The ShcA adaptor protein transduces oncogenic signals downstream of receptor tyrosine kinases. We show here that breast tumors engage the ShcA pathway to increase their metabolism. ShcA signaling enhanced glucose catabolism through glycolysis and oxidative phosphorylation, rendering breast cancer cells critically dependent on glucose. ShcA signaling simultaneously increased the metabolic rate and flexibility of breast cancer cells by inducing the PGC-1α transcriptional coactivator, a central regulator of mitochondrial metabolism. Breast tumors that engaged ShcA signaling were critically dependent on PGC-1α to support their increased metabolic rate. PGC-1α deletion drastically delayed breast tumor onset in an orthotopic mouse model, highlighting a key role for PGC-1α in tumor initiation. Conversely, reduced ShcA signaling impaired both the metabolic rate and flexibility of breast cancer cells, rendering them reliant on mitochondrial oxidative phosphorylation. This metabolic reprogramming exposed a targetable metabolic vulnerability, leading to a sensitization of breast tumors to inhibitors of mitochondrial complex I (biguanides). Genetic inhibition of ShcA signaling in the Polyoma virus middle T (MT) breast cancer mouse model sensitized mammary tumors to biguanides during the earliest stages of breast cancer progression. Tumor initiation and growth were selectively and severely impaired in MT/ShcA-deficient animals. These data demonstrate that metabolic reprogramming is a key component of ShcA signaling and serves an unappreciated yet vital role during breast cancer initiation and progression. These data further unravel a novel interplay between ShcA and PGC-1α in the coordination of metabolic reprogramming and demonstrate the sensitivity of breast tumors to drugs targeting oxidative phosphorylation.

Significance: This study uncovers a previously unrecognized mechanism that links aberrant RTK signaling with metabolic perturbations in breast cancer and exposes metabolic vulnerabilities that can be targeted by inhibitors of oxidative phosphorylation. Cancer Res; 78(17); 4826–38. ©2018 AACR.

Metabolic reprogramming represents an emerging hallmark in order for cancers to meet their energetic and biosynthetic demands (1). To achieve this, cancer cells coordinate ATP-producing (glycolysis and oxidative phosphorylation) and ATP-consuming (protein, nucleotide, and lipid synthesis) processes. This reprogramming is often coupled to distinct nutrient dependencies, exposing targetable metabolic vulnerabilities that are being exploited in cancer clinical trials. For example, Warburg-like cancers undergo glycolysis, even when oxygen levels are plentiful. Pyruvate is preferentially converted to lactate instead of being fully metabolized through the citric acid cycle (CAC) and oxidative phosphorylation for further ATP generation (2). This adaptive response permits increased biosynthesis to support cell division. Warburg-like tumors compensate for inefficient ATP production by increasing glucose uptake, which forms the basis for 18Fluorodeoxyglucose (FDG)–positron emission tomography (PET) imaging as a diagnostic tool to detect malignant lesions (3). Indeed, a nonhydrolyzable glucose analogue (2-deoxy-d-glucose) is being tested as a therapeutic in Warburg-like cancers (4), including in combination with tyrosine kinase inhibitors (5, 6).

Some aggressive cancers also use glutamine to increase ATP production and synthesis of CAC intermediates through reductive carboxylation to support lipid synthesis (7, 8). This requires the activity of two glutaminases (GLS1, GLS2), which convert glutamine to glutamate. Glutamate can be used as an intermediate for amino acid synthesis, redox balance, or it can enter the CAC through its conversion into α-ketoglutarate (8). Indeed, glutaminase inhibitors exert potent antineoplastic effects in a subset of cancers, including triple-negative breast cancers (9).

Cancers that are strictly glutamine addicted require a functional CAC to meet their energetic demands. This is consistent with the observation that decreased glutamine flux through the CAC sensitizes tumors to the antineoplastic effects of metformin (10), a biguanide that is commonly used to treat type II diabetes (11). Biguanides, including metformin and phenformin, block complex I of the electron transport chain (12–14) to inhibit mitochondrial ATP production, rendering cells critically dependent on glycolysis (10, 12, 15). This forms the basis for increased sensitivity of tumors to glycolysis inhibitors in combination with biguanides (16).

Breast cancers are classified into distinct subtypes, including ER+, HER2+ and triple negative. Of these, HER2+ and triple negative breast cancers are associated with the worst outcome and are driven by receptor tyrosine kinase (RTK) signaling. These breast cancers display significant metabolic rewiring, including increased rates of glucose and glutamine metabolism (17). RTKs increase breast tumor growth, in part, by promoting glycolytic metabolism (18–20). The ShcA adaptor is a modular protein that transduces phospho-tyrosine (pY)–dependent signals downstream of RTKs (21). Genetic studies showed that the ShcA pathway is essential for breast cancer progression (22–25).

The rapid growth of aggressive breast tumors requires the coordination of two fundamental processes. First, tumors must increase angiogenesis, to acquire the requisite oxygen and nutritional support to fuel their growth (26). Second, cancer cells must capitalize on this rise in nutrient supply with a simultaneous increase in their metabolic rate to meet the energetic demands of biosynthesis required for unrestrained cell proliferation. Given previous observations that RTKs engage ShcA to coordinately increase breast cancer proliferation and angiogenesis (23, 25), we reasoned that this pathway would also play a central role in metabolic rewiring of breast cancer cells. Indeed, we identified a novel role for ShcA in coordinating metabolic reprogramming of breast cancers. ShcA upregulates the expression and activity of the key metabolic regulator, PGC-1α, in breast cancer cells, leading to an elevated global bioenergetic capacity and increased glucose dependency. Attenuation of ShcA signaling reduces the metabolic rate of breast cancer cells and makes them more dependent on glutamine. Moreover, the antineoplastic effects of biguanides are potentiated when ShcA signaling is impeded in mammary tumors. Thus, the ShcA/PGC-1α axis is a novel determinant of metabolic reprogramming and sensitivity to metabolic disrupters, such as biguanides.

Cell lines

NMuMG-NT2197 are immortalized NMuMG mouse mammary epithelial cells that were obtained from the ATCC and subsequently transformed with an oncogenic variant of rat ErbB2 (Neu) and have been described previously (27). The stable cell lines overexpressing wild-type ShcA (ShcAWT) or a ShcA-Y239/240/313F mutant (Shc3F) were also described (25). ErbB2/Shc3F cells were further transfected with pQCXIB/PGC1α or pQCXIB empty vector to generate (PGC1α O/E) and Shc3F (EV) respectively. To generate the PGC-1α/pQCXIB vector, PGC-1α was subcloned from pcDNA-f:PGC1 (Addgene Plasmid #1026). The MT/ShcA+/+ and MT/Shc3F/+ cell lines used in this study were generated from MMTV/MT transgenic mouse mammary tumors and described elsewhere (25, 28). Nutrient deprivation media consists of all the components of control media except for DMEM, where no glucose DMEM (0 mmol/L glucose; Cat #319-061-CL, Wisent), low glucose DMEM (5 mmol/L glucose; Cat #319-010-CL, Wisent), or no glutamine DMEM (0 mmol/L glutamine; Cat #319-025-CL, Wisent) was used instead. For the nutrient starvation experiments, the media were refreshed every 2 to 3 days. All cell lines were not cultured for longer than 1 month prior to replenishing with a fresh stock of cells. Cell authentication was not performed as the NMuMG cell line used was purchased from the ATCC and the MT/ShcA cell lines were established by our group from MMTV/MT transgenic tumors. All cell lines were routinely tested for Mycoplasma using the Mycoprobe Mycoplasma Detection Kit (R&D Systems; Cat #CUL00IB).

Mice

MMTV/MT have been described (28). Mice expressing mutant ShcA protein containing tyrosine-to-phenylalanine point mutations at residues 239/240/313 (Shc3F) under the control of the endogenous ShcA promoter have been described (29). Nu/Nu and SCID-beige mice were purchased from The Charles River Laboratories. For mammary fat pad injection, 5 × 104 cells were injected into the fourth mammary fat pad of female mice (6–10 weeks of age). Tumor growth was monitored every 2 days via caliper measurements, and tumor volumes were calculated as previously described (25). Phenformin (Cat #P296900, Toronto Research Chemicals) treatment was carried out via daily intraperitoneal injection (50 mg/kg) for the orthotopic studies. For the transgenic studies, MMTV/ShcA+/+ and MMTV/ShcA3F/+ transgenic mice were monitored for tumor onset by biweekly physical palpation and following detection of one tumor-bearing gland, were either given water alone or received 200 mg/kg phenformin in their drinking water with 50 mg/mL of sucrose for 4 to 5 weeks, where fresh phenformin was administered 2 to 3 times per week. All animal studies were approved by the Animal Resources Council (ARC) at McGill University and comply with guidelines set by the Canadian Council of Animal Care.

Seahorse respirometry

Extracellular acidification rates (ECAR) and oxygen consumption rates (OCR) were measured using an XF24 Seahorse instrument (Extracellular Flux Analyzer, Agilent) per the manufacturer's instructions. Briefly, cells were seeded at 30 000 cells per well in 250 μL of culture media and, after incubating overnight in a 37°C incubator, cells were washed twice with XF base media (Agilent), supplemented with 25 mmol/L glucose, 4 mmol/L glutamine, and 1 mmol/L sodium pyruvate. A final volume of 525 μL supplemented XF media was added, and the plate was set to incubate for 1 hour in a CO2-free incubator at 37°C. OCR and ECAR were obtained by repeated cycles of mix (3 minutes), pause (3 minutes), and measurement (3 minutes). Measurements were normalized on protein content at the end of the experiment. The percentage of ECAR rate and the percentage of respiration rate were obtained by normalizing ECAR and OCR values, respectively, to that of ErbB2/ShcAWT cells.

