Tumor cells alter their metabolism to meet their demand for macromolecules and support a high rate of proliferation as well as cope with oxidative stress. The transcription factor yin yang 1 (YY1) is upregulated in various types of tumors and is crucial for tumor cell proliferation and metastasis. However, its role in tumor cell metabolic reprogramming is poorly understood. Here, we show that YY1 alters tumor cell metabolism by activating glucose-6-phosphate dehydrogenase (G6PD), the rate-limiting enzyme in the pentose phosphate pathway. By stimulating the pentose phosphate pathway, YY1 enhanced production of nucleotides and DNA synthesis, decreased intracellular reactive oxygen species levels, and promoted antioxidant defense by supplying increased reducing power in the form of NADPH. Importantly, YY1-mediated regulation of the pentose phosphate pathway in tumor cells occurred not through p53, but rather through direct activation of G6PD transcription by YY1. Regulation of pentose phosphate pathway activity through G6PD was strongly related to YY1-induced proliferation of tumor cells and tumorigenesis. Together, our results describe a novel role for YY1 in regulating G6PD in a p53-independent manner, which links its function in tumorigenesis to metabolic reprogramming in tumor cells.

Significance: This study reveals a novel role for YY1 in regulating G6PD and activating the pentose phosphate pathway, linking its function in tumorigenesis to metabolic reprogramming. Cancer Res; 78(16); 4549–62. ©2018 AACR.

Metabolic reprogramming is an important hallmark of tumor cells and a key feature of the malignant phenotype (1). As first discovered by Otto Warburg, compared with normal cells, tumor cells consume glucose at a significantly high rate and prefer to metabolize it through glycolysis even in the presence of an adequate oxygen supply (2–4). Through this metabolic alteration, known as the Warburg effect, more than 90% of glucose is converted to lactate (4), which conditions the tumor cell environment to favor their invasion (5). Metabolic reprogramming not only upregulates tumor cell adaptation potential to a fluctuating oxygen tension (6), but also allows production of glycolytic intermediates, which could in turn fuel the synthesis of macromolecules required for the formation of new cells (7). Metabolic reprogramming also enhances the pentose phosphate pathway (PPP), which produces ribose-5-phosphate for de novo nucleotide biosynthesis (5). The PPP also generates nicotinamide adenine dinucleotide phosphate (NADPH), which is crucial for supporting cell proliferation via the reductive biosynthesis of macromolecules such as lipids. Moreover, NADPH acts as a reactive oxygen species (ROS) scavenger and thus is critical for the antioxidant defense of the cells (5, 8). However, despite the crucial role of the PPP in tumorigenesis, its regulatory mechanism is not fully understood.

Yin yang 1 (YY1) is the first identified member of the Yin Yang (YY) family, a GLI-Krüppel zinc finger protein family that contains four highly conserved C2H2 zinc finger domains (9, 10). YY1 is a multifunctional transcription factor responsible for activating or inhibiting transcription of target genes depending on the context and the recruitment of specific cofactors (11). YY1 has been estimated to regulate about 7% of mammalian genes (12). Despite its function as a transcriptional regulator, recent studies have indicated that YY1 could regulate gene expression even through other mechanisms, for example, by stabilizing hypoxia-inducible factor 1 alpha (HIF1α; ref. 13). Not surprisingly, YY1 is implicated in various biological and physiological processes, including DNA replication, cell proliferation, embryonic development, differentiation, and disease pathways such as dystrophic muscle disease (9, 13–15). Meanwhile, YY1 overexpression could be observed in various cancers, including colon carcinoma, breast carcinoma, pancreatic cancer, and bladder cancer (16–19). We and other groups have also reported that YY1 plays crucial roles in tumorigenesis, promoting tumor cell proliferation and tumor angiogenesis in both a p53-dependent and p53-independent manner (11, 13, 18–20).

In spite of its documented roles in tumorigenesis, whether YY1 is involved in the regulation of tumor cell metabolic reprogramming, and whether this regulation could have a determinant effect on tumor cell proliferation are still unclear. Here, we report that YY1 activates the expression of glucose-6-phosphate dehydrogenase (G6PD), the key enzyme of the PPP. By stimulating the PPP, YY1 alters tumor cell metabolism, enhances nucleotide production and DNA synthesis, and promotes antioxidant defense by increasing NADPH. All of these adaptations are crucial for tumor cell proliferation. Concomitantly, G6PD overexpression rescues the suppression of cell proliferation and tumorigenesis induced by YY1 silencing. Importantly, we show that YY1 regulates G6PD through direct binding to the G6PD promoter, even in the absence of the tumor suppressor p53. Together, our findings not only unravel a pivotal role of YY1 in promoting tumor cell proliferation by inducing a metabolic shift, especially toward the PPP, but also provide new insights regarding the complex role of YY1 in driving tumorigenesis. Furthermore, our results strongly suggest that YY1 might be a potential target for cancer therapy regardless of the p53 status of the patient.

Cell lines and cell culture

HeLa, U2OS, and HepG2 cells were maintained in Dulbecco's modified Eagle's medium (Gibco, Life Technologies). HCT116WT and HCT116p53null cells were kindly provided by Dr. Bert Vogelstein at John Hopkins University School of Medicine, Baltimore, MD (21) and maintained in McCoy's 5A medium (Gibco). All cell lines were cultured with 10% fetal bovine serum (Biological Industries) and 1% penicillin–streptomycin. All cell lines were verified using short-tandem repeat profiling method at Genetic Testing Biotechnology (for HCT116p53null cell) and at Cell Bank, Type Culture Collection, Chinese Academy of Science (for other cell lines), and tested for Mycoplasma contamination using Mycoplasma Detection Kit-Quick Test (Biotool) routinely every 6 months. Cells were transfected with indicated vectors using Lipofectamine 2000 (Invitrogen Life Technologies) according to the manufacturer's protocol.

For gene-silencing experiments, cells were seeded in 6-well plates, and transfected with 2 μg of indicated shRNA expression vectors. Twenty-four hours after transfection, puromycin selection was performed to eliminate untransfected cells. For establishing YY1-silenced HCT116p53null cells (HCT116p53null/shYY1 cells) and YY1-silenced, G6PD-overexpressed HCT116p53null cells (HCT116p53null/shYY1/G6PD cells), cells were seeded in 10-cm well plates, and transfected with 12 μg of shCon or shYY1 and 6 μg of pcEF9-Puro or pG6PD-Puro vectors. Stable cell lines were established by performing puromycin selection. For overexpression experiments, cells were seeded in 6-well plates and transfected with 2 μg of pcYY1. For rescue experiments, cells were transfected with 1 μg of shYY1 and 1 μg of pcG6PD vectors and subjected to puromycin selection to eliminate untransfected cells.

Clinical human colon carcinoma specimens

Human colon carcinoma fresh specimens were obtained from patients undergoing surgery at Chongqing University Cancer Hospital (Chongqing, China). Patients did not receive chemotherapy, radiotherapy, or other adjuvant therapies prior to the surgery. The specimens were snap frozen in liquid nitrogen. Prior patients' written informed consents were obtained. The studies were approved by the Institutional Research Ethics Committee of Chongqing University Cancer Hospital, and conducted in accordance with the Declaration of Helsinki.

Animal experiment

For the in vivo tumor study, BALB/c-nu/nu mice (male; body weight, 18–22 g; 6 weeks old) were purchased from the Third Military Medical University (Chongqing, China; permit number SYXK-PLA-20120031). Animal studies were approved by the Laboratory Animal Welfare and Ethics Committee of the Third Military Medical University and carried out in the Third Military Medical University. All animal experiments conformed to the guidelines of the Animal Care and Use Committee of the Third Military Medical University. All efforts to minimize suffering were made. To generate an experimental subcutaneous tumor model, BALB/c-nu/nu mice were randomly divided into three groups (n = 6), and each group was injected subcutaneously with 5 × 106 HCT116p53null/Con, HCT116p53null/shYY1, or HCT116p53null/shYY1/G6PD stable cells. Tumor size (V) was evaluated by a caliper every 2 days with reference to the following equation: V = a × b2/2 (13), where a and b are the major and minor axes of the tumor, respectively. The investigator was blinded to the group allocation and during the assessment.

