PLK1-mediated phosphorylation of PHGDH reprograms serine metabolism in advanced prostate cancer

Abstract

Metabolic reprogramming is a hallmark of cancer, enabling tumor cells to meet their increased biosynthetic and energetic demands. While cells possess the capacity for de novo serine biosynthesis, most transformed cancer cells heavily depend on exogenous serine uptake to sustain their growth, yet the regulatory mechanisms driving this metabolic dependency remain poorly understood. Here, we uncover a novel mechanism by which Polo-like kinase 1 (PLK1), often overexpressed in prostate cancer, orchestrates a metabolic shift in serine and lipid metabolism through the phosphorylation of phosphoglycerate dehydrogenase (PHGDH), the rate-limiting enzyme of the serine synthesis pathway (SSP). We demonstrate that PLK1 phosphorylates PHGDH at three specific sites (S512, S513, S517), leading to a marked reduction in its protein level and enzymatic activity. This downregulation of SSP forces cancer cells to increase their reliance on exogenous serine uptake via the ASCT2 transporter, which, in turn, fuels the biosynthesis of lipids, including sphingolipids essential for tumor growth and survival. Targeting the SSP, serine uptake, or downstream lipid biosynthetic pathways may offer promising therapeutic avenues in PLK1-high advanced cancers.

Keywords: PHGDH; PLK1; metabolism; serine; sphingolipids.

Figure 1.. PLK1 promotes sphingolipid metabolism.

(A). The ANOVA analysis of PLK1 TMA score and representative images of PLK1 expression in the benign, low grade, intermediate/high grade prostate cancer. Scale bars, 1mm and 100 μm. (B). The PLK1 expression in the prostate cancer cell lines. (C). Elemental molecular formula (EMF) EMF-EMF correlation heatmap. (D). EMF log2 fold-change in PLK1 over Control plotted by M/Z. Points are colored by their voted lipid Category, with “multiple”, “not_lipid” and “not_categorized” removed. (E). Binomial test statistics for each lipid category. (F). Schematic diagram showing the pathways related to sphingolipid metabolism. (G). Immunoblotting of enzymes involved in de novo serine synthesis, serine-glycine transition, de novo lipid synthesis, and sphingolipid metabolism in control and PLK1 overexpressed LNCaP cells.

Figure 2.. PLK1 downregulates PHGDH in cancer.

(A). The expression of enzymes from de novo serine synthesis in prostate cancer cell lines. (B). The expression of enzymes from de novo serine synthesis in prostate cancer PDXs (LuCaP35CR-clone1, LuCaP35CR-clone2, LuCaP77CR, LuCaP145.1). (C). The ANOVA analysis for PLK1 TMA score and representative images of PHGDH expression in the benign, low grade, intermediate/high grade prostate cancer. Scale bars, 100 μm. (D). IHC analysis of PLK1 and PHGDH for prostate tissues from mice with indicated phenotypes. Scale bars, 20 μm. (E). PHGDH level is decreased with the increasing dose of PLK1 expression. GFP-PHGDH and HA-PLK1 co-transfected 293T cells were subjected to IB. (F). PLK1 overexpression (OE) decreases the level of PHGDH. LNCaP, C4-2 and 22Rv1 cells (WT or PLK1-OE) were subjected to immunoblotting (IB). (G). Knockdown (KD) of PLK1 increases the level of PHGDH. C4-2 and Du145 cells (control (scr) and PLK1-KD (sh1/2)) were subjected to IB. (H). Downregulation of PLK1 increases the level of PHGDH. Tet-inducible PLK1 KD C4-2 cells by vehicle (Veh) and 5 nM doxycycline (Dox) were subjected to IB. (I). Downregulation of PLK1 increases the level of PHGDH. Tet-inducible PLK1 KD Du145 cells by doxycycline were subjected to IB. (J). PLK1 overexpression (OE) decreases the level of PHGDH. A549 cells (WT or PLK1-OE) were subjected to immunoblotting (IB). (K). Downregulation of PLK1 increases the level of PHGDH and constitutively active PLK1 (T210D) overexpression (TD) decreases the level of PHGDH. U2OS cells (empty vector (EV), PLK1 knocking down (KD) or PLK1-TD) were subjected to immunoblotting (IB).

Figure 3.. PLK1 phosphorylation of PHGDH leads to its degradation and decrease of enzymatic activity.

(A). The binding between PLK1 and PHGDH. Different forms of Flag-tagged PLK1 (WT, T210D, K82M) were co-transfected with PHGDH plasmids in HEK293T cells. The whole cell lysates (WCL) were used for immunoprecipitation (IP) with the M2-Flag beads, followed by western blotting (WB) with targeted antibodies. (B). Kinase assay for purified recombinant WT PHGDH incubated with PLK1, PLK1 as well as PLK1 inhibitor (Onvansertib), and CDK1. (C). Kinase assay for purified recombinant PHGDH (1–300aa) fragments. (D). Kinase assay for purified recombinant PHGDH single A mutants. (E). The kinase assay for PHGDH. Purified His-tagged WT and double/triple A mutants PHGDH were incubated with recombinant PLK1 in the presence of [γ−³²P]-ATP, followed by SDS-PAGE and film exposure. (F). The consensus amino acid sequencing of PHGDH peptides among species. (G). The protein degradation of PHGDH and mutants. Different forms of PHGDH (WT, 3D, 3A) plasmids were transfected into Hela cells, followed by the treatment of 200 μg/ml CHX. The WCL were examined by PHGDH, and bands were quantified using Image J. (H). The enzyme activity assay for PHGDH. Flag-tagged PLK1 (WT, TD, KM) plasmids were co-transfected with PHGDH (WT, 3A) plasmids. The enzyme activity was measured using a PHGDH enzyme activity assay kit and normalized by the WB band intensity calculation via Image J. Statistical analysis was performed using unpaired two-tailed t-tests. Data are shown as means ± s.e.m. *p<0.05. (I). Phase transition of PHGDH protein. GFP-tagged PHGDH (WT, 3D) plasmids were transfected into HEK293T cells. 2.5% 1,6-hexanediol was added into the WT PHGDH and PHGDH-3D plasmids transfected HEK293T cells for 10 min. The cells were imaged under a fluorescence microscope before and after the treatment of 1,6-hexanediol. Bar=50 μm.