Metabolites extraction from cells and stable isotope tracer experiment

Cells were seeded in 6-well plates and were then treated with phenformin (0.5 mmol/L; Sigma-Aldrich) or ddH2O (control) for 24 hours. For stable isotope tracer experiments, cells were preincubated with media containing unlabeled glucose and glutamine for 2 hours, and the media were then changed by labeled media ([U-13C]-glucose (25 mmol/L in media; Cambridge Isotope Laboratories; CLM-1396; d-glucose ([U-13C], 99%) or with [U-13C]-glutamine (4 mmol/L in media; Cambridge Isotope Laboratories; CLM-1822; l-glutamine [U-13C], 99%)). Tracing to CAC intermediates with [U-13C]-glucose or [U-13C]-glutamine were performed at 15, 30, 60, and 120 minutes, and DHAP, lactate, and pyruvate isotopomer analyses were performed with [U-13C] glucose for 30, 60, 90, and 300 seconds to ensure measurements were in the dynamic range of labeling (30). After washing twice in cold saline solution (NaCl, 0.9 g/L), cells (at 80% confluence) were quenched with 600 μL methanol 80% (v/v) on dry ice. The homogenates were then sonicated in ice water slurry for 10 minutes with the cycling 30 seconds on/off at the highest setting (bath sonicator Biorupter) and centrifuged at 14,000 × g (4°C) for 10 minutes. The supernatants, supplemented with internal standard (750 ng myristic acid-D27; Sigma-Aldrich), were dried overnight in cold vacuum centrifuge (Labconco).

GC/MS

The dried samples were dissolved in 30 μL of methoxyamine hydrochloride (10 mg/mL diluted in pyridine; Sigma-Aldrich) with sonication (for 30 seconds) and vortex (for 30 seconds). After incubation for 30 minutes at room temperature, the samples were derivatized with 70 μL of N-tert-butyldimethylsilyl-N-methyltrifluoroacetamide (MTBSTFA, > 97%; Sigma-Aldrich), transferred in sealed glass vials, and then incubated for 1 hour at 70°C. GC/MS analysis was performed by injection of 1 μL of derivatized sample into GC/MS instrument (5975C; Agilent). Data acquisition was done in both Scan and SIM (Selected Ion Monitoring) modes as described in refs. 31 and 32. The isotopic distribution of integrated ions was determined using Agilent MassHunter software. The steady-state level was determined for each metabolite and normalized by cell number and internal standard (myristic acid-D27) intensity. Isotopic distributions were corrected for naturally occurring isotopes to isolate contributions from our tracers using an in-house algorithm according to refs. 32 and 33. Enrichments for specific isotopomers (e.g. m+3), termed proportional ion amount, were presented as proportions of the total isotopic distribution for each metabolite. Total abundance of specific isotopomers, termed relative ion amount, was assessed for each metabolite by multiplying proportional ion amounts by the normalized steady-state levels.

NMR sample preparation

Tumors were flash frozen directly after dissection, ground to a fine powder using a mortar and pestle on liquid nitrogen, weighed (15–59 mg) and transferred to 2 mL Eppendorf microcentrifuge tubes. Six prewashed and chilled 1.4 mm ceramic beads were added to each sample along with 1.5 mL of cold 80% methanol in water (both LC/MS grade solvents). The samples were subjected to 2 minutes of bead beating (TissueLyser II, Qiagen). Precipitated proteins and tissue debris were removed by centrifugation at 15 krpm and 1°C for 10 minutes. Supernatants were transferred to chilled tubes and dried by vacuum centrifugation (Labconco) with sample temperature maintained at −4°C. Samples were subsequently resuspended in 230 μL water centrifuged at 15 krpm for 10 minutes at 1°C. A volume of 144 μL of tumor extract along with 16 μL NMR sample buffer were transferred to 3 mm NMR tubes using a Gilson Liquid handler 215 sample preparation robot. The final buffer concentration was 150 mmol/L KH2PO4, 0.5 mmol/L 3-(trimethylsilyl) propionic-2,2,3,3-D4 acid, and 0.2 mmol/L NaN3 at pH 7.4 in 10% D2O.

NMR data collection and analysis

NMR data were collected at the Drug Discovery Platform (McGill University Health Centre Research Institute) utilizing a Bruker Avance III HD 600MHz NMR spectrometer equipped with a CPQCI 1H-31P/13C/15N cryogenically cooled probe and Sample Jet autosampler (Bruker Biospin Ltd). NMR spectra of samples were collected using the first increment of the NOESY pulse sequence supplied with the instrument. Metabolite profiling and quantitation was achieved using Chenomx NMR Suite Professional (v8.2) where 0.5 mmol/L TSP-D4 was used as the internal concentration standard (https://pubs.acs.org/doi/abs/10.1021/ac060209g). The amount of metabolites was normalized to the weight of the extracted tumor.

Statistical analysis

Unless otherwise indicated, all in vitro studies were carried out with three biological replicates with four technical replicates per experimental group. Data were normalized to the standard curve or the control group as appropriate. For the in vivo studies, power analysis showed a sample size of 10 tumors per group provided 80% power to detect a mean 1.5-fold difference in tumor volume with a significance level of 0.05 (two-tailed) between two groups. The following statistical analyses were used throughout this study: Two-tailed, paired Student t test (Fig. 1C and D; Fig. 2A and B; Fig. 5A); two-tailed, unpaired Student t test (Fig. 1B and E–H; Fig. 2C–E; Fig. 3A and B; Supplementary Fig. S6); two-tailed, unpaired Student t test Holm–Sidak method (Fig. 6D–F); one-way ANOVA with a Tukey multiple comparisons test (Fig. 3C–E, Fig. 5F–I; Supplementary Fig. S1B-C, S2A); two-way ANOVA with a Tukey multiple comparisons test (Fig. 4B and D; Fig. 5B–E; Fig. 6C; Supplementary Figs. S2B–S2D, S5B–S5C, and S7). Statistical significance is shown as follows: *, P = 0.05; **, P < 0.01; ***, P = 0.001; ****, P < 0.0001.

Figure 1.

ShcA signaling increases the metabolic activity and glucose dependency of ErbB2-driven breast cancer cells. A, Schematic diagram depicting the ShcA alleles (ShcAWT, Shc3F) used in this study. B, ShcAWT and Shc3F tumor growth rate is shown as average tumor volume (mm3) ± SEM (n = 8 tumors each). C, The glycolytic and respiratory activity of ShcAWT and Shc3F cells was assessed by quantifying their ECARs and OCR, respectively (means ± SEM; n = 9 per group). D, Immunoblot analysis from ShcAWT and Shc3F cells using pAMPK-, AMPK-, and tubulin-specific antibodies. The data are representative of four independent experiments (± SEM). E, Mammary tumors were subjected to immunohistochemical staining using pAMPK-specific antibodies (% positively stained cells ± SEM; n = 8 tumors each). Representative images are shown (scale bar, 50 μm). F, AMP levels were quantified in ShcAWT (n = 10) and Shc3F (n = 11) tumors by NMR analysis (moles AMP/mg tumor tissue ± SEM). G, Cells were cultured under variable glucose concentrations and the number of viable cells was quantified by trypan blue exclusion. The data show one representative experiment (n = 4) ± SD, which was performed in duplicate. H, Glucose and lactate levels were quantified in ShcAWT (n = 10) and Shc3F (n = 11) mammary tumors by NMR analysis (moles of each metabolite/mg tumor tissue ± SEM). The lactate/glucose ratio was also calculated. *, P = 0.05; **, P < 0.01; ***, P = 0.001; ****, P < 0.0001.

Figure 1.

ShcA signaling increases the metabolic activity and glucose dependency of ErbB2-driven breast cancer cells. A, Schematic diagram depicting the ShcA alleles (ShcAWT, Shc3F) used in this study. B, ShcAWT and Shc3F tumor growth rate is shown as average tumor volume (mm3) ± SEM (n = 8 tumors each). C, The glycolytic and respiratory activity of ShcAWT and Shc3F cells was assessed by quantifying their ECARs and OCR, respectively (means ± SEM; n = 9 per group). D, Immunoblot analysis from ShcAWT and Shc3F cells using pAMPK-, AMPK-, and tubulin-specific antibodies. The data are representative of four independent experiments (± SEM). E, Mammary tumors were subjected to immunohistochemical staining using pAMPK-specific antibodies (% positively stained cells ± SEM; n = 8 tumors each). Representative images are shown (scale bar, 50 μm). F, AMP levels were quantified in ShcAWT (n = 10) and Shc3F (n = 11) tumors by NMR analysis (moles AMP/mg tumor tissue ± SEM). G, Cells were cultured under variable glucose concentrations and the number of viable cells was quantified by trypan blue exclusion. The data show one representative experiment (n = 4) ± SD, which was performed in duplicate. H, Glucose and lactate levels were quantified in ShcAWT (n = 10) and Shc3F (n = 11) mammary tumors by NMR analysis (moles of each metabolite/mg tumor tissue ± SEM). The lactate/glucose ratio was also calculated. *, P = 0.05; **, P < 0.01; ***, P = 0.001; ****, P < 0.0001.

Close modal
Figure 2.