Plasmids and constructs

The shRNA control vector (shCon), shRNA expression vectors targeting YY1, YY2, and shp53, as well as YY1 overexpression vector were constructed as described previously (13, 22). The target sequences specific for YY1 are shYY1-1 (5′-GCAAGAAGAGTTACCTCAG-3′) and shYY1-2 (5′- GTCCAGAATACTTATAATT-3′), while those for YY2 are shYY2-1 (5′-GCATCAACATCAACATCAA-3′) and shYY2-2 (5′-ACATCAACATCAACCCAGA-3′), and those for p53 are shp53 (5′- GCAAGAAGGGAGACAAGAT-3′). For G6PD overexpression vector (pcG6PD), the coding region of human G6PD was amplified using the Takara Ex Taq Kit (Takara Bio) from human cDNA obtained by reverse-transcribing total RNA extracted from HCT116WT cells using the PrimeScript RT Reagent Kit with gDNA Eraser (Takara Bio). The amplicon was inserted into the HindIII and EcoRI sites of pcDNA3.1(+; Invitrogen Life Technologies). For pG6PD-Puro vector, the amplicon was inserted into the pcEF9-Puro vector bringing puromycin resistance gene.

For G6PD luciferase reporter vectors, we cloned the −419 to +1,428 (for G6PD-Luc) or the +58 to +1,428 (for G6PD-del-Luc) of the G6PD promoter region into the EcoRV and HindIII sites of the pGL4.13 vector (Promega). Human genome DNA extracted from HCT116WT cells using the TIANamp Genomic DNA Kit (Tiangen Biotech) was used as template for amplifying the promoter regions. G6PD-luciferase vector with mutated YY1 binding site (G6PDmut-Luc) was constructed based on the site-specific mutagenesis method (17).

Western blotting and quantitative PCR analysis

Detailed methods for performing Western blotting and quantitative PCR analysis are described in the Supplementary Materials and Methods. The sequences of the primers and antibodies used are shown in Supplementary Tables S1 and S2, respectively. Images of uncropped blots are shown in Supplementary Fig. S8A to S8F.

Glucose consumption, lactate production, G6PD and P6GD enzyme activities, and the intracellular NADPH level

Cells were transfected with indicated shRNA expression vectors or overexpression vectors as described above. For glucose consumption and lactate production, the culture medium was then changed, and the cells were incubated for an additional 24 hours. Glucose levels in the culture medium were measured using the Glucose Colorimetric Assay Kit (BioVision). Lactate levels were determined using a Lactate Assay kit (KeyGen Biotech). The values were normalized with total protein amount determined using the BCA Protein Assay Kit (Beyotime). G6PD enzyme activity and the intracellular NADPH level were determined using G6PD assay kit (Yuanye) or Amplite Colorimetric NADP/NADPH Ratio Assay Kit (AAT Bioquest), respectively, after the cells being lysed.

For measuring the activities of G6PD and 6PGD in clinical tumor tissues and xenografted tissues, the tissues were homogenized, then the enzyme activities were measured using the G6PD assay and 6PGD assay kits (Solarbio), and the relative level to the control was calculated.

Intracellular ROS level

Cells were transfected with indicated expression vectors as described above, and then the same number of the cells was reseeded to 24-well plates. The intracellular ROS level was stained using CellROX Green Reagent (Life Technologies). Images were taken with DMI6000B (Leica) and analyzed using ImageJ software. The results were expressed as the mean fluorescence intensity per cell.

Liquid chromatography–mass spectrometry

For profiling PPP metabolites 6-phosphogluconate and ribose-5-phosphate, cells transfected with indicated shRNA expression vectors as described above were washed twice with cold PBS, and then the intracellular metabolites were extracted by sonication in ice-cold 80% methanol. Supernatants were collected with centrifugation, rotary evaporated, and dissolved in 95% acetonitrile. Liquid chromatography–mass spectrometry (LC-MS) analysis was carried out on Agilent 1290 Infinity II UPLC tandem Agilent Triple Quard (QQQ) MS equipped with electrospray source operating in the negative-ion multiple-reaction monitoring mode. The source parameters were as follows: gas temperature 300°C, gas flow 8 L/minute, nebulizer 35 psi, sheath gas temperature 350°C, sheath gas flow 11 L/minute, capillary 3500 V, and nozzle voltage 1500 V. Chromatographic separation was performed on Acquity UPLC BEH HILIC (2.1 × 100 mm, 1.7 μm) column. Mobile phase A used consists of water, while mobile phase B consists of 95% acetonitrile and 5% water. The quantitative multiple-reactions monitoring transition of each particular metabolite were m/z 275 > 80 for 6-phosphogluconate, and m/z 229 > 183 for ribose-5-phosphate, respectively. Data were acquired and processed using MassHunter Workstation Acquisition (Version B.08.00). Peak areas of individual metabolites were normalized against the total protein amount. The relative levels of each metabolite to those of the control are calculated.

Statistical analysis

All values of the experimental results were presented as mean ± SEM of three independent experiments. Statistical analysis was performed using the Student t test. For clinical samples and xenograft experiments, statistical analysis was performed using one-way ANOVA. A value of P < 0.05 was considered statistically significant.

YY1 regulates glucose metabolism in tumor cells

To meet their highly proliferative growth, tumor cells alter their metabolism. We and others have shown that YY1 is crucial for tumor cell proliferation (11, 13, 17, 19, 20); however, it remains unclear whether this occurs via metabolic reprogramming. Therefore, we first analyzed whether YY1 expression levels affected glucose metabolism by measuring glucose consumption and lactate production in YY1-silenced tumor cells. To this end, we constructed two small-hairpin (shRNA) vectors against YY1 (Supplementary Fig. S1A and S1B) and a YY1 overexpression vector (Supplementary Fig. S1C). Notably, our shRNA expression vectors against YY1 and the YY1 overexpression vector did not affect the expression level of YY2, another member of the YY family (23, 24). Our results showed that knocking down YY1 significantly reduced glucose consumption by HCT116WT colon carcinoma cells, whereas overexpressing YY1 robustly enhanced it (Fig. 1A), indicating that YY1 positively regulates glucose metabolism in tumor cells. Moreover, YY1 silencing grossly suppressed lactate production (Fig. 1B, left), whereas YY1 overexpression conspicuously increased it (Fig. 1B, right). Thus, our results clearly demonstrated that YY1 alters glucose metabolism in tumor cells. To elucidate the underlying molecular mechanism, we screened for factors involved in glucose metabolism that could be potentially affected by YY1. As shown in Fig. 1C, YY1 silencing clearly reduced the expression of several factors involved in glucose metabolism, such as glucose transporter 1 (GLUT1), G6PD, 6-phosphofructokinase 2 (PFK2), TP53-induced glycolysis regulatory phosphatase (TIGAR), phosphoglycerate kinase 1 (PGK1), pyruvate kinase M1/2 (PKM2), and lactate dehydrogenase A (LDHA). Among them, the expression level of G6PD was the most affected, decreasing below 50% in YY1-silenced cells.

Figure 1.

YY1 regulates tumor cells' glucose metabolism. A, Glucose consumption level in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells. Total protein was used for normalization. B, Lactate production level in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells. Total protein was used for normalization. C, The mRNA expression levels of glucose metabolism-related genes in YY1-silenced HCT116WT cells were analyzed using qPCR. Cells transfected with shCon or pcCon were used as control, and the results are shown as relative to control. β-Actin was used for qPCR normalization. Quantitative data were expressed as mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.01; NS, not significant; pcCon, pcDNA3.1(+).

Figure 1.

