Discussion
Despite the pioneer chemotherapeutic strides made in other malignancies, no major breakthroughs have been achieved in the treatment of urothelial bladder carcinoma, with cisplatin remaining the only cornerstone for its clinical management in the past three decades. However, acquisition of resistance to cisplatin and emergence of systemic toxicity severely limit remedy’s success, ultimately resulting in failure of long-term disease remission and threat to patients’ survival [
71]. Therefore, there is an urgent need to develop more effective and less toxic regimens for bladder cancer therapy.
To this direction, and by exploiting the metabolic switch of cancer cells from mitochondrial oxidative phosphorylation to aerobic glycolysis (Warburg effect), we have herein attempted to target bladder cancer cells with 3-BrPA, a previously reported inhibitor of glycolysis. Since T24, but not RT4, cells carry activated -oncogenic- Ras, increased PI3K/Akt signaling, mutant p53 and mitochondrial defects (MPC repression), critical features for glycolytic fueling [
7,
9,
72], it seems that they (T24) significantly rely on Warburg effect-like metabolism to support their energy demands and rapid cell divisions. Although 3-BrPA has been previously used for targeting several types of cancer cells [
10,
11,
73], it has never been, so far, employed for bladder cancer treatment. Despite its initial view as a specific inhibitor of glycolysis [
12‐
16], we herein demonstrate that 3-BrPA can severely perturb the integrity of genome, transcriptome (including splicing), proteome (including glycolytic regulators) and kinome of T24 (and T24-X) cells, in a gene/transcript/protein/kinase-specific manner, successfully orchestrating bladder cancer cell elimination at the preclinical level. Besides its potency to generate DNA breaks, 3-BrPA proved able to harshly harm the transcription machinery in a gene-specific manner. In contrast to their vast majority, only few of the examined genes remained unaffected, with
FasL being the sole upregulated one in response to 3-BrPA. Moreover, none of them, except
GLUT4, presented with strong splicing silencing features, after exposure to 3-BrPA, indicating drug’s power to impair spliceosome machinery in a transcript-dependent fashion (Fig.
7, Additional file
5: Figure S5, Additional file
6: Figure S6 and Additional file
7: Figure S7).
To the same direction, our data support the notion that 3-BrPA is not just a conventional alkylating factor that promiscuously attacks cysteine residues in proteins [
12,
15,
57], but rather a powerful and multifaceted chemotherapeutic agent that strongly pyruvylates its target proteins in a highly specific and selective manner.
In toto, 3-BrPA proved capable to (a) severely inhibit phosphorylation-dependent signaling activities (e.g. Akt), (b) reduce total protein cellular contents (e.g. LC3B), (c) impose protein structural fragmentations (e.g. PARP), (d) promote protein post-translational modifications (e.g. pyruvylation; HK2), (e) upregulate typical (e.g. H2A.X) and aberrant phosphorylation-mediated signaling (e.g. 47 and 39 kDa), and (f) alter splicing patterns of selected signaling determinants (e.g. 47 kDa), following dose-, cell type- and protein-dependent modes. 3-BrPA can target proteins for pyruvylation and induce depletion of cellular ATP, hence harming and/or eliminating the respective phosphorylated forms and cognate activities (Figs.
2 and
4, and Additional file
2: Figure S2). Nevertheless, the production of 47 and 39 kDa phosphorylated kinases (Fig.
4), and the induction of
FasL gene (Fig.
7), in response to 3-BrPA, suggest that affected cells preserve marginal ATP levels to successfully encounter the bioenergetic expenses of regulated necrosis.
