Biophysical and biochemical properties of PHGDH revealed by studies on PHGDH inhibitors

Abstract

The rate-limiting serine biogenesis enzyme PHGDH is overexpressed in cancers. Both serine withdrawal and genetic/pharmacological inhibition of PHGDH have demonstrated promising tumor-suppressing activities. However, the enzyme properties of PHGDH are not well understood and the discovery of PHGDH inhibitors is still in its infancy. Here, oridonin was identified from a natural product library as a new PHGDH inhibitor. The crystal structure of PHGDH in complex with oridonin revealed a new allosteric site. The binding of oridonin to this site reduced the activity of the enzyme by relocating R54, a residue involved in substrate binding. Mutagenesis studies showed that PHGDH activity was very sensitive to cysteine mutations, especially those in the substrate binding domain. Conjugation of oridonin and other reported covalent PHGDH inhibitors to these sites will therefore inhibit PHGDH. In addition to being inhibited enzymatically, PHGDH can also be inhibited by protein aggregation and proteasome-mediated degradation. Several tested PHGDH cancer mutants showed altered enzymatic activity, which can be explained by protein structure and stability. Overall, the above studies present new biophysical and biochemical insights into PHGDH and may facilitate the future design of PHGDH inhibitors.

Keywords: Allosteric inhibition; Cancer mutation; Compound screening; Nucleotide-binding domain; Protein degradation.

Fig. 1

Identification of oridonin as a natural product PHGDH inhibitor. A The in vitro inhibitor screening assay. Diaphorase couples the generation of NADH to the generation of fluorescent resorufin. Hydrazine sulfate (N₂H₄ · H₂SO₄) prevents product feedback inhibition of PHGDH by removing 3-phosphohydroxypyruvate (3-PHP). B Inhibition rate against PHGDH and diaphorase by the hit compounds (5 μM, > 90% PHGDH inhibition), from one triplicate experiment. The hits were pre-incubated with diaphorase or PHGDH for 2 h before enzyme activity measurement. C The chemical structure of oridonin and withaferin A. D Isothermal titration analysis. PHGDH (200 μM) was titrated into 150 μM oridonin at 20 °C. E Time-dependent PHGDH inhibition. Samples of 200 nM PHGDH were pre-incubated with 1 μM oridonin for the indicated times, and then PHGDH activity was monitored for 5 h. The time line is shown above. Black arrow indicates the start of the reaction. The data are normalized to the 5 h reading of the ‘0 min’ pre-incubated enzyme. A short fragment of PHGDH (sPHGDH, a.a. 3–314) containing only the NAD-binding domain (NBD) and the substrate-binding domain (SBD) was used in these experiments. Data are expressed as mean ± standard error of mean (SEM), n = 3

Fig. 2

Co-crystal structure of sPHGDH:oridonin reveals a mechanism of PHGDH inhibition by R54 reposition. A Zoomed view of sPHGDH and oridonin interaction. Interacting residues are shown as green sticks. Hydrogen bonds and hydrophobic contacts are denoted as blue and gray dashed lines, respectively. B Relocation of C18 sulfhydryl and R54 guanidine group (yellow dashed lines) upon oridonin binding. Oridonin-bound sPHGDH (green) is superimposed onto the malate-bound sPHGDH (2G76, gray) using the NBD atoms. Hydrogen bonds are shown as blue dotted lines. For clarity, the side chains of S14, A76, and Q292 are not displayed. C Surface and stick representation of the active site around R54 in 2G76 (up) and the oridonin complex (down). D Enzyme activity comparison of WT, C18W, C18Y, and R54A at 100, 200, and 400 nM concentrations. Data are normalized to the reading of the 400 nM WT sample. Statistical differences relative to the WT sample at the same concentration are compared. *p < 0.05 and ***p < 0.001. E Lineweaver–Burk double reciprocal plots for WT, C18W, C18Y, and R54A

Fig. 3

PHGDH is very sensitive to cysteine mutations, especially those in the SBD. A Concentration-dependent enzyme inhibition analysis for C18S. C18S was used at 400 nM concentration. B Enzyme activity of WT and C18S in the absence of oridonin. C Location of oridonin-modified cysteines by mass spectrum analysis, at different concentrations of oridonin (1 μM, 5 μM and 25 μM). MS2 spectrum could not differentiate whether C18 or C19 was conjugated, but their simultaneous conjugation was not detected. D Enzyme activity of cysteine to tryptophan mutants. Statistical differences to the WT sample at the same concentration are compared. **p < 0.01; and ***p < 0.001. E Enzyme activity of the cysteine to serine mutants. F Concentration-dependent enzyme inhibition analysis for WT, C200S, C225S, C254S, and C281S in sPHGDH. Data are expressed as mean ± SEM, n = 3. G Oridonin IC₅₀ analysis under different concentrations of 3-PG or NAD⁺. In this enzyme assay, oridonin was not pre-incubation with sPHGDH

Fig. 4

PHGDH aggregates and precipitates in the presence of inhibitors. A First derivative of the melt curves of sPHGDH with DMSO or oridonin. The right panel shows the mean and standard error of measurements for three independent measurements. *p < 0.05. B Glutaraldehyde (GTA) cross-linking of sPHGDH. sPHGDH was pre-incubated with increasing concentrations of oridonin or DMSO overnight, and then cross-linked with GTA (0.05% v/v) for 2 h. Black arrows indicate cross-linked proteins. C Size exclusion chromatography profile for sPHGDH (20 μM) treated with DMSO- or different inhibitors (200 μM). Samples of sPHGDH were pre-incubated with DMSO or oridonin for 2 h on ice before size exclusion chromatography. D Precipitation of sPHGDH (5 mg/mL) upon binding to multiple oridonin molecules. Samples of sPHGDH were incubated with different molar ratios of oridonin (1:0.5, 1:1, 1:2, 1:4, 1:8, 1:16) at room temperature for 2 h. After centrifugation, the supernatant (S) and the pellet (P) fractions were analyzed by SDS-PAGE

Fig. 5

PHGDH could be inhibited through degradation in MDA-MB-468 cells. A Thermal stability of endogenous oridonin in the presence of DMSO or oridonin (40 μM). The heated samples were centrifuged and the supernatants were analyzed by western bot. B Western blot analysis of intracellular PHGDH level. Cells were incubated with DMSO or various concentrations of oridonin for 48 h. C Endogenous PHGDH level in the presence of oridonin and the proteasome inhibitor MG132. Cells were treated with 10 μM oridonin or DMSO in the presence or absence of MG132 for 8 h. D Intracellular PHGDH level after treatment with different inhibitors for 48 h

Fig. 6

PHGDH cancer mutants display altered enzyme activities. A Distribution diagram of selected PHGDH cancer mutations. B Enzyme activity of cancer mutants in full length PHGDH. Statistical differences relative to the WT sample at the same concentration are compared. ***p < 0.001. C Interaction between R236 (pink) and NADH (yellow), substrate analogue malate, SBD residues T78 and D81, and NBD residue A235. Dotted lines denote hydrogen bonds. D Vmax-normalized velocity (Vt) vs. time for cancer mutants (200 nM). Data are expressed as mean ± SEM, n = 3. E Location of D62 and A66 and their surrounding environment. PHGDH is shown as a cartoon and critical residues are shown as sticks