilized p53 protein. The biotinylated 166 bp probe formed a complex with the nuclear protein as indicated by a mobility shift ( Figure 3A, lane 3). This specific DNA-protein complex was dramatically competed by a 100-fold excess of the p53 consensus oligonucleotides (lane 4) but not the mutant p53-responsive element (lane 5). Furthermore, p53 existence in the complex was proved by generating a super-shifted band after incubation with the p53-specific antibody (lane 6). But, no obvious complex formed between the p53 Ab and the DNA probe alone (lane 1).
On the other hand, the biotinylated 173 bp probe revealed a specific interaction with the nuclear protein ( Figure 3B, lane 2), which was greatly attenuated by adding a 100-or 200-fold excess of p53 consensus oligonucleotide (lanes 3 and 4). A super-shifted band was evidently induced by the p53 Ab, confirming the presence of p53 protein in the complex (lane 6). On the contrary, the DNA-protein complex was unable to be depleted by the mutant p53-responsive element (lane 5). As well, no detectable complex was produced between the p53 Ab and the DNA probe alone ( Figure 3B, right panel).
Another DNA-protein complex was recognized while the nuclear extracts reacting with the biotinylated 199 bp probe ( Figure 3C clarified p53 being in the DNA-protein complex (lane 5). Moreover, no specific DNA-protein complex was generated between p53 Ab and the DNA probe ( Figure 3C, right panel). These data have evidently proved that p53 protein interacts with the promoter of MCT-1 gene.

MCT-1 inhibits p53 promoter activation and protein expression
ChIP assay was again evaluated if the MCT-1 protein reciprocally associated with p53 gene promoter ( Figure 4A). The p53 promoter-specific primers ranging from -420 to -84 were identified whether the p53 promoter DNA contained in the IP complex of MCT-1 Ab. In a PCR amplification analysis, approximately 8.9% of p53 promoter DNA was discovered in MCT-1 immune complex (IP). Control experiments had identified that GAPDH gene specifically associated with RNA polymerase II, but no genomic DNA was coupled with a normal IgG. Different from p53 specific binding with the MCT-1 promoter DNA (Figure 3), no interaction was detected between the p53 promoter DNA and MCT-1 protein in the EMSA study (data not shown). It is possible that MCT-1 could cooperate with other undisclosed molecule(s) that closely interact with the p53 gene promoter. The effects of wild-type MCT-1 (WT) and MCT-1 mutants that were directly mutated on Serine 118 residue (S118) to Alanine (S118A), Aspartic acid (S118D), and Glutamic acid (S118E) were studied. The dephosphorylation mutant (S118A) significantly improved the pGL3-p53 promoter function, but pGL3-p53 promoter was still repressed by cells expressing the phosphorylation-mimetic MCT-1 (S118D and S118E) and wild-type MCT-1. Statistical data was assessed with the Student's t test. **p < 0.002; ***p < 0.0001.
The proximal p53 promoter region (-188 to +12) contains a full promoter activity in the response to DNA damage [37,38]. To study whether ectopic MCT-1 antagonized p53 promoter activity, the p53 promoter DNA (-188 to +23) segment was cloned into the pGL3-Luc basic vector and then transfected into H1299 cells. The luciferase activity of pGL3-p53 promoter in ectopic MCT-1 group was decreased approximately to 70% of the respective control group (control) ( Figure 4B), indicating that MCT-1 functionally inactivated p53 promoter. The activity of CMV promoter of pCDNA3.1 that was cloned into the pGL3-Luc basic vector was not affected by the MCT-1, excluding the possibility that ectopic MCT-1 non-specifically affected the promoter function (Additional File 1B).