Impaired ShcA signaling increases the glutamine dependency of breast cancer cells. A, Glycolytic and CAC metabolite levels in ShcAWT and Shc3F cells cultured in high (25 mmol/L) or no (0 mmol/L) glucose for 3 days. The data are shown as fold change in metabolite levels following glucose withdrawal compared with 25 mmol/L glucose conditions ± SEM and is representative of three independent experiments. The raw data can be found in Supplementary Fig. S2A. B, The fold change in steady-state glutamine and glutamate levels, both under high glucose conditions and 48 hours following glucose withdrawal ± SEM. The data are representative of three independent experiments. C, ShcAWT and Shc3F cells were cultured in the presence or absence of glucose or glutamine for 3 days. The number of viable cells was quantified by trypan blue exclusion (n = 8). The data are representative of two independent experiments and are shown as fold change in viable cells relative to ShcAWT control cells ± SD. D, Glutamine and glutamate levels were quantified in ShcAWT (n = 10) and Shc3F (n = 11) mammary tumors by NMR analysis (moles of each metabolite/mg tumor tissue ± SEM). The glutamate/glutamine ratio was also calculated. *, P = 0.05; **, P < 0.01; ***, P = 0.001; ****, P < 0.0001.

Figure 2.

Impaired ShcA signaling increases the glutamine dependency of breast cancer cells. A, Glycolytic and CAC metabolite levels in ShcAWT and Shc3F cells cultured in high (25 mmol/L) or no (0 mmol/L) glucose for 3 days. The data are shown as fold change in metabolite levels following glucose withdrawal compared with 25 mmol/L glucose conditions ± SEM and is representative of three independent experiments. The raw data can be found in Supplementary Fig. S2A. B, The fold change in steady-state glutamine and glutamate levels, both under high glucose conditions and 48 hours following glucose withdrawal ± SEM. The data are representative of three independent experiments. C, ShcAWT and Shc3F cells were cultured in the presence or absence of glucose or glutamine for 3 days. The number of viable cells was quantified by trypan blue exclusion (n = 8). The data are representative of two independent experiments and are shown as fold change in viable cells relative to ShcAWT control cells ± SD. D, Glutamine and glutamate levels were quantified in ShcAWT (n = 10) and Shc3F (n = 11) mammary tumors by NMR analysis (moles of each metabolite/mg tumor tissue ± SEM). The glutamate/glutamine ratio was also calculated. *, P = 0.05; **, P < 0.01; ***, P = 0.001; ****, P < 0.0001.

Close modal
Figure 3.

Reduced ShcA signaling sensitizes breast tumors to phenformin. A, ShcAWT and Shc3F cells were cultured in the absence (PBS) or presence of phenformin (0.5 mmol/L) for 3 days. The percentage of viable cells was quantified by trypan blue exclusion (n = 9 each). The data are representative of three independent experiments and is shown as percentage of viable cells relative to ShcAWT control (± SD). B, Mammary fat pad injection of ShcAWT and Shc3F cells into immunodeficient mice. When mammary tumors reached 100–150 mm3, animals were treated daily with phenformin (50 mg/kg intraperitoneally) or PBS as control. Average fold increase in tumor volume relative to day 0 of treatment ± SEM (n = 10 tumors: PBS control groups; n = 12 tumors: phenformin-treated groups). C–E, Mammary tumors were subjected to IHC staining using pAMPK (C), Ki67 (D), and cleaved caspase-3–specific (E) antibodies. The data are shown as percentage of positively stained cells ± SEM (n = 8 tumors). Bottom, representative images (scale bar, 50 μm). *, P = 0.05; **, P < 0.01; ***, P = 0.001.

Figure 3.

Reduced ShcA signaling sensitizes breast tumors to phenformin. A, ShcAWT and Shc3F cells were cultured in the absence (PBS) or presence of phenformin (0.5 mmol/L) for 3 days. The percentage of viable cells was quantified by trypan blue exclusion (n = 9 each). The data are representative of three independent experiments and is shown as percentage of viable cells relative to ShcAWT control (± SD). B, Mammary fat pad injection of ShcAWT and Shc3F cells into immunodeficient mice. When mammary tumors reached 100–150 mm3, animals were treated daily with phenformin (50 mg/kg intraperitoneally) or PBS as control. Average fold increase in tumor volume relative to day 0 of treatment ± SEM (n = 10 tumors: PBS control groups; n = 12 tumors: phenformin-treated groups). C–E, Mammary tumors were subjected to IHC staining using pAMPK (C), Ki67 (D), and cleaved caspase-3–specific (E) antibodies. The data are shown as percentage of positively stained cells ± SEM (n = 8 tumors). Bottom, representative images (scale bar, 50 μm). *, P = 0.05; **, P < 0.01; ***, P = 0.001.

Close modal
Figure 4.

ShcA signaling increases glucose metabolism. A, Schematic representation of glycolysis and CAC, with forward (green; m+2) and reverse (blue; m+3) fluxes, using [U-13C]-glucose as the metabolite tracer [represented as black, green (m+2), and blue (m+3) dots]. Labeled intermediate metabolites were analyzed by GC/MS. White dots, endogenous 12C. B, ShcAWT and Shc3F cells were cultured for 24 hours with 0.5 mmol/L phenformin (Phen) or PBS controls, then incubated with [U-13C]-labeled glucose (25 mmol/L) for 30 seconds to measure incorporation into intracellular glycolytic intermediates (mean metabolite levels ± SEM; n = 3 for each group). C, Schematic representation of steady-state metabolite pools in ShcAWT and Shc3F cells cultured for 24 hours with 0.5 mmol/L phenformin (Phen) or PBS controls (graphed data are shown in Supplementary Fig. S2B). D, Cells were also incubated with [U-13C]-glucose for 15 minutes to label intermediates of CAC. Bar graphs indicate the relative ion amount per cell, expressed as mean metabolite level for CAC metabolites (± SEM; n = 5 for each group). Green bar, incorporation of glucose into the indicated metabolites by forward flow through the CAC (m+2); blue bar, incorporation of glucose into the indicated metabolites by reverse flow through the CAC (m+3). *, P = 0.05; **, P < 0.01; ***, P = 0.001.

Figure 4.

ShcA signaling increases glucose metabolism. A, Schematic representation of glycolysis and CAC, with forward (green; m+2) and reverse (blue; m+3) fluxes, using [U-13C]-glucose as the metabolite tracer [represented as black, green (m+2), and blue (m+3) dots]. Labeled intermediate metabolites were analyzed by GC/MS. White dots, endogenous 12C. B, ShcAWT and Shc3F cells were cultured for 24 hours with 0.5 mmol/L phenformin (Phen) or PBS controls, then incubated with [U-13C]-labeled glucose (25 mmol/L) for 30 seconds to measure incorporation into intracellular glycolytic intermediates (mean metabolite levels ± SEM; n = 3 for each group). C, Schematic representation of steady-state metabolite pools in ShcAWT and Shc3F cells cultured for 24 hours with 0.5 mmol/L phenformin (Phen) or PBS controls (graphed data are shown in Supplementary Fig. S2B). D, Cells were also incubated with [U-13C]-glucose for 15 minutes to label intermediates of CAC. Bar graphs indicate the relative ion amount per cell, expressed as mean metabolite level for CAC metabolites (± SEM; n = 5 for each group). Green bar, incorporation of glucose into the indicated metabolites by forward flow through the CAC (m+2); blue bar, incorporation of glucose into the indicated metabolites by reverse flow through the CAC (m+3). *, P = 0.05; **, P < 0.01; ***, P = 0.001.

Close modal
Figure 5.

ShcA signaling renders breast cancer cells dependent on PGC-1α. A, Relative Ppargc1a/Actb mRNA levels were determined by qRT-PCR analysis. The data are representative of two independent experiments, with four replicates per experiment ± SD. B, ShcAWT and Shc3F breast cancer cells were retrovirally infected with PGC-1α shRNAs or empty vector controls and cultured in the absence of glucose or in the presence of 0.5 mmol/L phenformin for 48 hours. The glutamine deprivation experiments were performed after 24 hours. Cell viability was quantified by trypan blue exclusion. The data are shown as percentage of viable cells relative to nutrient replete conditions (n = 4) ± SD for one representative experiment performed in duplicate. C, Two independent MT/ShcA+/+ (864; 4788) and MT/ShcA+/3F (6021; 6360) breast cancer cell lines expressing control or PGC-1α shRNAs were tested for their viability as described in B. D, The efficiency of generating PGC-1α-null cells using two CRISPR guides is shown. E, The indicated cell lines were cultured in high (25 mmol/L) or no (0 mmol/L) glucose for 24 hours and then subjected to qRT-PCR analysis. The relative Ppargc1a/Actb. Ppargc1b/Actb, and Esrra/Actb ratios were determined. The data are representative of four independent experiments (means ± SEM). *, ShcAWT vs. Shc3F (EV); ϵ, high vs. no glucose conditions; δ, Shc3F (EV) vs. PGC1α-O/E, CR#1 and CR#2. F–I, The indicated cell lines were cultured for 3 days under nutrient replete conditions (F) or in the absence of glucose (G) or glutamine (H) or in the presence of 0.5 mmol/L phenformin (I). Viable cells were quantified by trypan blue exclusion. The data are shown as # viable cells over three days (n = 4) ± SD for one representative experiment performed in duplicate. *, P = 0.05; **, P < 0.01; ***, P = 0.001; ****, P < 0.0001.

Figure 5.

