Pharmacological inhibition of PGK1 suppresses EGFR-positive esophageal squamous cell carcinoma by dual targeting of glycolysis and autophagy-dependent EGFR degradation

Abstract

Introduction

Esophageal squamous cell carcinoma (ESCC) is an aggressive malignancy with limited treatment options. Phosphoglycerate kinase 1 (PGK1), with both glycolytic and kinase activities, has been implicated in tumor progression, but its therapeutic potential in ESCC remains unclear.

Methods

We assessed PGK1 by genetic knockdown and developed compd 25 − 4, a structure-based small-molecule inhibitor. Biochemical and cellular assays determined its activity against PGK1 and ESCC proliferation. Mechanisms were explored through cellular glycolysis analysis, autophagy assessment, reverse-phase protein array (RPPA) analysis, and signaling pathway characterization.

Results

PGK1 knockdown significantly impaired ESCC cell growth both in vitro and in vivo, supporting its role as a therapeutic target. Compd 25 − 4 inhibited PGK1 glycolytic activity with nanomolar potency (IC₅₀ = 41 nM) and demonstrated > 5-fold selectivity toward EGFR-positive ESCC cells compared to EGFR-negative cells. Beyond glycolysis inhibition, compd 25 − 4 suppressed PGK1 kinase-mediated PRAS40 signaling and induced autophagy-dependent degradation of EGFR. This dual mechanism of action simultaneously disrupted cancer metabolism and EGFR-driven oncogenic signaling, leading to enhanced therapeutic efficacy.

Conclusions

Our findings establish PGK1 as a promising therapeutic target in ESCC and identify compd 25 − 4 as a potent chemical tool for probing PGK1’s dual enzymatic functions. By concurrently blocking glycolysis and promoting autophagy-mediated EGFR degradation, targeting PGK1 provides a novel therapeutic strategy for EGFR-positive ESCC.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13402-026-01215-4.

Introduction

Esophageal cancer (EC) is a malignant gastrointestinal cancer, ranking seventh among the most common malignancies and the sixth leading cause of cancer-related deaths globally [1, 2]. The two primary histological subtypes of esophageal cancer are esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC). Whereas ESCC is mostly found in East Asia, particularly in China and Japan, EAC is more common in Western nations [3, 4]. Overexpression of EGFR occurs in 43% to 97% of ESCC patients and is correlated with lymph node metastasis, overall survival, and pathological tumor staging, implying that EGFR has an important role in ESCC pathogenesis [5, 6] .

Phosphoglycerate kinase 1 (PGK1) is a key enzyme in the glycolytic process, playing a crucial role in cellular energy metabolism by catalyzing the conversion of 1,3-diphosphoglycerate to 3-phosphoglycerate, while simultaneously generating ATP. Numerous studies have shown that PGK1 overexpression or enhanced glycolytic activity is associated with poor prognosis [7–10] and chemoresistance [11–14] in tumors. Beyond its glycolytic function, PGK1 also acts as a protein kinase, significantly contributing to tumor progression. For instance, PGK1 phosphorylates threonine 338 of pyruvate dehydrogenase kinase isoenzyme 1 (PDHK1), thereby limiting mitochondrial pyruvate utilization and enhancing the Warburg effect [15]. Under normoxia, PGK1 promotes liver cancer cell proliferation and inhibits autophagy-mediated cell death by phosphorylating PRAS40 [16], while under hypoxia, it initiates autophagy by phosphorylating Beclin1 [17]. Overall, PGK1’s dual role as both a glycolytic enzyme and a protein kinase makes it a key driver of tumor progression by regulating cellular metabolism and signaling pathways.

Previous research has shown that esophageal carcinoma (ESCC) exhibits significantly higher levels of PGK1 mRNA in tumor tissues compared to normal tissues [18]. Xu et al. reported that the long non-coding RNA NRSN2-AS1, transcribed by SOX2, promotes ESCC cell proliferation, migration, invasion, and epithelial-mesenchymal transition by interacting with PGK1 and inhibiting its ubiquitination and proteasome-mediated degradation [19]. Additionally, Li et al. found that glycolytic metabolism is abnormally increased at both the genomic and proteomic levels during ESCC development, with PGK1 reprogramming glucose metabolism and contributing to ESCC progression [20]. Together, these studies indicate that PGK1 plays a crucial role in ESCC and represents a promising molecular target for therapeutic intervention. However, current research on PGK1 in ESCC primarily focuses on its role in regulating glucose metabolism, while its kinase activity and potential non-metabolic functions in ESCC progression remain to be further explored.

