p53 coordinates DNA repair with nucleotide synthesis by suppressing PFKFB3 expression and promoting the pentose phosphate pathway

Abstract

Activation of p53 in response to DNA damage is essential for tumor suppression. Although previous studies have emphasized the importance of p53-dependent cell cycle arrest and apoptosis for tumor suppression, recent studies have suggested that other areas of p53 regulation, such as metabolism and DNA damage repair (DDR), are also essential for p53-dependent tumor suppression. However, the intrinsic connections between p53-mediated DDR and metabolic regulation remain incompletely understood. Here, we present data suggesting that p53 promotes nucleotide biosynthesis in response to DNA damage by repressing the expression of the phosphofructokinase-2 (PFK2) isoform 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 (PFKFB3), a rate-limiting enzyme that promotes glycolysis. PFKFB3 suppression increases the flux of glucose through the pentose phosphate pathway (PPP) to increase nucleotide production, which results in more efficient DNA damage repair and increased cell survival. Interestingly, although p53-mediated suppression of PFKFB3 could increase the two major PPP products, NADPH and nucleotides, only nucleotide production was essential to promote DDR. By identifying the novel p53 target PFKFB3, we report an important mechanistic connection between p53-regulated metabolism and DDR, both of which play crucial roles in tumor suppression.

Figure 1. PFKFB3 is a p53 suppression target.

(A) p53^−/−, Mdm2^+/+;p53^ER/−, Mdm2^−/−;p53^ER/−, and WT MEF cells were treated (+/−) 100 nM 4-OHT for 24 h prior to lysis and immunoblotting. Densitometry analysis using Image J software was used to compare PFKFB3 expression after normalization to actin for 3 independent experiments. The PFKFB3 expression level observed in the (-)4OHT sample for each cell line was set to 1 for normalization. (B) Relative mRNA levels of Pfkfb3 were determined by qRT-PCR in p53^−/−, Mdm2^+/+;p53^ER/−, Mdm2^−/−;p53^ER/−and WT MEF cells 12 h after treatment with 100 nM 4-OHT. Levels were normalized to actin, and the levels of Pfkfb3 detected in the vehicle-treated samples for each cell line was set to 1 for normalization. (WT-ER p = 0.029 ^−/− ER p = 0.003 n = 3 for each sample) (C) WT and p53^−/− MEF cells were treated (+/−) 10 μM nutlin or 10 μM 5-FU for 24 h prior to immunoblotting for the indicated proteins. (D) Relative mRNA levels of Pfkfb3 were determined by qRT-PCR in p53^−/− and WT MEF cells 12 h after treatment with DMSO, 10 μM nutlin or 10 μM 5-FU. Levels were normalized to actin, and the levels of Pfkfb3 observed in the DMSO-treated samples for each cell line were set to 1. (WT-Nutlin p = 0.016;WT-5FU p = 0.001 n = 3 for each sample) (E) WT and p53^−/− MEF cells were treated with 30 J/m² UV for 24  h followed by immunoblotting for protein expression. (F) Relative mRNA levels of Pfkfb3, p21, and Mdm2 were determined by qRT-PCR in WT and p53^−/− MEF cells 24  h after treatment with 40 J/m² UV. Levels were normalized to actin, and the relative levels of each mRNA detected in the untreated cells was set to 1. (p = 0.0376 n = 6) G) Chromatin immunoprecipitation of the putative Pfkfb3 p53RE located within intron 1 and the p21 p53RE in U2OS cells 12 h after 10 μM nutlin treatment. (PFKFB3 p = 0.002; p21 p < 0.001 n = 3) (H) Exogenous overexpression of Pfkfb3 Intron 1 p53RE-luciferase (Pfkfb3) and mutant Pfkfb3 Intron 1 p53RE-luciferase (Pfkfb3m) in H1299 cells. Relative luminescence is the increase in luminescence signal compared with the vector control. (p = 0.0002 n = 3).

Figure 2. PFKFB3 down-regulation inhibits glycolysis.