ShcA signaling renders breast cancer cells dependent on PGC-1α. A, Relative Ppargc1a/Actb mRNA levels were determined by qRT-PCR analysis. The data are representative of two independent experiments, with four replicates per experiment ± SD. B, ShcAWT and Shc3F breast cancer cells were retrovirally infected with PGC-1α shRNAs or empty vector controls and cultured in the absence of glucose or in the presence of 0.5 mmol/L phenformin for 48 hours. The glutamine deprivation experiments were performed after 24 hours. Cell viability was quantified by trypan blue exclusion. The data are shown as percentage of viable cells relative to nutrient replete conditions (n = 4) ± SD for one representative experiment performed in duplicate. C, Two independent MT/ShcA+/+ (864; 4788) and MT/ShcA+/3F (6021; 6360) breast cancer cell lines expressing control or PGC-1α shRNAs were tested for their viability as described in B. D, The efficiency of generating PGC-1α-null cells using two CRISPR guides is shown. E, The indicated cell lines were cultured in high (25 mmol/L) or no (0 mmol/L) glucose for 24 hours and then subjected to qRT-PCR analysis. The relative Ppargc1a/Actb. Ppargc1b/Actb, and Esrra/Actb ratios were determined. The data are representative of four independent experiments (means ± SEM). *, ShcAWT vs. Shc3F (EV); ϵ, high vs. no glucose conditions; δ, Shc3F (EV) vs. PGC1α-O/E, CR#1 and CR#2. F–I, The indicated cell lines were cultured for 3 days under nutrient replete conditions (F) or in the absence of glucose (G) or glutamine (H) or in the presence of 0.5 mmol/L phenformin (I). Viable cells were quantified by trypan blue exclusion. The data are shown as # viable cells over three days (n = 4) ± SD for one representative experiment performed in duplicate. *, P = 0.05; **, P < 0.01; ***, P = 0.001; ****, P < 0.0001.

Close modal
Figure 6.

A functional ShcA/PGC1α signaling axis is required during the earliest stages of breast cancer development. A, Mammary fat pad injection of the indicated cells into SCID-beige mice (n = 10 tumors each). Kaplan–Meier analysis of the percent tumor-free mice over time. B, Average tumor volume (mm3) ± SEM following first palpation. C, The indicated cell lines were cultured in high (25 mmol/L) or no (0 mmol/L) glucose-containing media for 24 hours and then subjected to qRT-PCR analysis. The relative SOD2/Actb and GLS1/Actb ratios were determined. The data are representative of four independent experiments (means ± SEM) *, ShcAWT vs Shc3F (EV) cells; ϵ, high vs. no glucose conditions; δ, Shc3F (EV) vs. PGC1α-O/E, CR#1, or CR#2. D–F, MMTV/ShcA+/+ and MMTV/ShcA3F/+ transgenic mice were monitored for tumor onset by biweekly physical palpation and following detection of one tumor-bearing gland, were either given 200 mg/kg phenformin in their drinking water or received water alone for 4–5 weeks. Tumor-bearing days: MT/ShcA+/+ (control) = 39.1 ± 1.6 (n = 14); MT/ShcA+/+ (Phen) = 39.3 ± 2.5 (n = 6); MT/ShcA3F/+ (control) = 44.7 ± 6.2 (n = 10); MT/ShcA3F/+ (Phen) = 43.6 ± 11.5 (n = 5). D, The number of tumor-bearing glands at endpoint. The average (E) and total tumor volumes (mm3; F) ± SEM at endpoint. *, P = 0.05; **, P < 0.01; ***, P = 0.001.

Figure 6.

A functional ShcA/PGC1α signaling axis is required during the earliest stages of breast cancer development. A, Mammary fat pad injection of the indicated cells into SCID-beige mice (n = 10 tumors each). Kaplan–Meier analysis of the percent tumor-free mice over time. B, Average tumor volume (mm3) ± SEM following first palpation. C, The indicated cell lines were cultured in high (25 mmol/L) or no (0 mmol/L) glucose-containing media for 24 hours and then subjected to qRT-PCR analysis. The relative SOD2/Actb and GLS1/Actb ratios were determined. The data are representative of four independent experiments (means ± SEM) *, ShcAWT vs Shc3F (EV) cells; ϵ, high vs. no glucose conditions; δ, Shc3F (EV) vs. PGC1α-O/E, CR#1, or CR#2. D–F, MMTV/ShcA+/+ and MMTV/ShcA3F/+ transgenic mice were monitored for tumor onset by biweekly physical palpation and following detection of one tumor-bearing gland, were either given 200 mg/kg phenformin in their drinking water or received water alone for 4–5 weeks. Tumor-bearing days: MT/ShcA+/+ (control) = 39.1 ± 1.6 (n = 14); MT/ShcA+/+ (Phen) = 39.3 ± 2.5 (n = 6); MT/ShcA3F/+ (control) = 44.7 ± 6.2 (n = 10); MT/ShcA3F/+ (Phen) = 43.6 ± 11.5 (n = 5). D, The number of tumor-bearing glands at endpoint. The average (E) and total tumor volumes (mm3; F) ± SEM at endpoint. *, P = 0.05; **, P < 0.01; ***, P = 0.001.

Close modal

The ShcA adaptor increases the metabolic rate and glucose dependency of breast tumors

Because the ShcA adaptor is a critical signaling hub downstream of RTKs for mammary tumor initiation, growth, and angiogenesis (23, 34), we sought to establish whether it also plays a role in metabolic reprogramming. To test this, we modulated ShcA signaling in ErbB2-transformed NMuMG breast cancer cells by ectopically expressing FLAG-tagged ShcA-proficient (ShcAWT) or -deficient (Y239/240/313F; Shc3F) alleles. The Shc3F allele serves as a dominant-negative mutant (Fig. 1A) (25). We show comparable ShcAWT-FLAG and Shc3F-FLAG expression levels in both cell lines (Supplementary Fig. S1A). Moreover, expression of the Shc3F mutant severely impairs mammary tumor growth in an orthotopic mouse model relative to ShcAWT controls (Fig. 1B). We previously showed that impaired ShcA signaling reduces the proliferative and angiogenic capacities of these tumors in vivo (23, 24). We now show that Shc3F-expressing cells are less metabolically active, as evidenced by 20% and 10% decreased rates of glycolysis (ECAR) and mitochondrial respiration (OCR) compared with ShcAWT cells (Fig. 1C). Nonetheless, AMPK Thr-172 phosphorylation was reduced in Shc3F-expressing cells (Fig. 1D) and mammary tumors (Fig. 1E) compared with their ShcAWT counterparts. This is coincident with reduced AMP levels in Shc3F mammary tumors (Fig. 1F). Taken together, these data suggest that RTKs engage the ShcA pathway to promote metabolism.

With the knowledge that many breast tumors acquire a Warburg-like metabolism, we next assessed whether modulating the ShcA pathway in breast cancer cells altered their glucose dependency. While inhibition of ShcA signaling does not impact breast cancer cell growth in glucose replete conditions (25 mmol/L), Shc3F cells acquire a growth advantage when glucose levels are reduced (5 mmol/L) or absent (0 mmol/L; Fig. 1G). In contrast, ectopic expression of wild-type ShcA sensitizes ErbB2-NMuMG cells to glucose deprivation to levels that are comparable with parental cells (Supplementary Fig. S1B). To further support our findings, we tested the glucose dependency of MT-transformed breast cancer cells that originated from transgenic mice (MT/ShcA+/+ vs. MT/ShcA3F/+; ref. 25) expressing the Shc3F allele under the control of the endogenous ShcA promoter. We show that impaired ShcA signaling increases cell viability upon glucose withdrawal in cells expressing physiological levels of the Shc3F mutant (Supplementary Fig. S1C). These data suggest that reduced ShcA signaling endows breast cancer cells with the potential to grow under glucose-deprived conditions.

We next assessed whether reduced ShcA signaling is associated with impaired glucose metabolism in mammary tumors. We show that inhibition of ShcA signaling (Shc3F) significantly increases glucose levels in mammary tumors relative to ShcAWT controls (Fig. 1H). Despite this fact, Shc3F tumors possess reduced lactate levels and a significant reduction in the lactate/glucose ratio compared with ShcAWT tumors (Fig. 1H), coincident with their reduced growth potential (Fig. 1B). This suggests that attenuated ShcA signaling decreases the ability of mammary tumors to perform glycolysis, resulting in a reduced bioenergetic capacity.

To better understand the impact of glucose deprivation on metabolic adaptations in ShcAWT and Shc3F cells, we measured glycolytic and CAC metabolite levels in each cell line cultured in the presence or absence of glucose. Compared with ShcAWT cells, steady-state levels of pyruvate along with several CAC metabolites (citrate, α-ketoglutarate, fumarate, malate) are less depleted in Shc3F breast cancer cells upon glucose withdrawal (Fig. 2A, Fig. S2A). This suggests that decreased ShcA signaling reduces the metabolic activity of breast cancers, allowing them to withstand glucose deprivation.