To analyze if MCT-1 status directly involved in p53 gene deactivation, MCT-1 gene was knocked down by MCTS1 shRNA in H1299 cells that reduced cellular MCT-1 protein to different degrees (high, medium, low) ( Figure 4C). By contrast with suppression of MCT-1 protein levels, the pGL3-p53 promoter activity was progressively improved. As compared with low p53 promoter activity in high MCT-1 context, the reporter activity was conversely elevated up to a 2-fold induction while MCT-1 declined significantly. These verify that MCT-1 plays a critical role in regulation of p53 gene promoter.
We speculated that MCT-1 also repressed the p53 gene reactivation in H1299 cells, the metabolism of exogenic p53 mRNA was analyzed after de novo gene transcription being inhibited by actinomycin D ( Figure 4D). The remaining of p53 mRNA quantity was assessed with qRT-PCR analysis at each time point. Unlike exogenic p53 mRNA decayed quickly under MCT-1 oncogenic influence (t 1/2 = 2.66 h) (MCT-1 + p53, close square), the p53 mRNA was relatively stable in control group (t 1/2 = 4.28 h) (control + p53, open triangle). Further assessment with qRT-PCR study, we found that the total p53 mRNA quantities in ectopic MCT-1 cells (MCT-1 + p53) were dropped to 68% of that of controls (control + p53) ( Figure 4E). Taken together, these results firstly illustrate that MCT-1 inhibits the overall p53 mRNA expression by decreasing p53 promoter function and p53 mRNA stability.
MCT-1 overexpression activating the extracellular signal-regulated kinase activity (ERK) is link with p53 degradation (35). To test the functional domain responsible for p53 promoter deactivation ( Figure 4H), the Serine 118 (S118) residue of the potential MAPK kinase phosphorylation site on MCT-1 protein was specifically mutated to Alanine (S118A). As well, MCT-1 (S118) residue was modified to Glutamic acid (S118E) and to Aspartic acid (S118D) that were mimetic to the phosphor-S118 MCT-1. Surprisingly, only the S118A mutant restored the p53 promoter activity significantly, but the phosphorylation-mimetic MCT-1 (S118E and S118D) still decreased p53 promoter to an extent comparable to wild-type (WT). These suggest that the serine 118 residue on MCT-1 is essential and sufficient for inhibition of p53 promoter function.
Cell migratory ability was moreover investigated by Fluorimetric cell migration assay ( Figure 5C). While culturing in the complete media, cell mobility was notably enhanced in p53 knockdown (control-p53 and MCT-1-p53) but unaffected by inducing MCT-1. Cell motility mainly enhanced by loss of p53 function may be related with Rho signaling pathway that controls actin cytoskeleton organization as literatures demonstrate [39].
To confirm MCT-1 tumorigenic potency in a p53 wild-type background, A549 lung adenocarcinoma cells ectopically expressed MCT-1 oncogene also slightly suppressed p53 accumulation in the response to ETO genotoxin ( Figure 5E). Following A549 cells subcutaneously injected into the nude mice, the tumorigenic results had evidently proved that MCT-1 oncogene strongly promoted the tumor development to a 12.8-fold increase in comparison with the control A549 xenografts ( Figure 5D). Therefore, MCT-1 oncogenicity could go beyond p53 function in the tumorigenic development.

MCT-1 promotes tumorigenicity independent of p53 status
It was unidentified whether wild-type p53 gene reconstitution in the p53 null background relieved MCT-1 oncogenicity. The impact of p53 restoration on chromosome copy number was surveyed by a cytogenetic study [36]. The intrinsic gene mutations (chromosome amplification) in H1299 background (copy number = 101) did not obviously improve after p53 gene reconstitution (copy number = 97) or change as MCT-1 overexpressed (copy number = 97). These suggest that p53 reactivation probably cannot alter MCT-1 tumorigenic outcomes.