YY1 regulates tumor cells' glucose metabolism. A, Glucose consumption level in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells. Total protein was used for normalization. B, Lactate production level in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells. Total protein was used for normalization. C, The mRNA expression levels of glucose metabolism-related genes in YY1-silenced HCT116WT cells were analyzed using qPCR. Cells transfected with shCon or pcCon were used as control, and the results are shown as relative to control. β-Actin was used for qPCR normalization. Quantitative data were expressed as mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.01; NS, not significant; pcCon, pcDNA3.1(+).

Close modal

To further confirm the role of YY1 in regulating G6PD, we assessed G6PD mRNA expression levels using two shYY1 vectors and found that both of them significantly suppressed G6PD mRNA (Fig. 2A, left); in contrast, YY1 overexpression grossly upregulated G6PD mRNA (Fig. 2A, right). Furthermore, YY1 silencing suppressed the G6PD protein level (Fig. 2B, top), whereas YY1 overexpression upregulated it (Fig. 2B, bottom). Concomitantly, YY1 silencing significantly repressed G6PD enzyme activity (Fig. 2C, left), while YY1 overexpression induced it (Fig. 2C, right). To confirm that this regulation was specific to YY1, we silenced YY2 using two different shRNA expression vectors and found that the level of G6PD was not affected (Supplementary Fig. S1D). Next, to assess whether this regulation was common to other tumor cells, we silenced YY1 expression in human hepatocarcinoma HepG2 cells, human cervical carcinoma HeLa cells, and human bone osteosarcoma epithelial U2OS cells (Supplementary Fig. S1E–S1G). We found that G6PD expression was conspicuously suppressed in these cells when YY1 was silenced (Fig. 2D).

Figure 2.

YY1 positively regulates G6PD expression. A, The mRNA expression level of G6PD in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells was analyzed using qPCR. The results are shown as relative to control. B, The protein expression level of G6PD in YY1-silenced (top) and YY1-overexpressed (bottom) HCT116WT cells was analyzed using Western blotting. C, G6PD activity in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells. Total protein was used for normalization. D, The protein expression level of G6PD in YY1-silenced HepG2 (left), HeLa (middle), and U2OS (right) cells was analyzed using Western blotting. E and F, The mRNA (E; n = 20) and protein (F) expression levels of YY1 and G6PD in clinical human colon carcinoma and corresponding adjacent tissues were analyzed using qPCR and Western blotting, respectively. G, The results of immunohistochemistry staining against YY1 and G6PD in clinical human colon carcinoma and corresponding adjacent tissues (scale bars, 100 μm). H, G6PD enzyme activity in clinical human colon carcinoma and corresponding adjacent tissue. Total protein was used for normalization, and the results are shown as relative to control. Cells transfected with shCon or pcCon were used as control. β-Actin was used for qPCR normalization and Western blotting loading control. Quantitative data are expressed as mean ± SEM of three independent experiments. **, P < 0.01. pcCon, pcDNA3.1(+).

Figure 2.

YY1 positively regulates G6PD expression. A, The mRNA expression level of G6PD in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells was analyzed using qPCR. The results are shown as relative to control. B, The protein expression level of G6PD in YY1-silenced (top) and YY1-overexpressed (bottom) HCT116WT cells was analyzed using Western blotting. C, G6PD activity in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells. Total protein was used for normalization. D, The protein expression level of G6PD in YY1-silenced HepG2 (left), HeLa (middle), and U2OS (right) cells was analyzed using Western blotting. E and F, The mRNA (E; n = 20) and protein (F) expression levels of YY1 and G6PD in clinical human colon carcinoma and corresponding adjacent tissues were analyzed using qPCR and Western blotting, respectively. G, The results of immunohistochemistry staining against YY1 and G6PD in clinical human colon carcinoma and corresponding adjacent tissues (scale bars, 100 μm). H, G6PD enzyme activity in clinical human colon carcinoma and corresponding adjacent tissue. Total protein was used for normalization, and the results are shown as relative to control. Cells transfected with shCon or pcCon were used as control. β-Actin was used for qPCR normalization and Western blotting loading control. Quantitative data are expressed as mean ± SEM of three independent experiments. **, P < 0.01. pcCon, pcDNA3.1(+).

Close modal

Next, we examined the expression levels of YY1 and G6PD in human clinical colon carcinoma tissues. Compared with adjacent normal tissue, mRNA expression of both YY1 and G6PD was robustly upregulated in tumor tissue (Fig. 2E, left and right, respectively). Furthermore, as shown by Western blotting and immunohistochemistry, YY1 and G6PD protein expression was significantly higher in tumor tissues compared with normal ones (Fig. 2F and G). In line with these results, we also detected a significant increase in G6PD enzyme activity in tumor tissue (Fig. 2H). Together, these results clearly suggest that YY1 is a positive regulator of G6PD expression, and that this regulatory mechanism is common to various tumor cell types.

YY1/G6PD regulation is crucial for the PPP

G6PD is known as the pacesetter of the PPP, and its activity determines the first committing step of this pathway (8). Aberrant G6PD expression and regulation had been found in tumor cells (25, 26); however, it remains unclear whether YY1 is involved in PPP regulation. To unravel the role of YY1 in PPP, we first examined the effect of YY1 expression on the intracellular NADPH level. As shown in Fig. 3A, the expression of YY1 correlated positively with the intracellular NADPH level. Concomitantly, YY1 silencing robustly increased the amount of intracellular ROS (Fig. 3B). YY1 silencing also reduced cell survival upon exposure to oxidative stress, whereas YY1 overexpression upregulated it (Supplementary Fig. S2A). Furthermore, we measured the amounts of 6-phosphogluconate and ribose-5-phosphate, two typical metabolites of the PPP, using LC-MS (Supplementary Fig. S2B and S2C), and found that their amounts were also conspicuously lower as a result of YY1 silencing (Fig. 3C). Ribose-5-phosphate is an intermediate in nucleotide production, which is crucial for de novo DNA synthesis. Not surprisingly, as demonstrated by the 5-ethynyl-2′-deoxyuridine (EdU) incorporation assay, YY1 silencing suppressed de novo DNA synthesis (Fig. 3D). Concomitantly, YY1 silencing grossly reduced the total number of HCT116WT cells and the number of HCT116WT colonies (Fig. 3E and F). Together, these results reveal that YY1 is critical for regulating the PPP and, consequently, the proliferation and colony formation potentials of tumor cells.

Figure 3.

YY1 affects the PPP. A, Intracellular NADPH level in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells. Total protein was used for normalization, and the results are shown as relative to control. B, Intracellular ROS level in YY1-silenced HCT116WT cells. Representative images (left) and relative fluorescence intensity (right) were shown. C, The amount of 6-phosphogluconate (6-P-gluconate) and ribose-5-phosphate (ribose-5-P) as analyzed using LC-MS. Total protein was used for normalization, and the results are shown as relative to control. D, The ratio of de novo DNA synthesis determined using EdU incorporation assay. Representative images (left) and the quantitative results (right) are shown. Hoechst was used to stain the nuclei. The ratio of de novo DNA synthesis is shown as relative to control. E, Total cell number of YY1-silenced HCT116WT cells was examined at indicated time points. F, Colony formation potentials of YY1-silenced HCT116WT cells. Cells transfected with shCon or pcCon were used as control. Quantitative data are expressed as mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.01. pcCon, pcDNA3.1(+). Scale bars, 200 μm.

Figure 3.

YY1 affects the PPP. A, Intracellular NADPH level in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells. Total protein was used for normalization, and the results are shown as relative to control. B, Intracellular ROS level in YY1-silenced HCT116WT cells. Representative images (left) and relative fluorescence intensity (right) were shown. C, The amount of 6-phosphogluconate (6-P-gluconate) and ribose-5-phosphate (ribose-5-P) as analyzed using LC-MS. Total protein was used for normalization, and the results are shown as relative to control. D, The ratio of de novo DNA synthesis determined using EdU incorporation assay. Representative images (left) and the quantitative results (right) are shown. Hoechst was used to stain the nuclei. The ratio of de novo DNA synthesis is shown as relative to control. E, Total cell number of YY1-silenced HCT116WT cells was examined at indicated time points. F, Colony formation potentials of YY1-silenced HCT116WT cells. Cells transfected with shCon or pcCon were used as control. Quantitative data are expressed as mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.01. pcCon, pcDNA3.1(+). Scale bars, 200 μm.