Drug-directed severe dysfunction of their kinome and selective collapse of their proteome compel T24 (and T24-X), but not RT4, cells to undergo apoptotic and regulated necrotic death in a dose-specific manner. The two major issues we were challenged to herein resolve regarded the molecular signatures that control bladder cancer cell sensitivity to 3-BrPA and the molecular mechanisms that orchestrate T24 necrotic demise. In terms of the first issue, mutant p53, oncogenic Ras, constitutive autophagy, aberrant PI3K/Akt, p44/42 MAPK and AMPKα signaling, together with differential expression of MCT and MPC family members were examined. Surprisingly, despite the drug-induced genotoxicity, 3-BrPA proved unable to phosphorylate and activate the either -endogenously expressed- mutant (
ΔY126) or -transiently transfected- wild-type p53 form in T24, while cells could not be rescued from the drug in the presence of functional (responsive to Doxorubicin) wild-type p53 protein. Moreover, 3-BrPA could not transcriptionally induce any of the p53
bona fide target genes examined. However, 3-BrPA and PRIMA-1 could strongly synergize to specifically eradicate the mutant p53-carrying T24 and T24-X cells (Fig.
3), therefore shaping a novel and promising preclinical platform for bladder cancer therapy. In contrast to its dispensable role in 3-BrPA-induced death of bladder cancer cells, p53 was recently associated with susceptibility of breast cancer cells to 3-BrPA [
74], thus indicating the malignant environment-dependent engagement of p53 in drug’s cytotoxic activities.
Since
B-Raf
V600E
or
K-Ras
G13D
oncogenic mutations were previously reported to favor the successful targeting of 3-BrPA to colon cancer cells [
51], their involvement in bladder cancer cells’ differential sensitivity to the drug was herein examined. Interestingly, all five cell lines presented with wild-type
B-Raf
V600
and
K-Ras
G13
gene profiles, hence implicating a role of distinct aberrant determinants, such as PIK3CA
E545K (TCCSUP) and H-Ras
G12V (T24). Indeed, the
H-Ras
G12V
oncogenic mutation in T24 provides cells with strong constitutive autophagy (Fig.
2) and high basal levels of PI3K/Akt and MAPK (p44/42 and p38) signaling (Fig.
4). Given that T24 cells require H-Ras
G12V-dependent autophagy for growth and survival [
32], the 3-BrPA-orchestrated suppression of autophagy (Fig.
2) must critically contribute to 3-BrPA-treated T24 and T24-X necrotic death. Autophagy deficiency likely increases the cellular load of abnormal mitochondria and limits ATP availability, while the combined inactivation of apoptosis and autophagy may promote necrosis, drastically restricting tumor growth [
31,
75]. Therefore, in contrast to previous reports in hepatocellular carcinoma and glioblastoma [
76,
77], it is the autophagy impairment, and not its upregulation, that drives bladder cancer cells to ATP depletion-mediated death. Similarly to colon carcinoma [
51], resistance of bladder cancer cells to 3-BrPA was conversely associated with their tolerance to glucose deprivation (Figs.
1,
5 and
6, and Additional file
10: Table S2), again uncoupling
B-Raf
V600E
and
K-Ras
G13D
oncogenic mutations from bladder cancer cell survival in hypoglycemic environments. However, upon glucose lack, the T24-specific and H-Ras
G12V-driven constitutive autophagy could recycle intracellular components to satisfy pivotal metabolic requirements, hence promoting survival and tolerance to bioenergetic stress.
Besides constitutive autophagy, T24 cells are also addicted to strong basal activities of Akt signaling (Fig.
4). Since HK2 enhances autophagy upon glucose starvation [
42], preserves mitochondrial integrity through its (HK2) Akt-mediated phosphorylation [
38] and supports oncogenic K-Ras-directed tumorigenesis [
52], we reason that 3-BrPA preferably targets the -mitochondrial- HK2 phosphorylated form. Therefore, its -putative- constitutive phosphorylation by activated Akt (inhibited by LY294002) in T24 may render -mitochondrial- HK2 vulnerable to 3-BrPA, whereas the non-phosphorylated -cytosolic- form in RT4 cannot be recognized by the drug. Indeed, the dose-dependent induction of a 145 kDa protein directly reflects the T24-specific targeting of 3-BrPA to HK2, through most likely a pyruvylation process (Fig.
7). 3-BrPA might promote the dissociation of -phosphorylated- HK2 from mitochondria in T24, as previously shown in leukemia and hepatocellular carcinoma cells [
15,
16]. In accordance with several reports [
15,
16,
38,
78,
79], HK2-mitochondria disruption could presumably impair mitochondrial integrity, propelling death of T24, but not RT4, cells.