To explore whether MCT-1 oncogene continually antagonizes p53 function in vivo, different types of H1299 cells (control, MCT-1, control + p53, MCT-1 + p53) were subcutaneously inoculated into athymic BALB/c mice. The xenograft tumor burdens were enhanced dramatically in the mice engrafted with MCT-1 and MCT-1 + p53 expressed cells as compared with their corresponding control and control + p53 cells (p < 0.0001) ( Figure 6A). Moreover, the MCT-1 xenografts produced drastically larger tumors containing with higher hemoglobin levels (p < 0.0001) and vascular counts (p < 0.0001) than those were identified in the control and control + p53 xenografts. Though p53 functionally activated p21 expression ( Figure 4F), the wild-type p53 gene transfer still unsuccessfully repressed tumor growth in ectopic MCT-1 background that concomitantly increased in micro-vascularization as evaluated with the endothelial marker CD31 immunohistology staining ( Figure 6B). As a result, the tumorigenicity and angiogenecity are not suppressed by p53 renovation; probably due to p53 only function effectively in the early tumor initiation stage. Once the tumors have developed, p53 activation disables to repair the genetic mutations and control the tumor growth.
The interrelation between p53 and MCT-1 in the tumors were subsequently verified by immunohistochemistry study   Figure 6 Reactivation of p53 cannot prevent MCT-1 tumorigenicity. (A) Different types of H1299 cancer cells (control, MCT-1, control + p53, MCT-1 + p53) were inoculated into the athymic nude mice. As indicated with CD31 immunostaining analysis, the vascular index (capillaries/field) was calculated by six randomly selected fields from each tumor biopsy. Tumor weights, hemoglobin amounts, and capillary densities were promoted in the MCT-1 and MCT-1 + p53 xenografts. (B) Immunohistology study of the tumors with CD31 Ab revealed a higher density of microvessels (indicated with arrows) in the MCT-1 and MCT-1 + p53 xenografts than those identified in the control and control + p53 xenografts. (C) Immunohistology data indicated the proteins (brownish) and H&E counterstaining (blue) in the tumors. The basal levels of MCT-1 protein expressed in control and control + p53 samples (a, c). MCT-1 proteins were intensively enhanced in the ectopic MCT-1 tumors (b), whereas the MCT-1 + p53 tumors reduced in MCT-1 concentration (d). The control and MCT-1 tumors were p53 null (e, f). Lower p53 levels were noticed in the MCT-1 + p53 tumors (h) than that of the control + p53 xenografts (g). Cop1, Pirh2 and MDM2 were all enriched in the MCT-1 tumors (j, n, l, p, r). MDM2 was also enriched in the control + p53 xenografts (s), but it was dramatically decreased in the MCT-1 + p53 xenografts ( Figure 6C, a-t). Consistent with in vitro cellular results, MCT-1 was also decreased markedly in the tumors with p53 expression (MCT-1 + p53) (d). However, MCT-1 was produced highly in the tumors without p53 (MCT-1) (b). Though p53 was greatly restored in the H1299 background (control + p53) (g), it was still comparatively reduced because ectopic MCT-1 induction (MCT-1 + p53) (h). Intriguingly, the p53 suppressors, Cop1 and Pirh2, were stimulated predominantly in ectopic MCT-1 background (j, n, l, p). Although MDM2 amounts were significantly enhanced either by ectopic MCT-1 (MCT-1) (r) or by p53 restoration (control + p53) (s), the p53-mediated MDM2 induction in tumors was declined strikingly while simultaneously expressing MCT-1 and p53 (MCT-1 + p53) (t). These protein expressions were furthermore inspected in the tumors ( Figure 6D). Due to p53 influence, lower intrinsic and ectopic MCT-1 levels were detected in MCT-1 + p53 tumors than in MCT-1 tumors (lanes 2 vs. 4). On the other hand, less p53 quantities observed in the MCT-1 + p53 tumors than in control + p53 ones (lanes 3 vs. 4). These indicate that MCT-1 works against p53 in the tumor development. Consistent with the findings in MCF-10A ( Figure 1E), the auto-regulation of MCT-1 was manifestly detected in vivo because endogenic MCT-1 decreased in quantity as ectopic MCT-1 expressed (lanes 1 vs. 2).
