PLK1-mediated PDHA1 phosphorylation drives metabolic reprogramming in lung cancer

Abstract

Although the involvement of polo-like kinase 1 (PLK1) in metabolic reprogramming from oxidative phosphorylation (OXPHOS) to glycolysis has been previously described, the underlying molecular mechanism remains unclear. Pyruvate dehydrogenase (PDH) catalyzes the conversion of pyruvate into acetyl-CoA, the starting material for the tricarboxylic acid (TCA) cycle. In a companion study by Zhang et al., we demonstrated that PLK1 phosphorylation of PDHA1 at threonine 57 (PDHA1-T57) drives its protein degradation via mitophagy activation. Using a stable-isotope resolved metabolomics (SIRM) approach, we now show that PLK1 phosphorylation of PDHA1-T57 results in metabolic reprogramming from OXPHOS to glycolysis. Notably, cells mimicking PDHA1-T57 phosphorylation rely more on the aspartate-malate shuttle than on glucose-derived pyruvate to sustain the TCA cycle. This metabolic shift was also observed in mouse embryonic fibroblasts (MEFs) and transgenic mice conditionally expressing the PDHA1-T57D variant, highlighting the role of PLK1 in metabolic reprogramming in vivo. It is well-established that pyruvate dehydrogenase kinase (PDK)-mediated phosphorylation of PDH leads to its inactivation and that dichloroacetic acid (DCA), a PDK inhibitor, has been investigated in preclinical and early clinical studies as a potential therapeutic agent for lung cancer. We demonstrated that DCA combined with Onvansertib, a PLK1 inhibitor, synergistically inhibits lung tumor growth by enhancing mitochondrial ROS, inhibiting glycolysis, and inducing apoptosis. This study aims to elucidate how PLK1-associated activity drives the metabolic reprogramming from OXPHOS to glycolysis during cellular transformation, thereby contributing to lung carcinogenesis. Our results provide support for a clinical trial to evaluate the efficacy of Onvansertib plus DCA in treating lung cancer.

Figure 1.

Schematic of stable isotope-resolved metabolomics (SIRM) analysis of CrT cells expressing different PDHA1 variants (T57A or T57D). The cells were cultured with uniformly labeled ¹³C-glucose ([U-¹³C]-glucose), harvested for sample processing, and subjected to metabolite extraction. The analysis was conducted using nuclear maFnetic resonance (NMR) and ion chromatography-Fourier transform mass spectrometry (IC-FTMS). Black circle represents ¹²C, while red circle and blue circles indicate ¹³C contributions from the PDH and PC-derived TCA cycle, respectively. G6P, glucose-6-phosphate; F6P, fructose-6-phosphate; PEP: phosphoenolpyruvic acid; PC, Pyruvate carboxylase.

Figure 2.

PLK1-mediated PDHA1-T57 phosphorylation impacts glucose metabolism. (A) Time courses of ¹³C enrichment in glucose in the culture media. Representative NMR analysis of ¹³C-enriched glucose in the media from CrT cells expressing the PDHA1-T57A or -T57D variant. Media samples were collected at 3, 6, 12, and 24 hours and processed for NMR analysis. (B) Representative NMR analysis of ¹³C-enriched pyruvate in the media. (C) Total amounts of isotopologues in glycolytic metabolites. The x-axis indicates the number of ¹³C atoms in each compound. Total amounts of isotopologues are presented as means ± standard deviations (n = 3). (D) Representative IC-MS analysis of the fractional enrichment of ¹³C-labeled glycolytic metabolites in CrT cells expressing the PDHA1-T57A or PDHA1-T57D variant. The x-axis indicates the number of ¹³C atoms in each compound. Fractional enrichment of isotopologues is presented as means ± standard deviations (n = 3). (E) Total amounts of isotopologues in lactate. *P < 0.05; **P < 0.01; ***P < 0.001 by unpaired t-test.

Figure 3.