In addition to HK2 targeting, 3-BrPA proved able to drastically downregulate several determinants of glucose homeostasis in T24 and T24-X, with GAPDH and HK2 exhibiting intriguingly distinct expression profiles in response to the drug (Fig.
7). The 3-BrPA-induced -severe- reduction of GAPDH, but not HK2, cellular contents in T24 (and T24-X) can be associated with drug’s potency to strongly inhibit GAPDH, but not HK2, glycolytic activity in hepatocellular carcinoma cells [
13,
14]. Overall, 3-BrPA could promote, in T24, (a) a pyruvylation-mediated release of phosphorylated HK2 from mitochondria, without directly affecting its enzymatic activity, and (b) a structural elimination of GAPDH, likely through protein degradation, which together with injury of Rab10, Tug and GLUT4 could cause a detrimental bioenergetic crisis (Fig.
2 and Additional file
2: Figure S2).
LY294002 (PI3K/Akt inhibitor), U0126 [MAPKK(1/2)/MAPK(p44/42) inhibitor] and AICAR (AMPK agonist) could salvage, albeit at different levels each, T24 cells from 3-BrPA cytotoxic power (Fig.
5). Since Akt and p44/42 MAPK have been previously implicated in NHE1 phosphorylation and activation [
68,
69], and NHE1 represents a
bona fide target of EIPA [
67,
70], whose administration strikingly rescues T24 from 3-BrPA (Fig.
8), we herein reason that LY294002 and U0126 implement their beneficial roles in cell survival via attenuation of Akt- or p44/42 MAPK-dependent NHE1 activity (see below). In contrast to LY294002, U0126 and AICAR, the MCT1-specific inhibitor AR-C155858 [
65] presented with a unique proficiency to completely, and not just partly, rescue T24 from 3-BrPA (Fig.
8), thus underscoring the predominant contribution of MCT1-mediated drug influx to bladder cancer cell responsiveness to 3-BrPA. Apparently, the two MPC family members examined cannot offer any kind of 3-BrPA trafficking in T24 (Fig.
8), while in RT4 they likely allow the entry of pyruvate into functional mitochondria. Besides MCT1, the T24-specific expression pattern of SMCT1 agrees for its role as another major 3-BrPA importer in bladder cancer cells (Fig.
8). Furthermore, providing that MCT4 functions as drug exporter, its significantly reduced contents in T24 (Fig.
8) could likely enhance the MCT1-mediated accumulation of 3-BrPA in T24. Intriguingly, and despite their tolerance to 3-BrPA (Fig.
1), RT4 cells presented with notable downregulation of SAPK/JNK, GAPDH and Tug proteins (Figs.
4 and
7) and strong upregulation of unspliced
GLUT4 transcripts (Fig.
7). Therefore, it may be either the free, but slow, diffusion of 3-BrPA [
60], or the ability of MCT4 to import the drug [
63], with low affinity and rate, that specifically operates in RT4 cells. Perhaps, 3-BrPA effective concentration in RT4 is below a certain threshold and only few sensitive targets can be affected, such as GAPDH and
GLUT4-specific spliceosomal components, without, however, perturbing cell growth and survival. Altogether, and in accordance with previous reports in different cellular settings [
57,
62], we conclude that MCT family members undoubtedly control the uptake of 3-BrPA by bladder cancer cells and critically modulate their sensitivity to the drug.
Strikingly, in addition to MCT1, NHE1-mediated macropinocytosis proved also essential for 3-BrPA-induced cytotoxicity in T24 (Fig.
8). Since AR-C155858 and EIPA are both able to completely, and not just partly, salvage T24 from 3-BrPA (Fig.
8), we reason that their respective molecular targets, MCT1 and NHE1 [
65,
67,
70], must control cell sensitivity and death, in response to the drug, in a highly concerted and inter-dependent fashion. Hence, we propose that MCT1 and NHE1 could require each other for optimal activity. During the H-Ras
G12V-driven and NHE1-dependent macropinocytotic entry of 3-BrPA in T24 (an EIPA-inhibited process; Fig.