In support of the tumorigenic outcomes, the phosphor-activation of AKT and MAPK were found to be enhanced in MCT-1 + p53 xenograft tumors relative to the other cohorts (control, control + p53, MCT-1) (Additional File 3A). Additionally, the integrin-β4 was enriched particularly in MCT-1 + p53 xenograft tumors. Furthermore, the molecules involving in the oncogenic potential, H-Ras and HIF-1α mRNA levels, all showed approximately 1.6-fold increases in MCT-1 + p53 xenografts compared with those in the control + p53 xenografts (Additional File 3B). Stimulations of these anti-apoptotic molecules, which can enhance cancer cell proliferation and survival mechanisms, emphasize that the p53 reactivation under MCT-1 oncogenic stress fails to slow down the tumor development.

Discussion
The antagonism between MCT-1 oncogene and p53 tumor suppressor MCT-1 oncogene is a dangerous foe to p53 function, playing multiple roles in promoting chromosome instability and tumor growth [36]. Constitutively activating MCT-1 decreases p53 protein via a proteasome pathway [35]. We now demonstrate that MCT-1 reduces p53 mRNA levels accompanied with p53 gene promoter inactivation and p53 mRNA destabilization, which well correspond to inhibit p53-p21 pathway ( Figure 4B-4F). Conversely, MCT-1 gene interference stimulates p53 gene promoter and improves p53-p21 expression ( Figure  4C and 4G). Moreover, the original results have also demonstrated that p53 reciprocally interacts with the MCT-1 gene promoter that potentially suppresses MCT-1 oncogenicity in a feedback mechanism ( Figure  2B-2D and Figure 3).
The mutual counteractions between p53 and MCT-1 at the gene and protein stages are comparable to the negative regulation between MDM2 and p53 [44,45]. As well, the reciprocated transcriptional repression between MCT-1 and p53 resembles the mechanism of Twist oncogenic activity that obstructs the p53 tumor-suppressive function [46]. Another intriguing finding is that the induction of MCT-1 exerts a self-directed inhibition on intrinsic MCT-1 protein (Figure 1D-1E, and 6D), by which the overall MCT-1 protein levels can be systematically controlled using an autonomous regulation in its promoter function ( Figure 1E). The integral self-control could determine the steady state of MCT-1 activity that may critically regulate cell growth or tissue homeostasis. Similar effect has been identified that Myc overexpression contributes to tumorigenesis and myc expression is controlled through an autoregulatory circuit in non-transformed cells, by which elevated Myc protein amounts lead to down-regulation of myc transcription [47]. As well, overexpression of c-myc gene leads to a significant decrease in endogenous N-myc levels [48].

Reactivation of p53 cannot suppress MCT-1 tumorpromoting effect
The tumor promotion in MCT-1 expressed H1299 xenografts represents the synergistic consequences of p53 null and MCT-1 induction ( Figure 6). Our data have demonstrated that p53 reactivation cannot compromise the tumorigenic results induced by MCT-1 oncogene. MCT-1 + p53 xenograft mice thus develop significantly larger tumors with higher hemoglobin levels and micro-vessel density than the other xenograft tumors (control, MCT-1, control + p53). We have previously shown that p53 gene add-in cannot rescue the p53-deficient cells from MCT-1 oncogenic impact on genome destabilization, but actually increases incidence of aneuploidy from 42.8% to 95% [36]. Thus, p53 renovation fails to inhibit MCT-1-induced aneuploidization that could predispose to many carcinogenic endpoints as the documents report [49]. Another important fact is that MCT-1 oncogene confers cellular resistance to the oxidative pressure depending on p53 function ( Figure 5A). In the other way, the genotoxn-induced cytotoxicity is reduced particularly when p53 is abrogated in ectopic MCT-1 cells ( Figure 5B). But independent of MCT-1 function, cell migratory ability is promoted predominantly by p53 deficiency (Figure 5C), which could be coupled with the signaling activation of Rho pathway [39]. Regardless of p53 function, the tumor-promoting consequence is still largely promoted in A549 (p53 wild-type) lung cancer cells with a constitutive activation of MCT-1 ( Figure 5D), further revealing that MCT-1 oncogenicity could overcome p53 action in the tumor development.