Close modal

To investigate whether G6PD was involved in YY1-mediated PPP regulation, we constructed a G6PD overexpression vector (Supplementary Fig. S3A), which we used together with shYY1 vector to perform rescue experiment (Supplementary Fig. S3B). We observed that G6PD activity and cellular NADPH level, both of which were significantly suppressed by YY1 silencing, were significantly restored when G6PD was overexpressed (Fig. 4A and B, respectively). NADPH is a key factor in suppressing intracellular ROS and enhancing lipid synthesis. Consistent with the trend shown by NADPH, YY1 silencing robustly increased intracellular ROS, while suppressing cell viability; both of these were restored by G6PD overexpression (Fig. 4C; Supplementary Fig. S3C). Similarly, lipid accumulation was also significantly restored in YY1-silenced cells overexpressing G6PD (Fig. 4D). Furthermore, G6PD overexpression eliminated the suppressive effect of YY1 silencing on PPP intermediates (6-phosphogluconate and ribose-5-phosphate) and de novo DNA synthesis (Fig. 4E and F). Taken together, these results strongly suggest that YY1 regulation of the PPP occurs by controlling G6PD expression.

Figure 4.

G6PD restores the effect of YY1 silencing on tumor cell proliferation by enhancing PPP. A, G6PD activity in YY1-silenced, G6PD-overexpressed HCT116WT cells. Total protein was used for normalization. B, Intracellular NADPH level in YY1-silenced, G6PD-overexpressed HCT116WT cells. Total protein was used for normalization, and the results are shown as relative to control. C, Intracellular ROS level in YY1-silenced, G6PD-overexpressed HCT116WT cells. Representative images (left) and relative fluorescence intensity (right) are shown. D, Lipid accumulation in YY1-silenced, G6PD-overexpressed HCT116WT cells was stained using Nile Red. Representative images are shown. E, The amounts of 6-phosphogluconate (6-P-gluconate) and ribose-5-phosphate (ribose-5-P) in YY1-silenced, G6PD-overexpressed HCT116WT cells as analyzed using LC-MS. Total protein was used for normalization, and the results are shown as relative to control. F, The ratio of de novo DNA synthesis determined using EdU incorporation assay. Representative images (left) and the quantitative results (right) are shown. Hoechst was used to stain the nuclei. The ratio of de novo DNA synthesis is shown as relative to control. G, Total cell number of YY1-silenced, G6PD-overexpressed HCT116WT cells was examined at indicated time points. H, Colony formation potentials of YY1-silenced, G6PD-overexpressed HCT116WT cells. Cells transfected with shCon and pcCon were used as control. Quantitative data are expressed as mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.01. pcCon, pcDNA3.1(+). Scale bars, 200 μm.

Figure 4.

G6PD restores the effect of YY1 silencing on tumor cell proliferation by enhancing PPP. A, G6PD activity in YY1-silenced, G6PD-overexpressed HCT116WT cells. Total protein was used for normalization. B, Intracellular NADPH level in YY1-silenced, G6PD-overexpressed HCT116WT cells. Total protein was used for normalization, and the results are shown as relative to control. C, Intracellular ROS level in YY1-silenced, G6PD-overexpressed HCT116WT cells. Representative images (left) and relative fluorescence intensity (right) are shown. D, Lipid accumulation in YY1-silenced, G6PD-overexpressed HCT116WT cells was stained using Nile Red. Representative images are shown. E, The amounts of 6-phosphogluconate (6-P-gluconate) and ribose-5-phosphate (ribose-5-P) in YY1-silenced, G6PD-overexpressed HCT116WT cells as analyzed using LC-MS. Total protein was used for normalization, and the results are shown as relative to control. F, The ratio of de novo DNA synthesis determined using EdU incorporation assay. Representative images (left) and the quantitative results (right) are shown. Hoechst was used to stain the nuclei. The ratio of de novo DNA synthesis is shown as relative to control. G, Total cell number of YY1-silenced, G6PD-overexpressed HCT116WT cells was examined at indicated time points. H, Colony formation potentials of YY1-silenced, G6PD-overexpressed HCT116WT cells. Cells transfected with shCon and pcCon were used as control. Quantitative data are expressed as mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.01. pcCon, pcDNA3.1(+). Scale bars, 200 μm.

Close modal

Finally, we examined whether YY1 regulation of G6PD could be linked to its role in tumor cell proliferation. YY1 silencing was reported to reduce the proliferation rate of tumor cells (13, 17). However, whether the effect of YY1 on tumor cell proliferation could be linked to its regulation of the PPP, and particularly G6PD expression, remains unknown. As shown in Fig. 4G and H, G6PD overexpression robustly restored both the proliferation and colony formation potentials previously suppressed by YY1 silencing. Together, these results strongly indicate that YY1 exerts its critical effect on tumor cell survival, proliferation, and colonization by regulating G6PD expression and enhancing the PPP.

YY1 regulates G6PD transcriptional activity by direct binding to the G6PD promoter region

Given that YY1 is a transcriptional factor, we used the UCSC Genome Browser (http://genome.ucsc.edu/, ref. 27) to search for a potential YY1 binding site within the G6PD promoter. Given that a YY1-binding site (ATGG, ref. 28) was predicted at position −205 to −202 of the G6PD promoter region, we constructed two luciferase reporter assay vectors with or without the predicted binding site (G6PD-Luc and G6PD-del-Luc, Fig. 5A, respectively). Luciferase reporter assay results clearly showed that knocking down YY1 could grossly suppress the activity of G6PD-Luc; however, this suppressive effect could not be observed when G6PD-del-Luc was used (Fig. 5B, left). Similarly, although YY1 overexpression could conspicuously promote the luciferase reporter activity of G6PD-Luc, it failed to have any significant effect on G6PD-del-Luc (Fig. 5B, right). These results indicated that the −419 to +57 region was critical for YY1 to regulate G6PD transcriptional activity. Next, we designed a set of primers for a chromatin immunoprecipitation (ChIP) assay, with which to confirm the binding of YY1 to the predicted promoter region (Fig. 5C, top). As shown in Fig. 5C (bottom), ChIP assay results clearly showed that YY1 could bind to the −312 to −110 region of the G6PD promoter. Finally, to assess whether the predicted binding site was functional, we constructed a G6PD luciferase reporter vector with four point mutations in the YY1 core binding site (G6PDmut-Luc): the ATGG sequence from the wild-type G6PD promoter was mutated into TCTT (Fig. 5D). We found that while YY1 silencing could robustly suppress the luciferase activity of the wild-type G6PD reporter vector (G6PD-Luc), this effect was diminished when G6PDmut-Luc was used (Fig. 5E, left). Similarly, while YY1 overexpression greatly induced G6PD-Luc activity, it failed to affect the luciferase activity of G6PDmut-Luc (Fig. 5E, right). Together, these results strongly indicate that YY1 could directly bind to the G6PD promoter, and that such binding is critical for its transcriptional regulation.

Figure 5.