8), NHE1 likely exchanges intracellular H
+ with extracellular Na
+ ions. Providing that MCT1 and NHE1 share the same membrane micro-domain that orchestrates macropinocytosis, the NHE1-mediated enrichment of extracellular micro-environment with H
+ could strongly facilitate the coupled and simultaneous reverse transport of H
+ with 3-BrPA from the acidified extracellular micro-space into T24 cells through engagement of MCT1 carrier (an AR-C155858-inhibited process; Fig.
8). Now, the imported H
+ could be efficiently exported again through H
+-regulated activation of NHE1 [
80,
81], thus maintaining the functional integrity of macropinocytosis machinery in T24.
In toto, it seems that 3-BrPA can enter T24 via the two inter-dependent routes of MCT1 transporter and NHE1-mediated macropinocytosis.
Entry of 3-BrPA in T24 is followed by quick induction of dose-specific cell death. Pharmacological interventions indicated the critical roles of PARP- and MLKL/Drp1-mediated necrotic axes, together with a novel Nec-7-targeted pathway (Fig.
2). Low drug dose activates caspase-dependent apoptosis, featured by typically cleaved PARP (and Lamin A/C), whereas high drug doses force T24 to non-caspase-mediated death, characterized by irregular PARP (and Lamin A/C) cleavage profiles. This and the ability of PARP inhibitor PJ-34 [
18,
20] to significantly rescue 3-BrPA-treated T24 and T24-X (Fig.
2) provide evidence for implication of PARP over-activation in drug-orchestrated cell death. 3-BrPA-induced genotoxic stress (Fig.
3), together with the rather aberrant signaling functions of 47 (inhibited by U0126) and 39 kDa novel forms (Fig.
4) could cause uncontrolled PARP activity. Indeed, maximal or sustained activation of PARP requires its direct phosphorylation by p44/42 MAPK or JNK1 kinase, respectively, under cellular stress [
20,
82‐
84]. 3-BrPA-driven over-activated PARP extensively consumes NAD
+ pools, which together with suppressed autophagy and collapsed glycolysis (Figs.
2 and
7) result in detrimental depletion of ATP stores and, finally, necrotic death of T24 (and T24-X) (Figs.
1‐
2 and Additional file
2: Figure S2). Bioenergetic stress can be alleviated by AICAR-mediated activation of AMPK metabolic sensor and subsequent replenishment of ATP and NADPH pools (Fig.
5) [
43,
85]. PARP can also mediate caspase-independent cell death by triggering translocation of AIF from mitochondria to nucleus, where it (AIF) promotes large-scale chromatinolysis [
18‐
21,
84,
86]. However, and in contrast to a previous report in hepatoma cells [
87], we were unable to detect transport of AIF from cytoplasm (presumably mitochondria) to nucleus upon exposure of T24 to 3-BrPA. Furthermore, T24 cells presented with significantly reduced AIF cellular content after their exposure to necrotic drug doses (Additional file
8: Figure S8A-C).
Pharmacological intervention in the other two fundamental necrotic routes RIPK1/MLKL/Drp1 and p53/CypD proved the dispensable roles of RIPK1 (targeted by Nec-1 and Nec-5) [
19,
21‐
25] and CypD (targeted by CsA) [
28] mediators in 3-BrPA-driven death of T24 (Fig.
2). Neither p53 (Fig.
3) nor CypD (Fig.
2) can essentially direct cytotoxicity of 3-BrPA in T24, therefore uncoupling p53/CypD-complex necrotic activity (which mainly controls mitochondrial PTPC opening) from drug-induced regulated necrosis of bladder cancer cells. However, the ability of NSA, a
bona fide human MLKL inhibitor [
22‐
24], to provide cells with a strong survival advantage against 3-BrPA (Fig.