Genetic evidence has implicated that α6β4 integrin signaling in promoting tumor angiogenesis and invasion [50]. These can be enhanced by HIF-1α and Ras upexpression as well [51][52][53][54]. A selective enhancement of pro-survival molecules (Integrin-β4, p-AKT, p-MAPK, H-RAS and HIF-1α) under MCT-1 oncogenic stress could increase cancer cell growth and angiogenecity that are substantially advantageous for tumor development (Additional file 3). The MDM2-p53 pathway has been recognized as an ideal therapeutic target for cancer treatment [55]. MCT-1 promotes angiogenesis that might be achieved by deregulating p53 downstream targets, such as, TSP1, VEGF, and COX-2 [56][57][58]; by inhibiting the MDM2-HIF-1α interaction [59,60]; or by enhancing the Twist-HIF-1α regulation [53]. Understanding of the crosstalk between MCT-1 and p53 in depth may facilitate the development of a new promising cancer therapeutic strategy that improves the therapeutic efficacy.
How do MCT-1 and p53 counteract each other? MCT-1 modulates p53 degradation through the extracellular signal-regulated kinase activity (ERK) because the ERK antagonist effectively restores p53-p21 expression [35]. In a negative feedback loop, p53 may deactivate ERK function to change MCT-1 stability [27]. Our important novel findings indicate that the S118A mutant of MCT-1 fails to inhibit p53 promoter activity but that is still affected by the mimetic ERK-phosphorylated MCT-1 (S118E and S118D) ( Figure 4H). For that reason, MCT-1 could regulate the p53 gene promoter involving ERK pathway, or direct interaction with ERK molecule [27,35]. In addition, p53 regulates and represses RNA poly III transcription activity that may control MCT-1 protein synthesis or the oncogenic effects on cell growth as the literatures indicate [61,62]. MCT-1 may also functionally resemble E6 and MDM2 oncoproteins, releasing RNA poly III from repression by p53 that highly enhances pol III transcription activity for protein synthesis, cell growth, and malignant phenotypes [63,64].

Conclusions
Our results uncover an important reciprocated regulation between MCT-1 oncogene and p53 tumor suppressor. Achieving a counterbalance between them may determine tumor prevention or development. The wildtype p53 gene reactivation is not capable to suppress the tumor growth promoted by MCT-1. Thus, MCT-1 gene knockout or dysfunction of MCT-1 activity could be another significant stratagem for inhibition of the tumorigenicity.
The CMV promoter was removed from pCDNA3.1 (+)/hygro vector using MluI and NheI and inserted into the pGL3-Luciferase basic vector with the same restriction sites to generate the CMV reporter construct (pGL3-CMV).