Impaired ShcA signaling increases the glutamine dependency of breast cancer cells

Given its importance as an alternative energy source in a subset of cancers (8), we investigated whether impaired ShcA signaling instead increases the glutamine dependency of breast cancer cells. We first measured how ShcAWT and Shc3F cancer cells modulate glutamine and glutamate levels following glucose withdrawal. Increased glutamate production may indicate enhanced glutamine metabolism (8). Although glutamine levels are unaffected, Shc3F cells display a robust increase in glutamate levels following glucose deprivation, suggesting that they efficiently metabolize glutamine in the absence of glucose (Fig. 2B). We next compared the relative viability of ShcAWT and Shc3F cells in response to glucose versus glutamine deprivation. Whereas Shc3F cells better withstand glucose withdrawal, they are more sensitive to glutamine deprivation relative to ShcAWT controls (Fig. 2C). These data suggest that inhibition of ShcA signaling increases the glutamine dependency of breast cancer cells.

Considering these observations, we measured glutamine and glutamate levels in ShcAWT and Shc3F mammary tumors. Compared with ShcAWT controls, tumors with reduced ShcA signaling display increased glutamine levels, coincident with decreased glutamate levels (Fig. 2D). The increased lactate/glucose (Fig. 1H) and glutamate/glutamine (Fig. 2D) ratios in ShcAWT-expressing tumors suggest that they increase their bioenergetic capacity through enhanced glucose and glutamine metabolism yet are exquisitely reliant on glucose availability. Even though Shc3F cells reduce their bioenergetic capacity and are less dependent on glucose, they instead reprogram their metabolism to rely on glutamine.

Impaired ShcA signaling sensitizes breast cancers to biguanides

Biguanides (metformin and phenformin) block complex I of the electron transport chain (12–14) and render cells dependent on glycolysis (10, 12, 15). We next asked whether inhibition of ShcA signaling in breast cancer cells exposes a metabolic vulnerability to biguanides. In agreement with their reduced glycolytic capacity and enhanced dependence on oxidative phosphorylation, Shc3F cells are more sensitive to phenformin compared with ShcAWT cells in vitro (Fig. 3A). We next injected ShcAWT- and Shc3F-expressing cells into the mammary fat pad of immunodeficient mice to determine whether ShcA signaling impacted the sensitivity of mammary tumors to phenformin in vivo. ShcAWT tumor growth is minimally affected by phenformin treatment (Fig. 3B), even though we observe increased pAMPK immunohistochemical staining (Fig. 3C). Phenformin did not impact the proliferative capacity of ShcAWT tumors (Fig. 3D) but increased their apoptotic response (Fig. 3E). This is consistent with the inability of ShcAWT cells to reduce their metabolic rate under nutrient stress (Figs. 1G and 2A). The relative insensitivity of ShcAWT tumors to phenformin in vivo (Fig. 3B) can be explained by their enhanced bioenergetic capacity and metabolic flexibility to increase the rate of glycolysis.

In contrast, the growth potential of Shc3F tumors was significantly blunted (40% reduction in tumor volume) following phenformin treatment in vivo (Fig. 3B), in agreement with their greater dependency on mitochondrial metabolism. Phenformin increased pAMPK levels in Shc3F tumors and significantly impaired their proliferative capacity but did not alter their apoptotic rate (Fig. 3C–E). This suggests that reduced ShcA signaling sensitizes mammary tumors to phenformin through induction of a cytostatic response.

ShcA signaling increases glucose metabolism to confer biguanide resistance

To gain a mechanistic understanding of how ShcA signaling impacts the metabolic network, we performed stable isotope tracing experiments to follow the fate of [U-13C]-glucose in ShcAWT and Shc3F cells (Fig. 4A). These studies allowed us to determine whether ShcA signaling diverts glucose through glycolysis or the CAC for ATP production. ShcAWT and Shc3F cells displayed similar kinetics of [U-13C]-labeling for glycolytic intermediates (Fig. 4B). Accounting for the larger steady-state pool of each metabolite in ShcAWT cells (Fig. 4C; Supplementary Fig. S2B), these data support the fact that ShcA increases the glycolytic capacity of breast cancer cells (Fig. 1B), resulting in a higher quantitative glycolytic flux. Despite this fact, the lactate/pyruvate ratio is similar between ShcAWT and Shc3F cells (Supplementary Fig. S2C). These data suggest that ShcA signaling globally increases glucose metabolism both through glycolysis and the CAC. Indeed, we show that the ability of (m+3) pyruvate to enter the CAC, either to generate ATP (green bars) or for macromolecular synthesis (blue bars) is comparable between ShcAWT and Shc3F-expressing cells (Fig. 4D; Supplementary Fig. S3). Again, considering the larger pool of CAC metabolites in ShcAWT-expressing cells (Fig. 4C; Supplementary Fig. S2A), these data reinforce the fact that glucose flow through the CAC is globally increased by elevated ShcA signaling.

As expected, phenformin significantly increased glycolysis in both ShcAWT and Shc3F cells (Fig. 4B; Supplementary Fig. S3). Coupled with the drastic increase in the steady-state pools of glycolytic metabolites (Fig. 4C; Supplementary Fig. S2A), these data are consistent with studies reporting a strict reliance on aerobic glycolysis in the presence of biguanides (10, 12, 15). While phenformin increased the lactate/pyruvate ratio in ShcAWT and Shc3F-expressing cells, the induction was stronger in ShcAWT cells (Supplementary Fig. S2C). These data indicate that ShcA signaling increases the glycolytic capacity of breast cancer cells, rendering them better able to cope with biguanide treatment (Fig. 3A). Moreover, phenformin drastically reduced the forward flow of glucose-derived metabolites into the CAC in both ShcAWT and Shc3F-expressing cells (Fig. 4D; Supplementary Fig. S3). In contrast, phenformin promoted anaplerosis to replenish CAC intermediates. Combined, these data suggest that phenformin blocks mitochondrial glucose metabolism independently of the ShcA pathway.

Given that inhibition of ShcA signaling in breast cancer cells increases their glutamine reliance (Fig. 2C), we also examined the fate of [U-13C]-glutamine through the CAC. Labeled glutamine (m+5) enters the CAC through the stepwise conversion to glutamate (m+5) and α-ketoglutarate (m+5). α-Ketoglutarate-derived carbons either flow through the canonical CAC (m+4; illustrated in green) or reductive carboxylation (m+5 citrate; m+3 oxaloacetate, m+3 malate; m+3 fumarate; depicted in red) (Supplementary Fig. S4). Reductive carboxylation of glutamine occurs mainly in cancer cells with defective mitochondria or under reductive conditions, like hypoxia and exposure to biguanides (35–37). There was no significant difference in the metabolism of glutamine through the canonical CAC between ShcAWT and Shc3F-expressing cells (Supplementary Figs. S4; S5A and S5B). Phenformin treatment significantly decreased the flow of glutamine-derived carbons through the canonical CAC in both ShcAWT and Shc3F cells. However, phenformin promoted reductive carboxylation of glutamine in both cell types (Supplementary Fig. S4; S5A and S5C). These results indicate that perturbation of ShcA signaling does not affect how glutamine is metabolized through the canonical CAC. Although phenformin-treated Shc3F cells displayed an increase in the α-ketoglutarate/citrate ratio (Supplementary Fig. S2D), enhanced reductive carboxylation does not protect them from the antitumorigenic effects of this biguanide. Instead, the reduced capacity of Shc3F-expressing cells to engage glycolysis in the presence of phenformin likely explains their increased sensitivity to biguanides, compared with ShcAWT controls.

PGC-1α is essential for ErbB2-driven breast tumor initiation and resistance to biguanides

The PGC-1α transcriptional coactivator is a master regulator of energy metabolism (38) and is induced upon nutrient stress (39). Strikingly, we observed that ShcA signaling increases PGC-1α expression. Indeed, PGC-1α levels are 3.9-fold higher in ShcAWT cells compared with Shc3F cells (Fig. 5A). To interrogate whether ShcA signaling increases the PGC-1α dependency of breast cancer cells, we used shRNA approaches to stably reduce Ppargc1a mRNA levels in cells that differ in their ability to engage the ShcA pathway. This includes ErbB2/ShcAWT and ErbB2/Shc3F cells in addition to two independent explants from MT/ShcA+/+ and MT/ShcA+/3F mammary tumors (Supplementary Fig. S6). We show that breast cancer cells with an intact ShcA pathway are more reliant on PGC-1α to support their growth, even under nutrient replete conditions (Fig. 5B and C). Moreover, the ability of ShcA signaling to cope with phenformin requires PGC-1α (Fig. 5B and C). These data support an essential role for PGC-1α in increasing glucose metabolism (40). Finally, PGC-1α is also required to permit Shc3F cells to withstand glucose withdrawal, further reinforcing its essential role in mitochondrial metabolism (Fig. 5B and C).

To further substantiate an essential role for PGC-1α in ShcA-driven breast cancer progression, we ectopically expressed PGC-1α in Shc3F cells or stably deleted PGC-1α from ShcAWT and Shc3F cells by Crispr/Cas9 gene editing (Fig. 5D and E). We used two independent guides directed to the second exon of PGC-1α. PGC-1α-null cells encode the first 57 (Crispr #1) or 37 (Crispr #2) amino acids of PGC-1α prior to introduction of a premature stop codon. We show that Shc3F cells better tolerated PGC-1α loss, whereas PGC-1α was key for ShcAWT cell viability (Fig. 5D), confirming the fact that ShcA signaling increases the PGC-1α dependency of breast cancer cells (Fig. 5B and C). This suggests that the PGC-1α-null, ShcAWT clones likely significantly rewired their metabolism. Moreover, reduced ShcA signaling permits cancer cells to adapt to PGC-1α loss under nutrient replete conditions. Hence, we focused our attention on understanding the effects of PGC-1α loss on the tumorigenic potential of Shc3F cells.