YY1 directly binds G6PD promoter and regulates its transcriptional activity. A and B, The activities of G6PD-luciferase reporter vectors with or without predicted YY1 binding site (G6PD-Luc and G6PD-del-Luc, respectively). Schematic diagram (A) and relative luciferase activities to control (B) are shown. C, Binding of YY1 to the promoter region of G6PD was examined using ChIP assay with anti-YY1 antibody, followed by PCR in HCT116WT cells. The predicted YY1 binding site on the promoter region of G6PD and the location of primer set used for PCR are shown. D and E, The activity of YY1 binding site mutated G6PD-luciferase reporter vector (G6PDmut-Luc). Schematic diagram (D) and luciferase activities of G6PD-Luc and G6PDmut-Luc in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells (E) are shown. Luciferase activities are shown as relative to control. Mutated base pairs are indicated in gray. F and G, The mRNA (F) and protein (G) expression levels of G6PD in YY1-silenced (left) and YY1-overexpressed (right) HCT116p53null cells were analyzed using qPCR and Western blotting, respectively. β-Actin was used for qPCR normalization and Western blotting loading control. The mRNA expression levels are shown as relative to control. H, G6PD activity in YY1-silenced (left) and YY1-overexpressed (right) HCT116p53null cells. Total protein was used for normalization. Cells transfected with shCon or pcCon were used as control. Quantitative data are expressed as mean ± SEM of three independent experiments. **, P < 0.01; NS, not significant; pcCon, pcDNA3.1(+).

Figure 5.

YY1 directly binds G6PD promoter and regulates its transcriptional activity. A and B, The activities of G6PD-luciferase reporter vectors with or without predicted YY1 binding site (G6PD-Luc and G6PD-del-Luc, respectively). Schematic diagram (A) and relative luciferase activities to control (B) are shown. C, Binding of YY1 to the promoter region of G6PD was examined using ChIP assay with anti-YY1 antibody, followed by PCR in HCT116WT cells. The predicted YY1 binding site on the promoter region of G6PD and the location of primer set used for PCR are shown. D and E, The activity of YY1 binding site mutated G6PD-luciferase reporter vector (G6PDmut-Luc). Schematic diagram (D) and luciferase activities of G6PD-Luc and G6PDmut-Luc in YY1-silenced (left) and YY1-overexpressed (right) HCT116WT cells (E) are shown. Luciferase activities are shown as relative to control. Mutated base pairs are indicated in gray. F and G, The mRNA (F) and protein (G) expression levels of G6PD in YY1-silenced (left) and YY1-overexpressed (right) HCT116p53null cells were analyzed using qPCR and Western blotting, respectively. β-Actin was used for qPCR normalization and Western blotting loading control. The mRNA expression levels are shown as relative to control. H, G6PD activity in YY1-silenced (left) and YY1-overexpressed (right) HCT116p53null cells. Total protein was used for normalization. Cells transfected with shCon or pcCon were used as control. Quantitative data are expressed as mean ± SEM of three independent experiments. **, P < 0.01; NS, not significant; pcCon, pcDNA3.1(+).

Close modal

p53 is known as a critical inhibitor of G6PD enzyme activity (8), whereas YY1 is known as a negative regulator of p53 (20). Our results indicated that YY1 regulated G6PD directly at its transcriptional level, most plausibly by binding to its promoter region. These facts suggested the possibility that YY1 might regulate G6PD expression in a p53-independent way. To verify this hypothesis, we first investigated the level of G6PD in HCT116WT cells, in which both YY1 and p53 were transiently silenced. Our results showed that YY1 silencing suppressed G6PD expression in p53-silenced cells (Supplementary Fig. S4A). Consequently, YY1 silencing reduced G6PD activity and glucose consumption previously induced by p53 silencing (Supplementary Fig. S4B and S4C, respectively), indicating the presence of a p53-independent way for YY1 to regulate G6PD. To further confirm this pathway, we investigated the effect of YY1 silencing in HCT116p53null cells. We first confirmed the knockdown efficacy of our shYY1 vectors in HCT116p53null cells (Supplementary Fig. S5A) and measured G6PD expression in YY1-silenced or YY1-overexpressing HCT116p53null cells. Our results clearly showed that in HCT116p53null cells, YY1 silencing suppressed G6PD mRNA and protein expression, whereas YY1 overexpression induced it (Fig. 5F and G). Furthermore, manipulation of YY1 levels affected G6PD activity even in the absence of p53 (Fig. 5H). The fact that YY1 could directly regulate G6PD transcription and alter tumor metabolism even in the absence of p53 was further confirmed by the suppression of glucose consumption and lactate production in p53-null colon cancer cells upon YY1 silencing, and their induction upon YY1 overexpression (Supplementary Fig. S5B and S5C). These results suggest that transcriptional regulation of G6PD by YY1 could occur in a p53-independent manner, further confirming the direct action of YY1 in regulating G6PD activity.

G6PD-mediated YY1 regulation of the PPP occurs in a p53-independent manner

To further verify the functionality of the p53-independent YY1/G6PD axis on tumor cell PPP, we first analyzed the effect of YY1 expression on the intracellular NADPH level in HCT116p53null cells. In the absence of p53, YY1 silencing significantly suppressed the level of intracellular NADPH (Fig. 6A, left); in contrast, YY1 overexpression grossly upregulated it (Supplementary Fig. S6A). Moreover, the effect of YY1 silencing on intracellular NADPH could be diminished by overexpressing G6PD (Fig. 6A, right). Concomitantly, the intracellular ROS level in HCT116p53null cells was robustly increased by YY1 silencing (Fig. 6B), but was restored by G6PD overexpression (Fig. 6C). YY1-dependent activation of G6PD and the PPP also regulate the antioxidant defenses of HCT116p53null cells. Indeed, cell viability upon exposure to oxidative stress correlated positively with YY1 expression (Supplementary Fig. S6B), whereas G6PD overexpression restored the antioxidant defense suppressed by YY1 silencing (Supplementary Fig. S6C). Intracellular lipid accumulation showed a similar trend: G6PD overexpression prevented the inhibition of lipid accumulation caused by YY1 silencing (Fig. 6D).

Figure 6.

YY1 regulates PPP p53 independently. A, Intracellular NADPH level in YY1-silenced (left) and YY1-silenced, G6PD-overexpressed (right) HCT116p53null cells was determined. Total protein was used for normalization, and the results are shown as relative to control. B and C, The level of ROS in YY1-silenced (B) and YY1-silenced, G6PD-overexpressed (C) HCT116p53null cells. Representative images (left) and relative fluorescence intensity (right) are shown. D, Lipid accumulation in YY1-silenced, G6PD-overexpressed HCT116p53null cells was stained using Nile Red. Representative images are shown. E, The ratio of de novo DNA synthesis determined using EdU incorporation assay in YY1-silenced, G6PD-overexpressed HCT116p53null cells. Representative images (left) and the quantitative results (right) are shown. Hoechst was used to stain the nuclei. The ratio of de novo DNA synthesis is shown as relative to control. F, Total cell number of YY1-silenced, G6PD-overexpressed HCT116p53null cells was examined at indicated time points. G, Colony formation potentials of YY1-silenced, G6PD-overexpressed HCT116p53null cells. H, Relative cell number of YY1-silenced HCT116p53null cells cultured in the presence of nucleosides mixture (Nuc, final concentration, 200 μmol/L each) and/or ROS scavenger (N-acetyl-l-cysteine, Nac, final concentration, 2 mmol/L) as measured at the third day after the addition of nucleosides mixture and/or ROS scavenger. Cells transfected with shCon or shCon and pcCon were used as control. Quantitative data are expressed as mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.01. pcCon, pcDNA3.1(+). Scale bars, 200 μm.

Figure 6.

YY1 regulates PPP p53 independently. A, Intracellular NADPH level in YY1-silenced (left) and YY1-silenced, G6PD-overexpressed (right) HCT116p53null cells was determined. Total protein was used for normalization, and the results are shown as relative to control. B and C, The level of ROS in YY1-silenced (B) and YY1-silenced, G6PD-overexpressed (C) HCT116p53null cells. Representative images (left) and relative fluorescence intensity (right) are shown. D, Lipid accumulation in YY1-silenced, G6PD-overexpressed HCT116p53null cells was stained using Nile Red. Representative images are shown. E, The ratio of de novo DNA synthesis determined using EdU incorporation assay in YY1-silenced, G6PD-overexpressed HCT116p53null cells. Representative images (left) and the quantitative results (right) are shown. Hoechst was used to stain the nuclei. The ratio of de novo DNA synthesis is shown as relative to control. F, Total cell number of YY1-silenced, G6PD-overexpressed HCT116p53null cells was examined at indicated time points. G, Colony formation potentials of YY1-silenced, G6PD-overexpressed HCT116p53null cells. H, Relative cell number of YY1-silenced HCT116p53null cells cultured in the presence of nucleosides mixture (Nuc, final concentration, 200 μmol/L each) and/or ROS scavenger (N-acetyl-l-cysteine, Nac, final concentration, 2 mmol/L) as measured at the third day after the addition of nucleosides mixture and/or ROS scavenger. Cells transfected with shCon or shCon and pcCon were used as control. Quantitative data are expressed as mean ± SEM of three independent experiments. *, P < 0.05; **, P < 0.01. pcCon, pcDNA3.1(+). Scale bars, 200 μm.