2) clearly indicates engagement of novel necrotic pathways that orchestrate regulated necrosis of 3-BrPA-treated T24 cells in RIPK1-independent but RIPK3/MLKL-dependent manner. Indeed, in contrast to RIPK3 (when elevated), RIPK1 does not seem to have an obligatory role in necroptosis (challenged by TNF) signaling [
88], while catalytically active RIPK3 can mainly cause MLKL-mediated necroptotic death in RIPK1-deficient environment of mouse embryonic fibroblasts [
89]. Most interestingly, via employment of chemically enforced dimerization systems, it proved that -artificial- homodimerization or oligomerization of RIPK3 cannot only eliminate reliance on RIPK1 but is also sufficient to trigger MLKL-dependent necroptosis [
90,
91]. Accordingly, by exploiting its selective alkylating capacity, 3-BrPA could promote, in T24 cells, the homodi(oligo)merization and sequential (auto)phosphorylation-dependent activation of RIPK3 that induces phosphorylation-driven recruitment of MLKL to the activated necrosome [
22,
23]. Now, MLKL may be licensed to orchestrate the downstream (potentially interrelated) necrotic routes of Drp1-mediated mitochondrial fragmentation [
24] and TRPM7-directed Ca
2+ influx [
92], both likely contributing to T24 regulated necrosis in response to 3-BrPA. The abilities of NSA and Mdivi-1 to significantly rescue T24 cells from 3-BrPA (Fig.
2) render MLKL and Drp1 major determinants of drug’s cytotoxicity in bladder carcinoma.
During necroptosis, phosphorylated MLKL forms trimers that translocate to plasma membrane to promote TRPM7-dependent entry of Ca
2+ ions [
92]. Alternatively, necroptosis execution can be mediated by translocation of MLKL tetramers to plasma membrane, consequent influx of Na
+ ions and, eventually, membrane rupture due to increased osmotic pressure [
93]. Remarkably, the
bona fide target of EIPA, which (EIPA) strikingly salvages T24 from 3-BrPA (Fig.
8), is the principal Na
+/H
+ exchanger NHE1 [
67,
70]. Hence, 3-BrPA-induced oligomerization of MLKL might direct translocation of the protein to plasma membrane, wherein it facilitates NHE1-mediated influx of Na
+ ions. Now, upon EIPA administration in T24, NHE1 is inhibited, Na
+ entry is impaired, osmotic stress is prohibited, plasma-membrane integrity is preserved and cells are, finally, protected from necrotic power of 3-BrPA (Fig.
8). It seems that NHE1 may direct not only the early (macropinocytosis) but also the late (osmotic pressure and membrane damage) stages of regulated necrosis in 3-BrPA-treated T24 cells.
Another route of RIPK-mediated necroptosis has proved to implicate p44/42 MAPK and SAPK/JNK signaling that propels AP-1-dependent activation of
TNFα gene [
47,
94]. Given that
FasL gene requires SAPK/JNK-driven engagement of AP-1 for its (
FasL) transcriptional induction [
95], a novel necrotic axis of RIPK3/SAPK/JNK might also operate in 3-BrPA-treated T24 cells. 3-BrPA-directed stimulation of RIPK3 could promote aberrant SAPK/JNK signaling (39 kDa; Fig.
4), sequentially targeting downstream AP-1-dependent activation of
FasL gene in T24 upon drug exposure (Fig.
7). Now, membrane-anchored or secreted FasL protein can amplify necrotic signaling via a positive feedback loop.
In toto, we herein demonstrate, for the first time, that 3-BrPA can be successfully employed to eradicate, at the preclinical level, human bladder cancer cells with highly oncogenic molecular signatures, thus underscoring drug’s potential to be embodied in the clinical practice for disease therapy, either as single agent or in cocktail regimens. Its strong cytotoxic synergism with PRIMA-1 foreshadows the development of 3-BrPA-based new protocols with potent chemotherapeutic capacity against urothelial bladder malignancies.