We first examined how nutrient deprivation modulated PGC-1α levels in breast cancer cells. Although endogenous Ppargc1a (PGC-1α) levels are basally low in Shc3F cells, Ppargc1a is strongly induced in these cells (40-fold) upon glucose deprivation. In contrast, PGC-1α overexpressing cells modestly upregulated Ppargc1a levels (2-fold) in the absence of glucose (Fig. 5E). Moreover, Ppargc1a mRNA levels were not appreciably induced in PGC-1α-null cells, supporting the fact that PGC-1α transcriptionally controls its own promoter (Fig. 5E; ref. 41). We also examined how PGC-1β and ERRα levels are controlled by alterations in ShcA signaling and/or PGC-1α expression. PGC-1β is a closely related family member (42) and ERRα is an orphan nuclear receptor that preferentially dimerizes with PGC-1α to coordinate expression of metabolic genes (43). Ppargc1b is modestly repressed in Shc3F cells (1.4-fold), whereas Esrra expression is unaffected by the ShcA pathway in high glucose conditions (Fig. 5E). Glucose deprivation upregulates Ppargc1b and Esrra expression in both cell lines. However, Shc3F cells show a greater Ppargc1b and Esrra induction (Fig. 5E). PGC-1α overexpression has little to no impact on endogenous Ppargc1b and Esrra levels, in glucose-depleted conditions (Fig. 5E), given that PGC-1α levels are already significantly induced (Fig. 5A). In contrast, PGC-1α loss significantly impaired the ability of breast cancer cells to upregulate Ppargc1b (∼2.5-fold) and Esrra (∼2 fold) following glucose deprivation (Fig. 5E).

Next, we tested the impact of PGC-1α overexpression or loss on the growth potential of Shc3F cells. Under control conditions, both PGC-1α overexpression and loss decreased the growth potential of Shc3F cells relative to empty vector controls, suggesting that PGC-1α levels must be tightly controlled to maintain metabolic functions (Fig. 5F). Previous studies suggest that cells have a narrow window of tolerance for PGC-1α levels (38, 44, 45). While PGC-1α overexpression increased the viability of breast cancer cells in response to glucose and glutamine withdrawal, PGC-1α loss further sensitized them to nutrient deprivation (Fig. 5G and H). Similarly, PGC-1α overexpression protected breast cancer cells to phenformin while PGC-1α loss further sensitized them to this biguanide (Fig. 5I). These data suggest that PGC-1α overexpression endows breast cancer cells with increased metabolic capacities, while PGC-1α loss may be deleterious for cancer cell growth, which requires significant metabolic investment. To test this in vivo, we injected ShcAWT, Shc3F (EV), PGC-1α overexpressing, PGC-1α Crispr #1, and PGC-1α Crispr #2 cells into the mammary fat pads of immunodeficient mice. As expected, Shc3F cells display delayed tumor onset (Fig. 6A) and impaired tumor growth (Fig. 6B) relative to ShcAWT cells. PGC-1α overexpression modestly increased tumor onset and growth, suggesting that PGC-1α is insufficient to restore the tumorigenic potential of breast cancers that are debilitated in ShcA signaling. This is consistent with the fact that PGC-1α overexpression did not rescue the proliferative, apoptotic, and angiogenic defects of Shc3F tumors (Supplementary Fig. S7).

We also tested whether PGC-1α is necessary for breast cancer progression. Markedly, PGC-1α loss profoundly delayed breast tumor onset (EV: T50 = 16 days; CR1: T50 = 48 days; CR2: T50 = 26 days; Fig. 6A). As expected, Sod2 levels (a PGC-1α target gene) were significantly reduced in PGC-1α-null cells upon glucose deprivation relative to empty vector controls (Fig. 6C). In contrast, Gls1 levels were comparable between PGC-1α (CR#2) and EV control cells (Fig. 6C), which may explain why tumor onset was less severely impaired with the PGC-1α (CR#2) cohort (Fig. 6A). However, PGC-1α was dispensable for tumor growth (Fig. 6B). This is consistent with the observation that PGC-1α-null tumors display comparable rates of proliferation, apoptosis and angiogenesis relative to empty vector controls (Supplementary Fig. S7). Collectively, these data suggest that PGC-1α serves an essential and nonredundant role during tumor emergence. Moreover, significant selective pressures are exerted upon PGC-1α-null tumors to reprogram their metabolism for subsequent tumor growth. Indeed, we recently showed that PGC-1α increases the global bioenergetic capacity and flexibility of breast cancer cells by increasing ATP production, both through increased glycolysis and oxidative phosphorylation (46).

The ability of biguanides to impair breast tumor initiation is attenuated by deregulated tyrosine kinase signaling

Our results raise the intriguing possibility that metabolic reprogramming is a central component of dysregulated tyrosine kinase signaling during the earliest stages of breast cancer progression. To test this, we used the polyoma virus middle T transgenic mouse model (MMTV/MT). While MT is a viral oncogene and lacks intrinsic kinase activity, it recruits tyrosine kinases and activates many of the same pathways as RTKs (47). We reduced endogenous ShcA signaling using a knock-in allele harboring phenylalanine substitution of all three tyrosine phosphorylation sites under the control of the endogenous ShcA promoter (MT/ShcA+/3F) (29). To inhibit mitochondrial metabolism, MT/ShcA+/+ and MT/ShcA+/3F mice were treated with phenformin in the drinking water at first tumor palpation. Compared with the placebo group, phenformin neither impacted the number of tumor-bearing glands (Fig. 6D) nor the average or total tumor volumes at endpoint in MT/ShcA+/+ animals (Fig. 6E and F). In stark contrast, phenformin significantly reduced the number of tumor-bearing glands (average 8 vs. 5 tumors) in MT/ShcA+/3F animals (Fig. 6D). Moreover, phenformin-treated MT/ShcA+/3F mice showed a significantly reduced tumor volumes (3.8-fold) compared with control animals (Fig. 6E and F). These data suggest that the ability of biguanides to reduce breast cancer incidence critically depends on the ShcA activation status in the breast epithelium.

Taken together, our observations demonstrate that elevated ShcA signaling engages PGC-1α to increase the metabolic rate and bioenergetic flexibility of breast cancers. This ShcA/PGC-1α axis contributes significantly to breast tumor emergence and growth along with resistance to biguanides.

Numerous studies have implicated dysregulated RTK signaling in metabolic reprogramming. For example, ErbB2 stimulates glycolytic metabolism in cancer cells, in part, through its ability to upregulate the expression of glycolytic enzymes (20). Breast cancer cells that develop resistance to ErbB2-targeted therapies increase their rate of glycolysis and can also be sensitized to glycolytic inhibitors (5, 6). This reinforces the idea that metabolic reprogramming toward a glycolytic phenotype is an essential feature of RTK-driven breast tumors. In this study, we provide the first genetic evidence that the ShcA adaptor transduces oncogenic signals that permit metabolic perturbations underlying the glucose dependency of breast cancers.

To cope with their increased metabolic rate, breast tumors that hyperactivate ShcA signaling significantly upregulate PGC-1α levels, leading to elevated mitochondrial metabolism and glucose supply (40). Thus, the ShcA/PGC-1α signaling axis is likely to create a feed-forward mechanism to ensure that the proliferative and metabolic needs of aggressive breast tumors are continuously met (Fig. 7). This is consistent with the fact that PGC-1α is essential to support a higher metabolic rate in aggressive breast cancers. We recently demonstrated that PGC-1α also promotes bioenergetic flexibility, meaning that elevated PGC-1α levels readily allow cancer cells to switch between mitochondrial and glycolytic metabolism for ATP production (46). Increased ShcA signaling decreases the dependence of breast tumors on either pathway for energy production, rendering them less sensitive to biguanides, which block mitochondrial respiration. In contrast, impaired ShcA signaling lowers PGC-1α levels, which decreases the mitochondrial capacity of breast cancer cells yet renders them more dependent on mitochondrial metabolism, which sensitizes them to biguanides (Fig. 7).

Figure 7.

Breast cancer cells engage the ShcA/PGC-1α pathway to increase bioenergetic flexibility and resistance to biguanides. Schematic diagram illustrating how tyrosine kinases engage the ShcA pathway to promote metabolic flexibility by increasing PGC-1α expression. This response permits breast cancer cells to efficiently metabolize glucose and glutamine, providing adequate sources of fuel both through aerobic glycolysis and glucose oxidative phosphorylation, rendering them resistant to biguanides. By decreasing signaling downstream of ShcA, glucose metabolism is debilitated in breast cancer cells, which increases their dependency on mitochondrial metabolism for ATP production, exposing a therapeutic vulnerability to increased biguanide sensitivity.

Figure 7.

Breast cancer cells engage the ShcA/PGC-1α pathway to increase bioenergetic flexibility and resistance to biguanides. Schematic diagram illustrating how tyrosine kinases engage the ShcA pathway to promote metabolic flexibility by increasing PGC-1α expression. This response permits breast cancer cells to efficiently metabolize glucose and glutamine, providing adequate sources of fuel both through aerobic glycolysis and glucose oxidative phosphorylation, rendering them resistant to biguanides. By decreasing signaling downstream of ShcA, glucose metabolism is debilitated in breast cancer cells, which increases their dependency on mitochondrial metabolism for ATP production, exposing a therapeutic vulnerability to increased biguanide sensitivity.