(A) Fructose-(2,6)-bisphosphate (F2,6BP) levels were determined in WT, Mdm2^+/+;p53^ER/− and Mdm2^−/−;p53^ER/− MEF cells in the presence or absence of 4-OHT after 24 h. The amount of F2,6BP detected in the vehicle-treated WT MEF cells was set at 100%. (WT ER p = 0.0218; ^−/− ER p = 0.0006 n = 5) (B) Fructose-(2,6)-bisphosphate (F2,6BP) levels were assayed in WT and p53^−/− MEF cells 24 h after treatment with 40 J/m² UV, and the amount of F2,6BP detected in the untreated samples for each cell line was set to 100%. (WT p < 0.0001; ^−/− p = 0.0013 n = 5) (C) Lactate levels were measured in WT and p53^−/− MEF cells 24  h after treatment with 0 J/m², 25 J/m², or 40 J/m² UV, and the amount of lactate detected in the untreated samples for each cell line was set to 100%. (WT 0/25J p < 0.0001; WT 0/40J p < 0.0001; ^−/− 0/25J p = 0.0308 n = 3) (D) Expression levels of PFKFB3 were determined by western blot for p53^−/− MEF cells stably infected with 4 different shRNA constructs targeting Pfkfb3. (E) Extracellular lactate levels were measured in p53^−/− MEF cells transduced with 3 unique shRNA constructs specifically targeting Pfkfb3 (sh-PFKFB3) and compared with cells transduced with the non-specific scrambled control. The non-specific scrambled control samples were designated as 100% (#1 p < 0.0001; #2 p < 0.0001, #3 p < 0.0001 n = 3). (F) Lactate levels were measured 24 h after exogenous overexpression of PFKFB3 in U2OS cells, and the lactate level detected in the control cells was set to 100%. (p = 0.022 n = 3).

Figure 3. PFKFB3 down-regulation facilitates DNA damage repair and survival.

(A) si-NS and si-PFKFB3 pretreated U2OS cells were treated with 15 J/m² UV for varying amounts of time prior to immunoblotting for protein expression. (B) U2OS cells infected with lentiviral particles expressing shRNA constructs specific for PFKFB3 (sh-PFKFB3) and non-specific scrambled control (sh-NS) were treated with caspase inhibitor QVD-OPh for 30 minutes prior to treatment with 25 J/m² UV to prevent apoptosis. Fresh medium containing QVD-OPh was added after treatment, and the cells were incubated at 37 °C for 48 hours prior to fixation and staining for the DNA damage marker γ-H2AX along with DAPI to visualize the total number of nuclei present. (p < 0.0001, n = 6) (C) Mdm2^+/+;p53^ER/− MEF cells stably infected with lentiviral particles harboring PFKFB3-GFP or GFP constructs were treated with caspase inhibitor QVD-OPh for 30 minutes prior to treatment with UV 10 J/m² to prevent apoptosis. Fresh medium containing QVD-OPh was added after treatment, and the cells were incubated at 37 °C for 48 hours prior to fixation and staining for the DNA damage marker γ-H2AX along with DAPI to visualize the total number of nuclei present. (p < 0.0001 n = 5) (D) Mdm2^+/+;p53^ER/− MEF cells treated (+/−) 4-OHT were processed and analyzed as in panel C. (p = 0.0041  n = 7) (E) Mdm2^+/+;p53^ER/− MEF cells stably infected with lentiviral particles harboring PFKFB3-GFP or GFP constructs were treated with 4-OHT before processing and analyzing as in panel C. (p < 0.0001  n = 5) (F) U2OS cells infected with lentiviral particles expressing shRNA constructs specific for PFKFB3 (sh-PFKFB3) and non-specific scrambled control (sh-NS) were irradiated with UV 40  J/m². Twenty-four hours after treatment, the cells were trypsinized and counted using a Bio-Rad TC20 automated cell counter (p = 0.0036  n = 3) (G) U2OS cells overexpressing PFKFB3 by adenoviral infection were processed and analyzed as in panel F (p = 0.0033  n = 3).

Figure 4. UV irradiation induces p53-dependent down-regulation of PFKFB3 and up-regulation of nucleotide levels.