Site-directed mutagenesis on MCT-1
Three PCR primer sets were designed to generate the mutant strands of MCT-1 on Serine 118 residue (S118). The primer set for S118A included the forward primer 5'-GTCCAGGCTTAACTGCTCCTGGAGCTAAG-3' and the reverse primer 5'-CTTAGCTCCAGGAGCAGT-TAAGCCTGGAC-3'. The primer set for S118D included the forward primer 5'-CATGTGTCCAGGCT-TAACTGACCCTGGAGCTAAGCTTTAC-3' and the reverse primer 5'-GTAAAGCTTAGCTCCAGGGT-CAGTTAAGCCTGGACACATG-3'. The primer set for S118E included the forward primer 5'-CATGTGTC-CAGGCTTAACTGAGCCTGGAGCTAAGCTTTAC-3' and the reverse primer 5'-GTAAAGCTTAGCTC-CAGGCTCAGTTAAGCCTGGACACATG-3'. Following the manufacturer's protocol for the GeneEditor™in vitro site-directed mutagenesis system (Promega, Madison, WI), the insertion of wild-type MCT-1 constructed on pGEX-5X-1 plasmid was used as the mutagenesis template. The plasmid DNA was denatured, phosphorylated, annealed with the mutagenic oligonucleotides, and incubated with T4 DNA polymerase and T4 DNA ligase (Promega) at 37°C for 90 min. Mutant plasmids were transformed into BMH 71-18 mutS competent cells and selected with the GeneEditor™antibiotic selection mix, and subsequently transformed into high-efficiency JM109 competent cells followed by the selection of the ampicillin and GeneEditor™antibiotic selection mix. For long term storage, the mutants were transformed into the DH5α competent cells.

Chromatin immunoprecipitation (ChIP) assay
ChIP experiments were performed according to the manufacturer's protocol (Upstate Biotechnology, Lake Placid, NY). MCF-10A (p53 proficient) or p53-restored H1299 (2 × 10 7 ) cells were exposed to 40 μM etoposide for 4 h, fixed with 1% formaldehyde for 10 min at room temperature, neutralized with 125 mM glycine for 5 min, washed twice with PBS, and the cells were scraped off with PBS containing the protease inhibitor cocktail. Cell pellets were suspended in a 400 μl SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl, pH 8.0, protease inhibitor cocktail) and incubated on ice for 15 min followed by shearing of the genomic DNA into 200-1000 bp fragments by a sonicator (Bioruptor UCD-200). After cleaning the insoluble materials by centrifugation, supernatants were diluted with a 900 μl ChIP dilution buffer (0.01% SDS, 1% Triton X-100, 1.2 mM EDTA, 167 mM NaCl, 16.7 mM Tris-HCl, pH 8.1, protease inhibitor cocktail). Samples were pre-cleared with a 60 μl salmon sperm DNA/Protein G agarose slurry for 1 h at 4°C. An aliquot (10 μl) of the supernatants were kept as input materials, and the remaining samples (990 μl) were respectively incubated with 2 μg p53 Ab, 2 μg MCT-1 Ab, 1 μg RNA polymerase II Ab (positive control), or 1 μg normal mouse IgG (negative control) for 24 h at 4°C . The protein-DNA immune complexes were incubated with 60 μl salmon sperm DNA/Protein G agarose slurry for 1 h at 4°C. Beads were washed sequentially with the low salt buffer (20 mM Tris-HCl, pH 8.0, 2 mM EDTA, 150 mM NaCl, 0.1% SDS, 1% Triton X-100), the high salt buffer (20 mM Tris-HCl, pH 8.0, 2 mM EDTA, 500 mM NaCl, 0.1% SDS, 1% Triton X-100), the LiCl buffer (250 mM LiCl, 1% IGEPAL-CA630, 1 mM EDTA, 10 mM Tris-HCl, pH 8.0, 1% deoxycholic acid), and then rinsed twice with TE buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA). Afterward, the protein-DNA complexes were eluted from the beads with 1% SDS (200 μl) at room temperature for 15 min. To reverse the cross-linked protein-DNA complexes, samples were diluted to 50 mM NaCl followed by incubation with 10 μg RNaseA and 10 μg proteinase K at 65°C for 24 h. The eluted DNA was purified with a PCR Purification Kit (Qiagen, Valencia, CA) and subjected to the conventional PCR and q-PCR. Conventional PCR products were resolved on a 2% agarose gel and stained with ethidium bromide. For q-PCR analysis, we employed the SYBR green system using the ABI Prism 7900 Fast Real-Time PCR system and determined the threshold cycle numbers (Ct). All the relative Ct values were normalized to the inputs and compared between samples.