Methods
Drugs, reagents and chemicals
3-BrPA, PJ-34, Nec-1, Nec-5, Nec-7, Mdivi-1, CsA and UK-5099 were obtained from Sigma-Aldrich (Missouri, USA). NSA and PRIMA-1 were provided by Merck Millipore-Calbiochem (Merck KGaA, Darmstadt, Germany). LY294002, U0126 and AICAR were supplied by Cell Signaling Technology Inc. (Massachusetts, USA). Nec-1 (besides Sigma-Aldrich reagent) and EIPA were purchased from Santa Cruz Biotechnology Inc. (Texas, USA). AR-C155858 was provided by AdooQ BioScience (California, USA). Doxorubicin was obtained from EBEWE Arzneimittel GmbH (Unterach, Austria), while Taxol was supplied by Bristol-Myers Squibb (New York, USA). Rabbit polyclonal antibodies against Caspase-3, Caspase-9, PARP, Lamin A/C, Atg5, Atg7, Atg12, Beclin-1, LC3B, Akt, p-Akt-Ser473, FoxO1, p-FoxO1-Thr24/p-FoxO3a-Thr32, GSK-3α, p-GSK-3α/β-Ser21/9, p44/42 MAPK, p-p44/42 MAPK-Thr202/Tyr204, p38 MAPK, p-p38 MAPK-Thr180/Tyr182, SAPK/JNK, p-SAPK/JNK-Thr183/Tyr185, mTOR, p-mTOR-Ser2448, AMPKα, p-AMPKα1-Ser485/p-AMPKα2-Ser491, p53, p-p53-Ser15, p-p53-Ser392, H2A.X, p-H2A.X-Ser139, p-AS160-Thr642, Rab10, Tug and Actin (Pan-Actin) were purchased from Cell Signaling Technology Inc. (Massachusetts, USA). Rabbit monoclonal antibodies against p-Akt-Thr308, GSK-3β, p-GSK-3β-Ser9, S6, p-S6-Ser235/236, p-AMPKα-Thr172, HK2 and AS160 were obtained from Cell Signaling Technology Inc. (Massachusetts, USA). Rabbit polyclonal antibodies recognizing ICAD, GAPDH and GLUT1 were supplied by Santa Cruz Biotechnology Inc. (Texas, USA). Mouse monoclonal antibody against Caspase-8 was provided by Cell Signaling Technology Inc. (Massachusetts, USA). Mouse monoclonal antibody against GLUT4 was obtained from Santa Cruz Biotechnology Inc. (Texas, USA). Rabbit polyclonal antibodies recognizing MCT1 and MCT4 were supplied by Merck Millipore (Merck KGaA, Darmstadt, Germany). Rabbit polyclonal antibodies against SMCT1 and MPC1 were purchased from Novus Biologicals LLC (Connecticut, USA). Rabbit polyclonal antibody recognizing MPC2 was provided by Proteintech Group Inc. (Illinois, USA). Anti-rabbit IgG-HRP and anti-mouse IgG-HRP secondary antibodies used in Western blotting were supplied by Sigma-Aldrich (Missouri, USA). Anti-rabbit IgG-DyLight® 488 secondary antibody used in immunofluorescence was obtained from Thermo Fisher Scientific Inc. (Massachusetts, USA). ECL Western blotting reagents were purchased from GE Healthcare Life Sciences-Amersham (Buckinghamshire, England). Gene-specific oligonucleotide primers were synthesized by Metabion (Steinkirchen, Germany) and Eurofins Genomics-Operon Biotechnologies (Alabama, USA). All other chemicals were provided by Sigma-Aldrich (Missouri, USA) and AppliChem GmbH (Darmstadt, Germany).
Cell lines and culture conditions
The RT4, RT112, T24 and TCCSUP human cell lines used herein have been generated from urothelial carcinomas of the bladder. RT4 cells were obtained from ECACC-Sigma-Aldrich (Missouri, USA), while RT112 cells were kindly provided by Professor John R. Masters (London, England). T24 and TCCSUP cells originated from ATCC-LGC Standards GmbH (Wesel, Germany). T24-X cells derived from T24-specific tumor xenografts, sequentially created in SCID mice (Additional file
1: Figure S1). Cell cultures were grown in complete DMEM medium, supplemented with 10 % FBS, at 37 °C and 5 % CO
2. DMEM with no glucose (glucose-free) was purchased from Thermo Fisher Scientific Inc.-Life Technologies™-Gibco® (Massachusetts, USA). All other cell culture media and reagents were supplied by Merck Millipore-Biochrom AG (Merck KGaA, Darmstadt, Germany).