(A) WT and p53^−/− MEF cells were treated with the pyrimidine synthesis inhibitor leflunomide (25 μM) 24 h prior to immunoblotting for protein expression. (B) WT MEF cells were treated with UV 15 J/m² and incubated for 24 h (+/−) 0.2 mM nucleoside supplementation prior to immunoblotting for protein expression. (C) WT MEF cells were treated with 10 μM 5-FU and incubated for 24 h (+/−) 0.2  mM nucleoside supplementation prior to immunoblotting for protein expression. (D) WT MEF cells were treated with UV 20 J/m² and incubated for 12 h prior to methanol extraction of nucleotides. Extracts were analyzed by LC-MS for relative nucleotide levels between UV-treated and untreated control cells, and untreated cell nucleotide levels were normalized to 100% for each experiment (Error bars represent the SEM n = 4). (E) p53^−/− MEF cells were treated with UV 20 J/m² and incubated for 12 h prior to methanol extraction of nucleotides. Extracts were analyzed by LC-MS for relative nucleotide levels between UV-treated and untreated control cells, and the untreated cell nucleotide levels were normalized to 100% for each experiment (Error bars represent the SEM n = 4).

Figure 5. Inhibition of PFKFB3 expression augments PPP-dependent nucleotide production.

(A) NADPH/NADP⁺ ratios were determined in p53^−/− MEF cell lines stably expressing scrambled or sh-PFKFB3 constructs. (#1 p = 0.007 #2 p = 0.0341 n = 3 for each sample) (B) U2OS cells stably expressing scrambled or sh-PFKFB3 constructs were treated with QVD-OPh and (+/−) the PPP inhibitor DHEA (0.25 mM) for 30 minutes prior to 20 J/m² UV treatment. After 48 h, the cells were fixed and stained for γ-H2AX and DAPI. (p = 0.0044 n = 3) (C) U2OS cells stably expressing scrambled or sh-PFKFB3 lentiviral constructs were treated with 40 J/m² UV (+/−) DHEA (0.25 mM) for 24  h, after which the cells were trypsinized and counted using a Bio-Rad TC20 automated cell counter. (p = 0.0002 n = 3) (D) U2OS cells were treated with QVD-OPh and an equimolar nucleoside mixture or NADPH (0 mM, 0.1 mM, or 0.3 mM) for 30 minutes. Cells were then treated with 40 J/m²UV 48 h prior to fixation and then were stained for γ-H2AX and DAPI. (0.1 mM p = 0.0016;0.3 mM p < 0.001 n = 3) (E) U2OS cells were treated with a 0 mM, 0.1 mM, or 0.3 mM equimolar nucleoside mixture for 30 minutes, after which the cells were treated with UV 40 J/m². Surviving cells were imaged and counted 24 h after UV treatment. (0.1 mM p = 0.0185;0.3 mM p < 0.0001 n = 5) (F) Nucleotide abundance was assessed by HPLC in p53^−/− MEF cells stably expressing scrambled or sh-PFKFB3 lentiviral constructs. (p = 0.01 n = 3) (G) U2OS cells stably expressing scrambled or sh-PFKFB3 constructs were treated (+/−) 0.2 mM equimolar nucleoside mixture for 30 minutes. After nucleoside treatment, cells were exposed to 40 J/m² UV for 24 h, after which the surviving cells were counted by microscopy. (p = 0.0001 n = 5) (H) Nucleotide abundance was assessed by HPLC in p53^−/− MEF cells stably expressing GFP or PFKFB3-GFP. (p = 0.0242 n = 3) (I) HCT116 cells stably expressing GFP or PFKFB3-GFP were treated with QVD-OPh (+/−) 0.2 mM equimolar nucleoside mixture for 30 minutes. Cells were treated with 25 J/m² UV 48 h prior to fixation and then were stained for γ-H2AX and DAPI (Vector vs. PFK p = 0.0002; PFK vs. PFK NT p = 0.0031 n = 5).

Figure 6. Model showing that p53 suppresses PFKFB3 expression, which results in increased de novo nucleotide production via the PPP to facilitate DNA damage repair and survival.

Model representation of the role of p53 in the concomitant regulation of glycolysis, the PPP, and nucleotide production through PFKFB3 as well as through the regulation of DNA damage repair target genes to promote DNA damage repair.