Close modal

Next, we examined the role of the YY1/G6PD pathway on de novo DNA synthesis in p53-null tumor cells. We found that while YY1 silencing significantly suppressed the number of EdU-positive HCT116p53null cells (Supplementary Fig. S6D), overexpression of G6PD restored it (Fig. 6E). Furthermore, YY1 silencing suppressed the proliferation and colony formation potentials of HCT116p53null cells (Supplementary Fig. S6E and S6F), whereas G6PD blocked these effects (Fig. 6F and G).

Given that the PPP could produce ribose-5-phosphate and NADPH, and thus is crucial for the biosynthesis of nucleotides and for suppressing intracellular ROS, we examined the effect of adding an ROS scavenger and/or a mixture of four deoxyribonucleosides and four ribonucleosides on the proliferation potential of YY1-silenced HCT116p53null cells as described previously (25, 26). As shown in Fig. 6H, addition of both ROS scavenger and nucleosides mixture significantly reverted the effect of YY1 silencing on the proliferation of HCT116p53null cells. Thus, increased intracellular ROS level and impaired nucleotide biosynthesis contribute to the disruption of tumor cell proliferation caused by YY1 silencing. Together, our results demonstrate, for the first time, that YY1 could activate the PPP in tumor cells in a p53-independent manner by directly regulating G6PD.

The YY1/G6PD axis alters tumor growth in a p53-independent manner

Finally, we examined the pathological function of the YY1/G6PD pathway in tumorigenesis in vivo by transplanting HCT116p53null/Con, HCT116p53null/shYY1, and HCT116p53null/shYY1/G6PD stable cell lines subcutaneously into BALB/c-nu/nu mice. As shown in Fig. 7A and B, while YY1 silencing significantly decreased the tumorigenesis potential of HCT116p53null cells, G6PD overexpression restored it. Furthermore, Western blotting revealed that expression of G6PD was negatively regulated in YY1-silenced cells (Fig. 7C), and immunohistochemistry of tissue sections from xenografted tumors further confirmed the above results (Fig. 7D). Moreover, a significant decrease in the enzyme activities of G6PD and 6-phosphogluconate dehydrogenase (6PGD), which had been reported to quantitatively correspond to G6PD activity (29), was observed in the tumors produced from HCT116p53null/shYY1 cells; however, the activities of both enzymes were restored in xenografted tumors originating from HCT116p53null/shYY1/G6PD cells (Fig. 7E and F).

Figure 7.

G6PD is involved in YY1-mediated tumorigenesis. A and B, Tumorigenesis potentials of HCT116p53null/Con (Control), HCT116p53null/shYY1 (shYY1), and HCT116p53null/shYY1/pcG6PD (shYY1/G6PD) stable cell lines were examined in vivo by subcutaneous injection into Balb/c-nu/nu mice (n = 6). Volume of the tumor generated was measured at indicated time points (A), and the representative images are shown (B). C, The protein expression levels of YY1 and G6PD in the tumor formed were examined using Western blotting. β-Actin was used as loading control. D, Immunohistochemistry staining images against YY1 and G6PD in tissue sections of xenografted tumors in Balb/c-nu/nu mice injected with indicated cell lines. Scale bars, 50 μm. E and F, G6PD (E) and 6PGD (F) enzyme activities in tissues of xenografted tumors in Balb/c-nu/nu mice injected with indicated cell lines. Total protein was used for normalization, and the results are shown as relative to control. G, Schematic diagram showing the mechanism of YY1 regulation on the tumorigenesis via G6PD-induced PPP. Quantitative data are expressed as mean ± SEM of three independent experiments. **, P < 0.01.

Figure 7.

G6PD is involved in YY1-mediated tumorigenesis. A and B, Tumorigenesis potentials of HCT116p53null/Con (Control), HCT116p53null/shYY1 (shYY1), and HCT116p53null/shYY1/pcG6PD (shYY1/G6PD) stable cell lines were examined in vivo by subcutaneous injection into Balb/c-nu/nu mice (n = 6). Volume of the tumor generated was measured at indicated time points (A), and the representative images are shown (B). C, The protein expression levels of YY1 and G6PD in the tumor formed were examined using Western blotting. β-Actin was used as loading control. D, Immunohistochemistry staining images against YY1 and G6PD in tissue sections of xenografted tumors in Balb/c-nu/nu mice injected with indicated cell lines. Scale bars, 50 μm. E and F, G6PD (E) and 6PGD (F) enzyme activities in tissues of xenografted tumors in Balb/c-nu/nu mice injected with indicated cell lines. Total protein was used for normalization, and the results are shown as relative to control. G, Schematic diagram showing the mechanism of YY1 regulation on the tumorigenesis via G6PD-induced PPP. Quantitative data are expressed as mean ± SEM of three independent experiments. **, P < 0.01.

Close modal

Together, these findings reveal a novel role of YY1 in regulating the PPP in tumor cells. Through the direct activation of G6PD, YY1 induces tumor cell metabolic reprogramming in a p53-independent manner; this function is a crucial feature explaining the regulatory role of YY1 in tumorigenesis (Fig. 7G).

Metabolic alteration is a hallmark of tumor cells as it confers them important growth advantages. Discovered first by Otto Warburg (2), it is now clear that metabolic reprogramming facilitates tumor cell growth not only through the generation of ATP, but also by providing tumor cells with nucleotide and lipid precursors for macromolecular synthesis, as well as NADPH (4). The latter acts as a reducing equivalent for synthesizing macromolecules such as lipids, and as a crucial factor for tumor cell antioxidant defense (8, 25, 26, 30, 31). Our findings show that YY1 positively regulates glucose consumption and lactate production by tumor cells, thus linking YY1 to tumor cell metabolic reprogramming.

We demonstrate that YY1 enhances the PPP, a highly conserved metabolic pathway (32). The PPP, which is activated in various tumor cells (33), generates ribose-5-phosphate for de novo synthesis of the DNA and RNA phosphoribosylamine backbone (3) and is the main source of intracellular NADPH (34, 35). Excessive oxidative stress is hazardous as it could damage cellular components including DNA, proteins, and lipids (3, 36, 37), eventually initiating cell death signaling pathways, while NADPH provides the reducing power for both the glutathione and thioredoxin systems, which are crucial for the clearance of ROS produced during rapid cell proliferation and for repairing ROS-induced damage (35). Hence, the PPP links glucose catabolism to the biosynthesis of macromolecules essential for supporting the rapid proliferation of tumor cells and also contributes to tumor growth and survival under oxidative stress condition, which arises due to rapid proliferation of tumor cells.

Indeed, our results show that an increase in intracellular oxidative stress and a decrease in nucleotides biosynthesis contribute to the impaired tumor cell proliferation caused by YY1 silencing. Interestingly, restoration of either oxidative stress or the intracellular amount of nucleosides failed to significantly affect tumor cell proliferation, indicating that these two factors are of equal importance. Thus, our results strongly suggest that YY1 enhances the PPP and subsequently supports tumor cell viability and proliferation by acting as a guardian against excessive oxidative stress and by providing precursor for producing nucleic acids.