Cell viability MTT assays
Cells were seeded onto 48-well plates at a confluency of ~60 % (unless stated otherwise) and treated with different doses of 3-BrPA, in the presence or absence of the indicated inhibitors (or activators), for 24 h (except described differently). Then, cells were incubated with MTT solution, for 4 h, and the formazan crystals produced were dissolved in pure isopropanol. Spectrophotometric absorbance was measured in a Dynatech MR5000 ELISA microtiter plate reader (Dynatech Laboratories, Virginia, USA) at 550 nm, using 630 nm as wavelength of reference. Each assay was repeated three times, using three wells per experimental condition. Regarding Figs.
2 and
5, MTT viability rates obtained from each cocktail of 3-BrPA plus inhibitor/activator were compared to the respective ones of 3-BrPA alone and sequentially normalized according to values derived from inhibitor/activator only, versus its cognate solvent: data were presented in fold (x) of inhibitor/activator-induced cell survival increase.
Flow cytometry
Control and 3-BrPA-treated cells were initially stained with 20 μl of AnnexinV-FITC solution, for 20 min at 4 °C in the dark, and subsequently incubated with 10 μl of 7AAD solution, for 15 min at 4 °C in the dark. All cell preparations were analyzed within 30 min by flow cytometry, using a Beckman Coulter-Cytomics FC500 cell sorter (Beckman Coulter Inc., California, USA).
T24-specific bladder cancer xenografts in SCID mice: the new T24-X cell line
NOD.CB17-
Prkdc
scid
/J (SCID) immunodeficient mice (The Jackson Laboratory, Maine, USA) were used for bladder cancer xenografts establishment. Mice were subcutaneously inoculated with ~10
6 T24 cells per animal and closely followed till the development of tumors. Each tumor was carefully excised for cell subculture and serial passage to another SCID mouse. The whole procedure was repeated four successive times. Finally, cells extracted from a tumor of the fourth xenograft passage established a new line named T24-X (X: xenograft) (Additional file
1: Figure S1). Animals were treated according to Greek laws (2015/92), guidelines of European Union and European Council (86/609 and ETS123, respectively), and in compliance with standards for human care and use of laboratory animals (NIH, USA, assurance no. A5736-01).
Western blotting
Approximately 40 μg of whole-cell protein extracts were separated by SDS-PAGE in 10-15 % gels and subsequently transferred onto nitrocellulose membranes (Whatman-Schleicher & Schuell GmbH, Dassel, Germany). Blocking process was carried out through treatment of membranes with TBS-T containing 5 % NFM (or 5 % BSA), for 2 h at room temperature. Each primary antibody was added at a concentration of 1:1000, for 2 h at room temperature and 16 h at 4 °C, while the appropriate IgG-HRP secondary antibody (anti-rabbit or anti-mouse) was used at a dilution of 1:2000, for 2 h at room temperature. Immunoreactive bands were visualized by ECL reactions, following manufacturer’s instructions. Actin was used as protein of reference.
Cellular ADP and ATP contents of control and 3-BrPA-treated cells were determined by a bioluminescence-based assay, using the ApoSENSOR™ ADP/ATP Ratio Assay Kit (BioVision Inc., California, USA), according to manufacturer’s advice. Processed samples were read in a Tecan Infinite® M200 microplate reader (Tecan Austria GmbH, Grödig, Austria) and obtained luminescence signals were normalized to the number of cells. For lactate detection, whole-cell lysates of control and 3-BrPA-treated cells were processed through the Lactate Assay Colorimetric Kit (BioVision Inc., California, USA), following provider’s instructions, and the optical densities of analyzed samples were measured at 570 nm. Lactate production values were normalized to the number of cells.