Close modal

Our previous work implicates elevated AKT/mTORC1 signaling as a critical mediator of ShcA-driven breast cancer progression (23). Numerous studies have highlighted an important role for this pathway in cellular metabolism. First, AKT increases glycolytic activity in cancer cells (48). Second, mTORC1 signaling activates a metabolic network, which couples increased glycolysis with protein, nucleotide, and lipid synthesis (49). Finally, mTORC1 signaling increases mRNA translation of nuclear-encoded mitochondrial genes in a 4E-BP–dependent manner to augment oxidative phosphorylation (50, 51). Our data suggest that ShcA functions as a molecular bridge that allows tyrosine kinases to drive metabolic reprogramming by engaging AKT/mTORC1. Thus, the ShcA pathway may allow breast cancer cells to coordinate signaling (AKT/mTORC1), transcriptional (PGC-1α), and translational (eIF4E) control of tumor metabolism.

Significant research efforts have focused on examining whether biguanides can be repurposed as anticancer agents. Epidemiologic and clinical studies have examined whether biguanides associate with decreased risk of developing breast cancer or influence clinical variables associated with tumor growth. For example, a phase I study where metformin was administered to nondiabetic patients with breast cancer in the neoadjuvant setting showed decreased cell proliferation, which was associated with reduced insulin receptor expression and signaling in metformin-treated tumors (52). This study provided the first experimental evidence that metformin may exert antineoplastic activities in breast cancers. Given these observations, activation of the ShcA signaling pathway may help determine patients with breast cancer who would achieve maximal benefit from biguanides.

Globally, our work highlights the importance of metabolism in fueling RTK/ShcA signaling during all stages of the tumorigenic program. We show that the ShcA pathway increases PGC-1α driven metabolic reprogramming to augment the metabolic rate of mammary tumors and render them resistant to agents that target mitochondrial metabolism. This suggests that the ShcA/PGC-1α axis may define sensitivity to appropriate metabolic-based therapies, both to decrease the risk of disease progression in patients with early-stage breast cancer and to treat women with invasive carcinoma.

No potential conflicts of interest were disclosed.

Conception and design: I. Topisirovic, M. Pollak, J. St-Pierre, J. Ursini-Siegel

Development of methodology: Y.K. Im, O. Najyb, S.-P. Gravel, S. McGuirk, J. St-Pierre, J. Ursini-Siegel

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y.K. Im, O. Najyb, S.-P. Gravel, S. McGuirk, R. Ahn, D.Z. Avizonis, V. Chénard, V. Sabourin, J. Hudson, T. Pawson

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y.K. Im, O. Najyb, S.-P. Gravel, S. McGuirk, M. Pollak, J. St-Pierre, J. Ursini-Siegel

Writing, review, and/or revision of the manuscript: Y.K. Im, O. Najyb, S. McGuirk, M. Pollak, J. St-Pierre, J. Ursini-Siegel

Study supervision: M. Pollak, J. St-Pierre, J. Ursini-Siegel

Other (design of diagrams and artwork for representation of isotope tracing): S. McGuirk