YY1 itself is a transcriptional regulator that could act positively or negatively, depending on its target genes (11). Aberrant YY1 expression has been confirmed in various cancers, including colon carcinoma, pancreatic cancer, bladder cancer, and breast cancer (13, 16–19). YY1 has been found to regulate various protumorigenic pathways by, for example, facilitating the interaction between MDM2 proto-oncogene and p53 to increase p53 ubiquitination (20), blocking p300-mediated p53 transcription and stabilization (38), stabilizing HIF1α (13), and inhibiting a tumor suppressor microRNA (18). Through these pathways, YY1 stimulates tumor cell proliferation, angiogenesis, and metastasis. Our present findings demonstrated a new role for YY1 in tumorigenesis, i.e., altering tumor cell metabolism by enhancing the PPP. Given that metabolic reprogramming is a key enabler of tumor cell proliferation, survival, and metastasis, the regulatory effect of YY1 on the PPP indicates that YY1 is critical for key functions of tumor cells.

As a transcription factor, YY1 operates on a wide range of targets involved in various biological pathways. Indeed, our results indicate that YY1 might affect, even if to a low extent, the expression of other glucose metabolism genes, including glycolysis-related factors GLUT1, PFK2, PKM2, and LDHA (Fig. 1C), as well as PPP-related factors 6PGD and RPE (Supplementary Fig. S7). Our results show that YY1-silencing robustly suppresses the amount of 6-phosphogluconate, a PPP intermediate produced in the upstream of 6PGD and RPE enzymatic reactions, and is the substrate of 6PGD; and that excessive ROS and impaired nucleotides biosynthesis critically influenced YY1 regulation of tumor cell growth, indicating that YY1 is critical for regulating PPP in tumor cells. Furthermore, given that G6PD restores cellular 6-phosphogluconate, ribose-5-phosphate, and NADPH levels, as well as tumor cell proliferation and colony formation potentials altered by YY1 manipulation, our results highlight the importance of the YY1/G6PD pathway. Hence, while YY1 might also affect other relevant metabolic pathways such as glycolysis, as well as ROS metabolism, our findings strongly suggest a novel, critical role of YY1 in regulating the PPP in tumor cells by acting on its pacesetter enzyme G6PD.

We show that regulation of the PPP by YY1 occurred regardless of p53 status, as it could bind directly and activate the G6PD promoter in p53-null cells. G6PD is the limiting enzyme that catalyzes the first committing step of the PPP, i.e., the oxidation of glucose-6-phosphate, and concomitantly reduces NADP+ to NADPH (34, 39). Deregulated G6PD expression could be observed in various tumors, such as colon cancer, glioblastoma, and breast cancer (40–42) and is related to metabolic alteration of tumor cells. YY1 has been known as an inhibitor of the p53 pathway (20, 38), and a recent study has shown that p53 is involved in suppressing the PPP through inactivation of G6PD (8); however, our findings clearly indicate that YY1 regulates G6PD even in the absence of p53. Thus, regulation of G6PD by YY1 could occur in both p53-dependent and p53-independent manners. Several current anticancer drugs inhibit tumor cell growth by activating p53 (43, 44) and p53 mutation and downregulation have been reported in many tumor patients (45). Therefore, together with our previous results describing the p53-independent function of YY1 in stabilizing HIF1α and inducing tumor angiogenesis (13), our present findings further emphasize the potential of targeting YY1 for cancer therapy. The benefit of such approach would entail a consistent clinical outcome regardless of p53 status.

In summary, we report a previously unknown YY1-dependent modulation of the PPP, which promotes tumor cell growth via enhanced biosynthesis of macromolecules required for tumor cell proliferation and protection from hazardous oxidative stress. These findings provide new mechanistic insights into the function of YY1 in tumorigenesis.

No potential conflicts of interest were disclosed.

Conception and design: S. Wu, V. Kasim

Development of methodology: H. Wang, Y. Li, Y. Xie, M. Miyagishi, V. Kasim

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Wu, H. Wang, Y. Li, Y. Xie, C. Huang

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Wu, H. Wang, Y. Li, Y. Xie, C. Huang, V. Kasim

Writing, review, and/or revision of the manuscript: S. Wu, H. Wang, Y. Li, V. Kasim

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): S. Wu, H. Zhao, M. Miyagishi, V. Kasim

Study supervision: S. Wu, V. Kasim

The authors thank Dr. Bert Vogelstein (Johns Hopkins University School of Medicine) for kindly providing HCT116WT and HCT116p53null cell lines and Dr. Dezhang Zhao (College of Pharmacy, Chongqing Medical University) for his technical assistance in LC-MS for metabolite analysis.

This work was supported by grants from the National Natural Science Foundation of China No. 81372202 (S. Wu and V. Kasim) and No. 31301119 (V. Kasim), the Natural Science Foundation of Chongqing No. cstc2014jcyjA10058 (V. Kasim), and the Fundamental Research Funds for the Central Universities No. 106112016CDJZR235516 (V. Kasim).