Effects of PLK1-mediated PDHA1-T57 phosphorylation on the TCA cycle. (A) Total amounts of isotopologues in citrate, cis-aconitate, and isocitrate in CrT cells expressing the PDHA1-T57A or PDHA1-T57D variant. (B) Fractional enrichment of ¹³C-labeled citrate, cis-aconitate, and isocitrate. (C) Total amounts of the isotopologues in α-ketoglutarate (α-KG), succinate, fumarate, and malate in CrT cells expressing the PDHA1-T57A or PDHA1-T57D variant. (D) Fractional enrichment of ¹³C-labeled α-KG, succinate, fumarate, and malate. (E, F) Total amounts and fractional enrichment of ¹³C-labeled glutamate and aspartate. The x-axis indicates the number of ¹³C atoms in each compound. Total amounts of isotopologues are presented as means ± standard deviations (n = 3), *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 by unpaired t-test.

Figure 4.

PDHA1-T57 phosphorylation shifts the carbon source from glucose to glutamate and aspartate. (A) Citrate m+2 / pyruvate m+3 ratio and citrate m+3 / pyruvate m+3 ratio are shown as surrogates for the activity of PDH and PC, respectively, in cells expressing the PDHA1-T57A or PDHA1-T57D variant. (B) Total fractional enrichments of ¹³C-labeled isotopologues. (C) Heat map showing the total amounts of ¹³C-enriched metabolites associated with glycolysis, the TCA cycle, purine synthesis, pyrimidine synthesis, and amino acid metabolism.

Figure 5.

The PDHA1-T57D mutation alters energy metabolism in the mouse model. (A) Glycolytic rate assay in PDHA1-T57D conditional KI MEFs. MEFs from PDHA1-T57D conditional KI mice were transfected with Lenti-GFP-Control or Lenti-GFP-Cre for 48 hours, followed by glycolytic rate assay. Quantification of metabolic parameters is presented as means ± standard deviations (n = 6). **P < 0.01 by unpaired t-test. Indirect calorimetry was conducted in PDHA1-T57D KI male and female mice. The PDHA1-T57D KI was activated via tamoxifen injection in male mice (T-M, n = 3) and female (T-F, n = 3) mice. Mice without tamoxifen treatment served as control groups (N-M and N-F, n = 3). Oxygen consumption (B), carbon dioxide production (C), energy expenditure (D), and respiratory exchange ratio (E) were monitored for all groups over 84 hours at room temperature. Results were presented as means ± standard deviations, *P < 0.05 by unpaired t-test.

Figure 6.

PLK1 inhibition enhances the chemosensitivity of DCA in vitro. (A) MTT analysis of CrT cells (shCtl or shPLK1) treated with increasing doses of DCA. Results are presented as means ± standard deviations (n = 3). (B) Quantification of mitochondrial ROS levels in CrT cells (shCtl or shPLK1) treated with 2 mM DCA. Results are presented as means ± standard deviations (n=3). **P < 0.01; ns P > 0.05 by unpaired t-test. (C) Representative images of colony formation analysis of CrT cells (shCtl or shPLK1) after 2 mM DCA treatment. (D-F) MTT analysis of CrT(D), H1299(E), PC9(F) cells treated with treated with 5 mM DCA and/or 50 nM Onvansertib. Results are presented as means ± standard deviations (n = 6). ****P < 0.0001, ***P < 0.001**P < 0.01 by two-way ANOVA. (G-I) Representative images of colony formation analysis of CrT(G), H1299(H), PC9(I) cells treated with treated with 2.5 mM DCA and/or 25 nM Onvansertib. Results are presented as means ± standard deviations (n = 6). (J) Immunoblotting analysis of PARP,c-PARP, Caspase3, cleaved Caspase3 protein levels in CrT, H1299 and PC9 cells treated with 5 mM DCA and/or 50 nM Onvansertib. (K) Flow cytometry apoptosis detection using Annexin-V FITC/7AAD staining in Crt cells treated with 5 mM DCA and/or 50 nM Onvansertib. Results are presented as means ± standard deviations (n = 3). ***P < 0.001 by one-way ANOVA. (L) Dose-response map showing the inhibition rate of cell viability in CrT cells treated with DCA + Onvansertib. (M) Synergy matrix plot showing ZIP score for CrT cells treated with DCA + Onvansertib. The average ZIP synergy score is 9.58, with a maximum ZIP synergy score of 22.7.