DNA sequencing of PCR products
Genomic DNA from bladder cancer cells was isolated and subsequently amplified by PCR using
B-Raf and
K-Ras gene-specific primers flanking the V600 and G12/G13 cognate codons, respectively (Additional file
9: Table S1). Cycle sequencing of purified (with Sephadex® G50) PCR products was carried out using the BigDye® Terminator Sequencing Kit (Thermo Fisher Scientific Inc.-Life Technologies™-Applied Biosystems®, Massachusetts, USA) and the processed samples were analyzed on an ABI Prism® 310 Genetic Analyzer (Thermo Fisher Scientific Inc.-Life Technologies™-Applied Biosystems®, Massachusetts, USA). Also, total RNA from 3-BrPA-treated (T24) cells was reverse transcribed and two
GLUT4-specific PCR products, of 411 bp (
Ex7-
Ex8) and 336 bp (
Ex9-
Ex10), were purified and sequenced as described above (Fig.
7e, Additional file
5: Figure S5 and Additional file
9: Table S1). All the obtained sequences were compared to available genome assemblies created by the Genome Reference Consortium (
http://www.ensembl.org/index.html).
Transient transfection
T24 cells were grown on 6-well plates, in the absence of antibiotics, to a final confluency of ~90 %. Cells were, then, incubated for 6 h with 500 μl of Opti-MEM® I Reduced Serum Medium (Thermo Fisher Scientific Inc.-Life Technologies™-Gibco®, Massachusetts, USA), containing 4 μg of purified plasmid DNA [CMV control (empty) vector or CMV-p53 (over-expressing the wild-type human p53 protein) vector; kindly provided by Professor Jean-Christophe Marine (Leuven, Belgium)] and 10 μl of Lipofectamine® 2000 (Thermo Fisher Scientific Inc.-Life Technologies™-Invitrogen™, Massachusetts, USA). Transfection medium was replaced by complete DMEM, supplemented with 10 % FBS, and p53 gene expression was tested after 48 h of cell growth, at 37 °C and 5 % CO2, in the presence or absence of 3-BrPA (or Doxorubicin).
RT-sqPCR
Total RNA from control and 3-BrPA-treated cells was extracted following a Trizol-based protocol (Thermo Fisher Scientific Inc.-Life Technologies™-Ambion®, Massachusetts, USA). RNA (1 μg) was reverse transcribed using an oligo(dT)
12–18 primer and the M-MLV enzyme (Thermo Fisher Scientific Inc.-Life Technologies™-Invitrogen™, Massachusetts, USA). cDNA was amplified by sqPCR, with a Biometra T3000 Thermocycler (Biometra GmbH, Goettingen, Germany), using gene-specific oligonucleotide primers (Additional file
9: Table S1). PCR fragments were resolved in 2-3 % agarose gels, according to standard procedures.
GAPDH served as gene of reference.
Immunofluorescence
Cells were seeded on poly-L-lysine coated slides (Thermo Fisher Scientific Inc., Massachusetts, USA) and treated with 3-BrPA, for 4 h at 37 °C and 5 % CO
2 (Fig.
8). Then, they (together with control slides) were fixed with 4 % paraformaldehyde, for 20 min at 37 °C and 5 % CO
2, and permeabilized with 0.3 % Triton X-100, for 20 min at 37 °C and 5 % CO
2. Slides were blocked with 1 % BSA, for 90 min at 37 °C and 5 % CO
2, and subsequently incubated with a rabbit polyclonal antibody against MCT1 (1:200), for 60 min at room temperature and 16 h at 4 °C. The anti-rabbit IgG-DyLight® 488 secondary antibody was used at a dilution of 1:250, for 2 h at room temperature. Cells were observed under a Nikon Digital Eclipse C1 confocal laser scanning microscope (Nikon Corporation, Tokyo, Japan).
Statistical analysis
Statistical significance of differences observed in drug-treated versus control cell values (in the absence or presence of an inhibitor/agonist) was determined using unpaired, two-sided Student’s t-test. Data were reported as mean ± standard deviation of the mean. P < 0.001 was considered statistically significant.