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.
Gatenby
RA
,
Gillies
RJ
. 
Why do cancers have high aerobic glycolysis?
Nat Rev Cancer
2004
;
4
:
891
9
.
2.
Warburg
O
. 
On the origin of cancer cells
.
Science
1956
;
123
:
309
14
.
3.
DeBerardinis
RJ
,
Chandel
NS
. 
Fundamentals of cancer metabolism
.
Sci Adv
2016
;
2
:
e1600200
.
4.
DeBerardinis
RJ
,
Lum
JJ
,
Hatzivassiliou
G
,
Thompson
CB
. 
The biology of cancer: metabolic reprogramming fuels cell growth and proliferation
.
Cell Metab
2008
;
7
:
11
20
.
5.
Kroemer
G
,
Pouyssegur
J
. 
Tumor cell metabolism: cancer's Achilles' heel
.
Cancer Cell
2008
;
13
:
472
82
.
6.
Pouyssegur
J
,
Dayan
F
,
Mazure
NM
. 
Hypoxia signalling in cancer and approaches to enforce tumour regression
.
Nature
2006
;
441
:
437
43
.
7.
Hanahan
D
,
Weinberg
RA
. 
Hallmarks of cancer: the next generation
.
Cell
2011
;
144
:
646
74
.
8.
Jiang
P
,
Du
W
,
Wang
X
,
Mancuso
A
,
Gao
X
,
Wu
M
, et al
p53 regulates biosynthesis through direct inactivation of glucose-6-phosphate dehydrogenase
.
Nat Cell Biol
2011
;
13
:
310
6
.
9.
Shi
Y
,
Seto
E
,
Chang
LS
,
Shenk
T
. 
Transcriptional repression by YY1, a human GLI-Kruppel-related protein, and relief of repression by adenovirus E1A protein
.
Cell
1991
;
67
:
377
88
.
10.
Seto
E
,
Shi
Y
,
Shenk
T
. 
YY1 is an initiator sequence-binding protein that directs and activates transcription in vitro
.
Nature
1991
;
354
:
241
5
.
11.
Gordon
S
,
Akopyan
G
,
Garban
H
,
Bonavida
B
. 
Transcription factor YY1: structure, function, and therapeutic implications in cancer biology
.
Oncogene
2006
;
25
:
1125
42
.
12.
Atchison
M
,
Basu
A
,
Zaprazna
K
,
Papasani
M
. 
Mechanisms of Yin Yang 1 in oncogenesis: the importance of indirect effects
.
Crit Rev Oncog
2011
;
16
:
143
61
.
13.
Wu
S
,
Kasim
V
,
Kano
MR
,
Tanaka
S
,
Ohba
S
,
Miura
Y
, et al
Transcription factor YY1 contributes to tumor growth by stabilizing hypoxia factor HIF-1alpha in a p53-independent manner
.
Cancer Res
2013
;
73
:
1787
99
.
14.
Donohoe
ME
,
Zhang
X
,
McGinnis
L
,
Biggers
J
,
Li
E
,
Shi
Y
. 
Targeted disruption of mouse Yin Yang 1 transcription factor results in peri-implantation lethality
.
Mol Cell Biol
1999
;
19
:
7237
44
.
15.
Atchison
ML
. 
Function of YY1 in long-distance DNA interactions
.
Front Immunol
2014
;
5
:
45
.
16.
Chinnappan
D
,
Xiao
D
,
Ratnasari
A
,
Andry
C
,
King
TC
,
Weber
HC
. 
Transcription factor YY1 expression in human gastrointestinal cancer cells
.
Int J Oncol
2009
;
34
:
1417
23
.
17.
Kasim
V
,
Xie
YD
,
Wang
HM
,
Huang
C
,
Yan
XS
,
Nian
WQ
, et al
Transcription factor Yin Yang 2 is a novel regulator of the p53/p21 axis
.
Oncotarget
2017
;
8
:
54694
707
.
18.
Yuan
P
,
He
XH
,
Rong
YF
,
Cao
J
,
Li
Y
,
Hu
YP
, et al
KRAS/NF-kappaB/YY1/miR-489 signaling axis controls pancreatic cancer metastasis
.
Cancer Res
2017
;
77
:
100
11
.
19.
Agarwal
N
,
Dancik
GM
,
Goodspeed
A
,
Costello
JC
,
Owens
C
,
Duex
JE
, et al
GON4L drives cancer growth through a YY1-androgen receptor-CD24 axis
.
Cancer Res
2016
;
76
:
5175
85
.
20.
Sui
G
,
Affar
el B
,
Shi
Y
,
Brignone
C
,
Wall
NR
,
Yin
P
, et al
Yin Yang 1 is a negative regulator of p53
.
Cell
2004
;
117
:
859
72
.
21.
Bunz
F
,
Dutriaux
A
,
Lengauer
C
,
Waldman
T
,
Zhou
S
,
Brown
JP
, et al
Requirement for p53 and p21 to sustain G2 arrest after DNA damage
.
Science
1998
;
282
:
1497
501
.
22.
Miyagishi
M
,
Taira
K
. 
Strategies for generation of an siRNA expression library directed against the human genome
.
Oligonucleotides
2003
;
13
:
325
33
.
23.
Nguyen
N
,
Zhang
X
,
Olashaw
N
,
Seto
E
. 
Molecular cloning and functional characterization of the transcription factor YY2
.
J Biol Chem
2004
;
279
:
25927
34
.
24.
Klar
M
. 
Yin Yang 2: the great unknown within the Yin Yang 1 regulatory network
.
Crit Rev Oncog
2011
;
16
:
239
43
.
25.
Du
W
,
Jiang
P
,
Mancuso
A
,
Stonestrom
A
,
Brewer
MD
,
Minn
AJ
, et al
TAp73 enhances the pentose phosphate pathway and supports cell proliferation
.
Nat Cell Biol
2013
;
15
:
991
1000
.
26.
Ma
X
,
Wang
L
,
Huang
,
Li
Y
,
Yang
D
,
Li
T
, et al
Polo-like kinase 1 coordinates biosynthesis during cell cycle progression by directly activating pentose phosphate pathway
.
Nat Commun
2017
;
8
:
1506
.
27.
Kent
WJ
,
Sugnet
CW
,
Furey
TS
,
Roskin
KM
,
Pringle
TH
,
Zahler
AM
, et al
The human genome browser at UCSC
.
Genome Res
2002
;
12
:
996
1006
.
28.
Klar
M
,
Stellamanns
E
,
Ak
P
,
Gluch
A
,
Bode
J
. 
Dominant genomic structures: detection and potential signal functions in the interferon-beta domain
.
Gene
2005
;
364
:
79
89
.
29.
Bijlsma
R
. 
Polymorphism at the G6pd and 6Pgd loci in Drosophila melanogaster. IV. Genetic factors modifying enzyme activity
.
Biochem Genet
1980
;
18
:
699
715
.
30.
Gillies
RJ
,
Gatenby
RA
. 
Adaptive landscapes and emergent phenotypes: why do cancers have high glycolysis?
J Bioenerg Biomembr
2007
;
39
:
251
7
.
31.
Jiang
P
,
Du
W
,
Wu
M
. 
Regulation of the pentose phosphate pathway in cancer
.
Protein Cell
2014
;
5
:
592
602
.
32.
Keller
MA
,
Zylstra
A
,
Castro
C
,
Turchyn
AV
,
Griffin
JL
,
Ralser
M
. 
Conditional iron and pH-dependent activity of a non-enzymatic glycolysis and pentose phosphate pathway
.
Sci Adv
2016
;
2
:
e1501235
.
33.
Patra
KC
,
Hay
N
. 
The pentose phosphate pathway and cancer
.
Trends Biochem Sci
2014
;
39
:
347
54
.
34.
Rao
X
,
Duan
X
,
Mao
W
,
Li
X
,
Li
Z
,
Li
Q
, et al
O-GlcNAcylation of G6PD promotes the pentose phosphate pathway and tumor growth
.
Nat Commun
2015
;
6
:
8468
.
35.
Cairns
RA
,
Harris
IS
,
Mak
TW
. 
Regulation of cancer cell metabolism
.
Nat Rev Cancer
2011
;
11
:
85
95
.
36.
Sena
LA
,
Chandel
NS
. 
Physiological roles of mitochondrial reactive oxygen species
.
Mol Cell
2012
;
48
:
158
67
.
37.
Weinberg
F
,
Hamanaka
R
,
Wheaton
WW
,
Weinberg
S
,
Joseph
J
,
Lopez
M
, et al
Mitochondrial metabolism and ROS generation are essential for Kras-mediated tumorigenicity
.
Proc Natl Acad Sci U S A
2010
;
107
:
8788
93
.
38.
Gronroos
E
,
Terentiev
AA
,
Punga
T
,
Ericsson
J
. 
YY1 inhibits the activation of the p53 tumor suppressor in response to genotoxic stress
.
Proc Natl Acad Sci U S A
2004
;
101
:
12165
70
.
39.
Wang
YP
,
Zhou
LS
,
Zhao
YZ
,
Wang
SW
,
Chen
LL
,
Liu
LX
, et al
Regulation of G6PD acetylation by SIRT2 and KAT9 modulates NADPH homeostasis and cell survival during oxidative stress
.
EMBO J
2014
;
33
:
1304
20
.
40.
Ju
HQ
,
Lu
YX
,
Wu
QN
,
Liu
J
,
Zeng
ZL
,
Mo
HY
, et al
Disrupting G6PD-mediated Redox homeostasis enhances chemosensitivity in colorectal cancer
.
Oncogene
2017
;
36
:
6282
92
.
41.
Wang
Z
,
Liang
S
,
Lian
X
,
Liu
L
,
Zhao
S
,
Xuan
Q
, et al
Identification of proteins responsible for adriamycin resistance in breast cancer cells using proteomics analysis
.
Sci Rep
2015
;
5
:
9301
.
42.
Ahmad
F
,
Dixit
D
,
Sharma
V
,
Kumar
A
,
Joshi
SD
,
Sarkar
C
, et al
Nrf2-driven TERT regulates pentose phosphate pathway in glioblastoma
.
Cell Death Dis
2016
;
7
:
e2213
.
43.
Piazza
GA
,
Rahm
AK
,
Finn
TS
,
Fryer
BH
,
Li
H
,
Stoumen
AL
, et al
Apoptosis primarily accounts for the growth-inhibitory properties of sulindac metabolites and involves a mechanism that is independent of cyclooxygenase inhibition, cell cycle arrest, and p53 induction
.
Cancer Res
1997
;
57
:
2452
9
.
44.
Bible
KC
,
Bible
RH
 Jr
,
Kottke
TJ
,
Svingen
PA
,
Xu
K
,
Pang
YP
, et al
Flavopiridol binds to duplex DNA
.
Cancer Res
2000
;
60
:
2419
28
.
45.
Huang
C
,
Wu
S
,
Ji
H
,
Yan
X
,
Xie
Y
,
Murai
S
, et al
Identification of XBP1-u as a novel regulator of the MDM2/p53 axis using an shRNA library
.
Sci Adv
2017
;
3
:
e1701383
.