Figure 4.. PLK1 promotes exogenous serine uptake via ASCT2.

(A). PLK1 overexpression increases serine uptake. The same number (5×10⁵) of LNCaP-Tet-CTL, LNCaP-Tet-PLK1-WT, LNCaP-Tet-PLK1-T210D, and LNCaP-Tet-PLK1-WT cells treated by 5 μM ASCT2 inhibitor V-9302 were seeded onto 6-well plates. Cells were washed and incubated with [¹⁴C]-serine upon PLK1 expression was induced by 4 nM doxycycline for 24h. Statistical analysis was performed using unpaired two-tailed t-tests. (B). ACST1 and ASCT2 expression after PLK1 overexpression in LNCaP cells. (C). ACST1 and ASCT2 expression among prostate cancer cell lines. (D). ASCT2 expression after PLK1 overexpression in LNCaP, C4-2 and 22Rv1 cells. (E). ASCT2 expression after doxycycline induced PLK1 knocking down in DU145-Tet-shPLK1 cells. (F). ASCT2 expression in PC3 and DU145 cells after the treatment of PLK1 inhibitor Onvansertib. (G). The expression of ASCT2 and enzymes involved in SSP upon the absence of serine for 24h and 48h in LNCaP-control and LNCaP-PLK1 cells. (H). ASCT2 expression in the PHGDH knockdown and PHGDH restored C4-2 cells.

Figure 5.. PHGDH phosphorylation by PLK1 causes metabolomic reprogramming.

(A). The comparison of labeling glycolytic intermediates using [U]-¹³C-glucose between LNCaP-PLK1 vs LNCaP-Control cells (right) and C4-2-PHGDH-3D vs C4-2-PHGDH-WT cells (left). (B). The comparison of ¹³C incorporation into Inosine monophosphate (IMP) from [U]-¹³C-glucose between LNCaP-PLK1 vs LNCaP-Control cells and C4-2-PHGDH-3D vs C4-2-PHGDH-WT cells. (C). The comparison of glycerol 3-phosphate labelling from [U]-¹³C-glucose between LNCaP-PLK1 vs LNCaP-Control cells and C4-2-PHGDH-3D vs C4-2-PHGDH-WT cells. (D). The 1D ¹H(¹³C)-HSQC NMR analysis for the glucose-derived glycine production for C4-2-3A and C4-2-WT cells. (E). The deuterium (²H) NMR analysis for D-serine (D-Ser) from the media of C4-2-3A and C4-2-WT cells. Statistical analysis was performed using unpaired two-tailed t-tests (A-E). Data are shown as means ± s.e.m. * p<0.05, ** p<0.01, *** p<0.001.

Figure 6.. PLK1 phosphorylation of PHGDH affects serine incorporation into lipids and lipid synthesis in vivo.

(A). The flow chart of ²H from [2,3,3-²H]-serine incorporation into fatty acids. (B). The [2,3,3-²H]-serine incorporation into lipid species in C4-2-PHGDH-WT cells and C4-2-PHGDH-3A cells. Fraction= ²H labeled specie/total specie. (C). The fat% ratio (fat/weight of 2D mice/fat/weight of WT litter mice) over the time after tamoxifen injection (n>5). (D). Representative images of the dissection of fat tissues from paired WT and 2D mice which are from the same litter. (E). Representative images of H&E staining for the liver from paired WT and 2D mice which are from the same litter.

Figure 7.. Therapeutic strategies for advanced prostate cancer.

(A). Colony formation for LNCaP, C4-2R and 22Rv1 cells after treatment by PLK1 inhibitor Onvansertib, PHGDH inhibitor NCT-503 and the combination of BI6727 and NCT-503. (B). Cell viability assay for C4-2-control/C4-2-PLK1 and 22Rv1-control/22Rv1-PLK1 with the treatment of different concentrations of ASCT2 inhibitor V-9302 for 24h. Statistical analysis was performed using two-way ANOVA. (C). Colony formation and quantification for the 22Rv1-control and 22Rv1-PLK1 cells after treatment by ASCT2 inhibitor V-9302. Statistical analysis was performed using two-way ANOVA. (D). Colony formation and quantification for the C4-2-3A and C4-2-3D cells after the deprivation of serine, glycine, serine and glycine. Statistical analysis was performed using two-way ANOVA. (E). The cell viability of 22Rv1-control and 22Rv1-PLK1 after treatment by SPHK2 inhibitor ABC294640. Statistical analysis was performed using two-way ANOVA. (F). Xenograft images and quantification of 22Rv1 and N2P1 with or without treatment of NCT-503. Statistical analysis was performed using unpaired two-tailed t-tests. (G). IHC analysis of PLK1, AR and cleaved caspase-3 for the xenograft of 22Rv1 and N2P1 with or without treatment of NCT-503. Scale bars, 100 μm. (H). Schematic diagram of metabolic reprogramming by PLK1 in prostate cancer. (I). Schematic diagram of therapeutic strategies Data are shown as means ± s.e.m. * p<0.05, ** p<0.01, n.s., not significant.