Electrophoretic-mobility shift assay (EMSA)
EMSA was conducted with a Gel-Shift kit according to the manufacturer's protocol (Panomics, Fremont, CA).
The nuclear extracts were prepared after MCF-10A (2 × 10 7 ) cells were exposed to 40 μM etoposide (ETO) for 4 h. The biotin-labeled MCT-1 promoter probes (166, 173 and 199 bp) corresponding to the nucleotides -1301 to -1135, -1142 to -969, and -1000 to -801 on the promoter regions were PCR amplified by forward and reverse primers as listed (Additional File 2). The PCR-amplified DNA probes were clarified by gel extraction kit (Qiagen). Nuclear extracts (5 μg) were pre-incubated with 1X EMSA binding buffer and 1 μg poly d(I-C) for 5 min at room temperature followed by incubation with 30 ng of biotin-labeled MCT-1 probe at 15°C for 30 min. The competition experiments were performed by including a 100-or 200-fold excess of unlabeled wild-type or mutant p53 consensus sequences in the reactions for 20 min prior to incubation with the biotin-labeled probe. For the super-shift assay, 1 μg of p53 antibody (SC-126 X) (Santa Cruz) was pre-incubated with the reaction for 1 h prior to adding the probe. Protein-DNA complexes were resolved with 6% non-denaturing polyacrylamide gel in 0.5X Tris-borate/EDTA buffer (100 mM Tris, 90 mM boric acid, 1 mM EDTA) at 4°C and transferred to an Immobilon positively-charged nylon membrane (Millipore, Billerica, MA) for 1 h at 300 mA. The transferred oligonucleotides were immobilized by UV crosslinking for 3 min. The membranes were reacted with the blocking buffer followed by reaction with Streptavidin-HRP and development with ECL reagent.

Cell apoptotic analysis
To evaluate apoptotic cell death, MCF-10A cells were treated with 5 μM H 2 O 2 or 40 mU Bleomycin for 24 h followed by staining with a Annexin V apoptosis detection kit (BD Biosciences) for 15 min. Afterward, apoptotic cells were evaluated by BD FACS Calibur Flow Cytometry (Becton-Dickinson, San Jose, CA).

Cell migration assay
MCF-10A cells were essayed for migratory ability with QCM™24-Well Fluorimetric Cell Migration Array Kit (Chemicon International Inc., Temecula, CA). Cells (5 × 10 5 cells) were seeded in the culture chamber with an 8 μm pore size polycarbonate membrane. Five hundred microliter of serum-free or the complete DMEM/F-12 medium was added to the lower chamber. After incubating for 16 h at 37°C in a CO 2 incubator, the non-migratory cells were carefully removed and the chamber membranes were inserted into a fresh well with 225 μl pre-warmed Cell Detachment Solution for 30 min in a 37°C incubator to detach cells, followed by adding 75 μl Lysis Buffer/ CyQuant GR® dye solution for 15 min at room temperature. Reaction mixtures (200 μl) were added into a 96-well micro-titer plate for detection of fluorescence absorbance at excitation/emission filter sets 485/530 nm using a Hidex Plate Chameleon (SisLab, Milano, Italy) apparatus.

Tumorigenicity, hemoglobin assay, and immunohistochemistry studies
Eight-week-old female BALB/c nude mice (BALB/cAnN-Foxn1nu/CrlNarl) were injected with H1299 cancer cells (control, MCT-1, control + p53, MCT-1 + p53). This animal experiment was approved by Animal Use Protocol in National Health Research Institutes (NHRI-IACUC-096049-A). Each mouse was inoculated with 2 × 10 6 cells suspended in 100 μl RPMI medium at both subcutaneous sites. When tumor sizes had reached approximately 4-6 mm, the tumors were resected and weighed. The portions of tumor tissues were processed for hemoglobin levels, immunoblotting, qRT-PCR, and immunohistochemistry analysis as previously described [36].