This work was supported by CIHR grants to J. Ursini-Siegel (MOP-111143) and J. Ursini-Siegel and J. St-Pierre (MOP-244105). We further acknowledge support from the small animal research and pathology core facilities at the Lady Davis Institute and Goodman Cancer Research Centre (GCRC). GC/MS and tracer analyses were performed at the Rosalind and Morris Goodman Cancer Research Centre Metabolomics Core Facility supported by The Dr. John R. and Clara M. Fraser Memorial Trust, the Terry Fox Foundation [TFF Oncometabolism Team Grant (TFF-116128) in partnership with the Foundation du Cancer du Sein du Quebec], and McGill University. We are thankful to Dr. Andrée Gravel and Dr. Anne-Laure Larroque for the final NMR sample preparation and data acquisition at the Drug Discovery Platform (MUHC-RI). J. Ursini-Siegel is the recipient of a Senior FRQS salary support award. J. St-Pierre and I. Topisirovic acknowledge Junior 2 FRQS salary support awards. Y.K. Im, R. Ahn, and J. Hudson were supported by an FRQS Doctoral Award. O. Najyb is supported by a Canderel fellowship and S. McGuirk by a Vanier Canada Graduate Scholarship–CIHR.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Ward
PS
,
Thompson
CB
. 
Metabolic reprogramming: a cancer hallmark even warburg did not anticipate
.
Cancer Cell
2012
;
21
:
297
308
.
2.
Vander Heiden
MG
,
Cantley
LC
,
Thompson
CB
. 
Understanding the Warburg effect: the metabolic requirements of cell proliferation
.
Science
2009
;
324
:
1029
33
.
3.
Groheux
D
,
Cochet
A
,
Humbert
O
,
Alberini
JL
,
Hindie
E
,
Mankoff
D
. 
(1)(8)F-FDG PET/CT for staging and restaging of breast cancer
.
J Nucl Med
2016
;
57
:
17S
26S
.
4.
Pelicano
H
,
Martin
DS
,
Xu
RH
,
Huang
P
. 
Glycolysis inhibition for anticancer treatment
.
Oncogene
2006
;
25
:
4633
46
.
5.
Ruprecht
B
,
Zaal
EA
,
Zecha
J
,
Wu
W
,
Berkers
CR
,
Kuster
B
, et al
Lapatinib resistance in breast cancer cells is accompanied by phosphorylation-mediated reprogramming of glycolysis
.
Cancer Res
2017
;
77
:
1842
53
.
6.
Zhao
Y
,
Liu
H
,
Liu
Z
,
Ding
Y
,
Ledoux
SP
,
Wilson
GL
, et al
Overcoming trastuzumab resistance in breast cancer by targeting dysregulated glucose metabolism
.
Cancer Res
2011
;
71
:
4585
97
.
7.
Fendt
SM
,
Bell
EL
,
Keibler
MA
,
Olenchock
BA
,
Mayers
JR
,
Wasylenko
TM
, et al
Reductive glutamine metabolism is a function of the alpha-ketoglutarate to citrate ratio in cells
.
Nat Commun
2013
;
4
:
2236
.
8.
Hensley
CT
,
Wasti
AT
,
DeBerardinis
RJ
. 
Glutamine and cancer: cell biology, physiology, and clinical opportunities
.
J Clin Invest
2013
;
123
:
3678
84
.
9.
Gross
MI
,
Demo
SD
,
Dennison
JB
,
Chen
L
,
Chernov-Rogan
T
,
Goyal
B
, et al
Antitumor activity of the glutaminase inhibitor CB-839 in triple-negative breast cancer
.
Mol Cancer Ther
2014
;
13
:
890
901
.
10.
Fendt
SM
,
Bell
EL
,
Keibler
MA
,
Davidson
SM
,
Wirth
GJ
,
Fiske
B
, et al
Metformin decreases glucose oxidation and increases the dependency of prostate cancer cells on reductive glutamine metabolism
.
Cancer Res
2013
;
73
:
4429
38
.
11.
Foretz
M
,
Guigas
B
,
Bertrand
L
,
Pollak
M
,
Viollet
B
. 
Metformin: from mechanisms of action to therapies
.
Cell Metab
2014
;
20
:
953
66
.
12.
Andrzejewski
S
,
Gravel
SP
,
Pollak
M
,
St-Pierre
J
. 
Metformin directly acts on mitochondria to alter cellular bioenergetics
.
Cancer Metab
2014
;
2
:
12
.
13.
Bridges
HR
,
Jones
AJ
,
Pollak
MN
,
Hirst
J
. 
Effects of metformin and other biguanides on oxidative phosphorylation in mitochondria
.
Biochem J
2014
;
462
:
475
87
.
14.
Wheaton
WW
,
Weinberg
SE
,
Hamanaka
RB
,
Soberanes
S
,
Sullivan
LB
,
Anso
E
, et al
Metformin inhibits mitochondrial complex I of cancer cells to reduce tumorigenesis
.
Elife
2014
;
3
:
e02242
.
15.
Gravel
SP
,
Hulea
L
,
Toban
N
,
Birman
E
,
Blouin
MJ
,
Zakikhani
M
, et al
Serine deprivation enhances antineoplastic activity of biguanides
.
Cancer Res
2014
;
74
:
7521
33
.
16.
Cheong
JH
,
Park
ES
,
Liang
J
,
Dennison
JB
,
Tsavachidou
D
,
Nguyen-Charles
C
, et al
Dual inhibition of tumor energy pathway by 2-deoxyglucose and metformin is effective against a broad spectrum of preclinical cancer models
.
Mol Cancer Ther
2011
;
10
:
2350
62
.
17.
Ogrodzinski
MP
,
Bernard
JJ
,
Lunt
SY
. 
Deciphering metabolic rewiring in breast cancer subtypes
.
Transl Res
2017
;
189
:
105
22
.
18.
Landis
J
,
Shaw
LM
. 
Insulin receptor substrate 2-mediated phosphatidylinositol 3-kinase signaling selectively inhibits glycogen synthase kinase 3beta to regulate aerobic glycolysis
.
J Biol Chem
2014
;
289
:
18603
13
.
19.
Natan
S
,
Tsarfaty
G
,
Horev
J
,
Haklai
R
,
Kloog
Y
,
Tsarfaty
I
. 
Interplay Between HGF/SF-Met-Ras signaling, tumor metabolism and blood flow as a potential target for breast cancer therapy
.
Oncoscience
2014
;
1
:
30
8
.
20.
Zhao
YH
,
Zhou
M
,
Liu
H
,
Ding
Y
,
Khong
HT
,
Yu
D
, et al
Upregulation of lactate dehydrogenase A by ErbB2 through heat shock factor 1 promotes breast cancer cell glycolysis and growth
.
Oncogene
2009
;
28
:
3689
701
.
21.
Pelicci
G
,
Lanfrancone
L
,
Grignani
F
,
McGlade
J
,
Cavallo
F
,
Forni
G
, et al
A novel transforming protein (SHC) with an SH2 domain is implicated in mitogenic signal transduction
.
Cell
1992
;
70
:
93
104
.
22.
Ahn
R
,
Sabourin
V
,
Bolt
AM
,
Hebert
S
,
Totten
S
,
De Jay
N
, et al
The Shc1 adaptor simultaneously balances Stat1 and Stat3 activity to promote breast cancer immune suppression
.
Nat Commun
2017
;
8
:
14638
.
23.
Im
YK
,
La Selva
R
,
Gandin
V
,
Ha
JR
,
Sabourin
V
,
Sonenberg
N
, et al
The ShcA adaptor activates AKT signaling to potentiate breast tumor angiogenesis by stimulating VEGF mRNA translation in a 4E-BP-dependent manner
.
Oncogene
2015
;
34
:
1729
35
.
24.
Ursini-Siegel
J
,
Cory
S
,
Zuo
D
,
Hardy
WR
,
Rexhepaj
E
,
Lam
S
, et al
Receptor tyrosine kinase signaling favors a protumorigenic state in breast cancer cells by inhibiting the adaptive immune response
.
Cancer Res
2010
;
70
:
7776
87
.
25.
Ursini-Siegel
J
,
Hardy
WR
,
Zuo
D
,
Lam
SH
,
Sanguin-Gendreau
V
,
Cardiff
RD
, et al
ShcA signalling is essential for tumour progression in mouse models of human breast cancer
.
EMBO J
2008
;
27
:
910
20
.
26.
Bergers
G
,
Benjamin
LE
. 
Tumorigenesis and the angiogenic switch
.
Nat Rev Cancer
2003
;
3
:
401
10
.
27.
Ursini-Siegel
J
,
Rajput
AB
,
Lu
H
,
Sanguin-Gendreau
V
,
Zuo
D
,
Papavasiliou
V
, et al
Elevated expression of DecR1 impairs ErbB2/Neu-induced mammary tumor development
.
Mol Cell Biol
2007
;
27
:
6361
71
.
28.
Guy
CT
,
Cardiff
RD
,
Muller
WJ
. 
Induction of mammary tumors by expression of polyomavirus middle T oncogene: a transgenic mouse model for metastatic disease
.
Mol Cell Biol
1992
;
12
:
954
61
.
29.
Hardy
WR
,
Li
L
,
Wang
Z
,
Sedy
J
,
Fawcett
J
,
Frank
E
, et al
Combinatorial ShcA docking interactions support diversity in tissue morphogenesis
.
Science
2007
;
317
:
251
6
.
30.
Buescher
JM
,
Antoniewicz
MR
,
Boros
LG
,
Burgess
SC
,
Brunengraber
H
,
Clish
CB
, et al
A roadmap for interpreting (13)C metabolite labeling patterns from cells
.
Curr Opin Biotechnol
2015
;
34
:
189
201
.
31.
Mamer
O
,
Gravel
SP
,
Choiniere
L
,
Chenard
V
,
St-Pierre
J
,
Avizonis
D
. 
The complete targeted profile of the organic acid intermediates of the citric acid cycle using a single stable isotope dilution analysis, sodium borodeuteride reduction and selected ion monitoring GC/MS
.
Metabolomics
2013
;
9
:
1019
30
.
32.
McGuirk
S
,
Gravel
SP
,
Deblois
G
,
Papadopoli
DJ
,
Faubert
B
,
Wegner
A
, et al
PGC-1alpha supports glutamine metabolism in breast cancer
.
Cancer Metab
2013
;
1
:
22
.
33.
Nanchen
A
,
Fuhrer
T
,
Sauer
U
. 
Determination of metabolic flux ratios from 13C-experiments and gas chromatography-mass spectrometry data: protocol and principles
.
Methods Mol Biol
2007
;
358
:
177
97
.
34.
Ursini-Siegel
J
,
Muller
WJ
. 
The ShcA adaptor protein is a critical regulator of breast cancer progression
.
Cell Cycle
2008
;
7
:
1936
43
.
35.
Griss
T
,
Vincent
EE
,
Egnatchik
R
,
Chen
J
,
Ma
EH
,
Faubert
B
, et al
Metformin antagonizes cancer cell proliferation by suppressing mitochondrial-dependent biosynthesis
.
PLoS Biol
2015
;
13
:
e1002309
.
36.
Metallo
CM
,
Gameiro
PA
,
Bell
EL
,
Mattaini
KR
,
Yang
J
,
Hiller
K
, et al
Reductive glutamine metabolism by IDH1 mediates lipogenesis under hypoxia
.
Nature
2011
;
481
:
380
4
.
37.
Mullen
AR
,
Wheaton
WW
,
Jin
ES
,
Chen
PH
,
Sullivan
LB
,
Cheng
T
, et al
Reductive carboxylation supports growth in tumour cells with defective mitochondria
.
Nature
2011
;
481
:
385
8
.
38.
Austin
S
,
St-Pierre
J
. 
PGC1alpha and mitochondrial metabolism–emerging concepts and relevance in ageing and neurodegenerative disorders
.
J Cell Sci
2012
;
125
:
4963
71
.
39.
Arany
Z
,
Foo
SY
,
Ma
Y
,
Ruas
JL
,
Bommi-Reddy
A
,
Girnun
G
, et al
HIF-independent regulation of VEGF and angiogenesis by the transcriptional coactivator PGC-1alpha
.
Nature
2008
;
451
:
1008
12
.
40.
Klimcakova
E
,
Chenard
V
,
McGuirk
S
,
Germain
D
,
Avizonis
D
,
Muller
WJ
, et al
PGC-1alpha promotes the growth of ErbB2/Neu-induced mammary tumors by regulating nutrient supply
.
Cancer Res
2012
;
72
:
1538
46
.
41.
Handschin
C
,
Rhee
J
,
Lin
J
,
Tarr
PT
,
Spiegelman
BM
. 
An autoregulatory loop controls peroxisome proliferator-activated receptor gamma coactivator 1alpha expression in muscle
.
Proc Natl Acad Sci U S A
2003
;
100
:
7111
6
.
42.
Liu
C
,
Lin
JD
. 
PGC-1 coactivators in the control of energy metabolism
.
Acta Biochim Biophys Sin (Shanghai)
2011
;
43
:
248
57
.
43.
Deblois
G
,
St-Pierre
J
,
Giguere
V
. 
The PGC-1/ERR signaling axis in cancer
.
Oncogene
2013
;
32
:
3483
90
.
44.
Ciron
C
,
Lengacher
S
,
Dusonchet
J
,
Aebischer
P
,
Schneider
BL
. 
Sustained expression of PGC-1alpha in the rat nigrostriatal system selectively impairs dopaminergic function
.
Hum Mol Genet
2012
;
21
:
1861
76
.
45.
Shin
JH
,
Ko
HS
,
Kang
H
,
Lee
Y
,
Lee
YI
,
Pletinkova
O
, et al
PARIS (ZNF746) repression of PGC-1alpha contributes to neurodegeneration in Parkinson's disease
.
Cell
2011
;
144
:
689
702
.
46.
Andrzejewski
S
,
Klimcakova
E
,
Johnson
RM
,
Tabaries
S
,
Annis
MG
,
McGuirk
S
, et al
PGC-1alpha promotes breast cancer metastasis and confers bioenergetic flexibility against metabolic drugs
.
Cell Metab
2017
.
47.
Fluck
MM
,
Schaffhausen
BS
. 
Lessons in signaling and tumorigenesis from polyomavirus middle T antigen
.
Microbiol Mol Biol Rev
2009
;
73
:
542
63
,
Table of Contents
.
48.
Elstrom
RL
,
Bauer
DE
,
Buzzai
M
,
Karnauskas
R
,
Harris
MH
,
Plas
DR
, et al
Akt stimulates aerobic glycolysis in cancer cells
.
Cancer Res
2004
;
64
:
3892
9
.
49.
Duvel
K
,
Yecies
JL
,
Menon
S
,
Raman
P
,
Lipovsky
AI
,
Souza
AL
, et al
Activation of a metabolic gene regulatory network downstream of mTOR complex 1
.
Mol Cell
2010
;
39
:
171
83
.
50.
Gandin
V
,
Masvidal
L
,
Hulea
L
,
Gravel
SP
,
Cargnello
M
,
McLaughlan
S
, et al
nanoCAGE reveals 5′ UTR features that define specific modes of translation of functionally related MTOR-sensitive mRNAs
.
Genome Res
2016
;
26
:
636
48
.
51.
Morita
M
,
Gravel
SP
,
Chenard
V
,
Sikstrom
K
,
Zheng
L
,
Alain
T
, et al
mTORC1 controls mitochondrial activity and biogenesis through 4E-BP-dependent translational regulation
.
Cell Metab
2013
;
18
:
698
711
.
52.
Dowling
RJ
,
Niraula
S
,
Chang
MC
,
Done
SJ
,
Ennis
M
,
McCready
DR
, et al
Changes in insulin receptor signaling underlie neoadjuvant metformin administration in breast cancer: a prospective window of opportunity neoadjuvant study
.
Breast Cancer Res
2015
;
17
:
32
.