Figure 7.

PLK1 inhibition enhances the chemosensitivity of DCA through enhancing mitochondrial ROS, inhibiting glycolysis. (A, B) Quantification of mitochondrial ROS levels in CrT cells treated with treated with 5 mM DCA and/or 50 nM Onvansertib(A). CrT cells treated with or without MitoTEMPO (5μM) Results are presented as means ± standard deviations (n=4). * P <0.05 by one-way ANOVA; **P < 0.01 by unpaired t-test. (C) Immunoblotting analysis of PARP,c-PARP, Caspase3, cleaved Caspase3 protein levels in CrT and H1299 cells treated with 5 mM DCA, 50 nM Onvansertib, 5μM MitoTEMPO. (D) Mitochondrial Stress Test of CrT cells. CrT cells treated with 5 mM DCA and/or 50 nM Onvansertib for 16h, followed by Mitochondrial Stress Test. Quantification of metabolic parameters is presented as means ± standard deviations (n=6). *P<0.05; **P<0.01 by , by one-way ANOVA. (E) Glycolytic Rate Assay of CrT cells. CrT cells treated with treated with 5 mM DCA and/or 50 nM Onvansertib, followed by Glycolytic Rate Assay. Quantification of metabolic parameters are presented as means ± standard deviations (n=6). ****P < 0.0001, ***P < 0.001**P < 0.01, *P<0.05, ns. not statistically significant, by one-way ANOVA.

Figure 8.

DCA plus Onvansertib synergistically inhibits tumor growth in vivo. Nude mice were inoculated with CrT cells, randomized into four groups (n = 5 per group), and treated with placebo, DCA, Onvansertib, or a combination for 4 weeks. (A) Tumor volumes over the 4-week treatment period. *P < 0.05 by two-way ANOVA. (B) Representative images of xenograft tumors at harvest. (C) Representative hematoxylin and eosin (H&E) staining of xenograft tumors derived from CrT cells. Red arrows indicate apoptotic cells. Images captured at 20x (scale bars, 200 μm) and 40x magnification (scale bars, 100 μm). (D) Immunofluorescence (IF) staining for cleaved-caspase 3, a marker of apoptosis, reveals increased apoptotic cells in tumors treated with DCA and Onvansertib. Quantification of cleaved-caspase 3 staining is shown on the right. ****P < 0.0001 by one-way ANOVA. Scale bars, 50 μm. (E) IF staining for Ki67 demonstrates reduced proliferation in tumors treated with DCA and Onvansertib. Quantification of Ki67 staining is shown on the right. ****P < 0.0001 by one-way ANOVA. Scale bars, 50 μm.

Figure 9.

High PLK1 and low PDHA1 expression predict poor clinical outcomes in NSCLC. (A, B) Immunohistochemistry (IHC) staining of PLK1 and PDHA1 was performed on a tissue microarray (TMA) containing 216 NSCLC patient samples from a clinical cohort. (A) Protein expression levels in NSCLC samples with well, moderate, and poor differentiation are shown. A statistically significant (*P < 0.05 by one-way ANOVA) increase in PLK1 staining is observed as differentiation decreases. (B) Representative images demonstrate the reciprocal staining patterns of PDHA1 and PLK1 in moderately and poorly differentiated NSCLC samples. PDHA1 expression shows granular cytoplasmic staining, while PLK1 expression shows both nuclear and cytoplasmic staining. (C) Kaplan-Meier 10-year overall survival analysis of NSCLC patients based on PLK1 and PDHA1 expression levels. High PLK1 expression is associated with significantly reduced survival probability (***P < 0. 001), and low PDHA1 expression is similarly associated with poorer outcomes (**P < 0. 01).