Recently, NG52, which was previously reported as a cell cycle-regulating kinase inhibitor in yeast, was shown to display a sub-micromolar IC₅₀ against PGK1 [21]. To increase the potency of NG52, a new PGK1 inhibitor compd 25 − 4 has been developed. In addition to the known function of PGK1 in glycolytic metabolism, compd 25 − 4 also exhibits strong anti-proliferation efficacy against EGFR-positive ESCC by inhibiting PRAS40 phosphorylation and inducing EGFR degradation via autophagy. Nonetheless, this study reinforces the concept that the kinase activity of PGK1 is crucial for ESCC development, and inhibiting PGK1 may thus offer a novel and effective therapeutic strategy for the treatment of EGFR-positive ESCC.

Materials and methods

Reagents

The PGK1 rabbit monoclonal antibody (mAb) (#ab154613) was obtained from Abcam (Cambridge, MA, USA). EGF receptor (D38B1) rabbit mAb (#4267), p44/42 MAPK (Erk1/2) rabbit mAb (#4695), phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) rabbit mAb (#4370), Phospho-mTOR(Ser2448) rabbit mAb(#2971), mTOR rabbit mAb (#2972), GAPDH rabbit mAb (#5174), and LC3B rabbit mAb (#3868) were obtained from Cell Signaling Technology (Danvers, MA, USA). PRAS40 rabbit mAb (#21097-1-AP), Phospho-PRAS40 (T246) rabbit polyclonal antibody (#29072-1-AP), and Tubulin rabbit mAb (#11224-1-AP) were obtained from Proteintech (Wuhan, China). All antibodies were used at a dilution of 1:1000 in immunoblotting experiments. Recombinant human PGK1 protein (Active) (#ab289789) was obtained from Abcam (Cambridge, MA, USA). PRAS40 Human Recombinant Protein (#TP300919) was purchased from Origene (MD, USA). Cycloheximide (CHX), MG132, chloroquine, and AZD9291 were purchased from MedChemExpress (Shanghai, China).

Cell lines

The ESCC cell lines OE19, OE33, OE21, KYSE450, KYSE30, and KYSE520 were obtained from the American Type Culture Collection (ATCC). All cell lines were cultured in RPMI 1640 medium (Corning, Manassas, VA, USA) supplemented with 10% certified fetal bovine serum (C0400, VivaCell, Shanghai, China) and penicillin/streptomycin/L-glutamine at 37 °C in a humidified incubator with 5% CO₂.

Synthesis of compd 25 − 4

The process for the synthesis of compd 25 − 4 is described in Supporting Information Fig. S1. Intermediate 2 was synthesized through a nucleophilic substitution reaction between 2,6-dichloro-9-(prop-2-yl)purine and 3-amino-5-bromobenzene-1-carbonitrile. Subsequently, Intermediate 2 and trans-4-aminocyclohexan-1-ol underwent the second nucleophilic substitution reaction in the presence of Et₃N to afford the final compd 25 − 4.

In vitro kinase assay

To each tube, 2 µL of PGK1 protein solution (1 µg/µL, pre-diluted in reaction buffer) was added, followed by 2 µL of the diluted compound. The mixture was gently vortexed and incubated at room temperature for 30 min to facilitate protein-compound binding. The enzymatic reaction was initiated by adding 2 µg of PRAS40 substrate and 0.5 mM ATP in reaction buffer (50 mM Tris-HCl, 100 mM KCl, 50 mM MgCl₂, 1 mM Na₃VO₄, 1 mM DTT, 5% glycerol, pH 7.5). The tubes were further incubated at room temperature for 30 min. Reactions were terminated by adding 5× loading buffer, and samples were denatured by boiling at 95 °C for 10 min. Phosphorylated PRAS40 (P-PRAS40) levels were analyzed by Western blotting using specific primary antibodies against P-PRAS40 followed by horseradish peroxidase (HRP)-conjugated secondary antibodies and chemiluminescent detection .

PGK1 glycolytic enzyme activity assay

To test the inhibitory effect of compd 25 − 4 on PGK1 glycolytic enzyme activity, an assay was performed as previously reported [22]. The drug powder was dissolved in DMSO to prepare a 1 mM stock solution, followed by serial threefold dilutions in DMSO, which were further diluted 250-fold with reaction buffer (20 mM Tris, 100 mM NaCl, 0.1 mM MgSO₄, 10 mM Na₂HPO₄, 2 mM DTT, pH 8.6). Then, purified recombinant PGK1 protein (0.2 ng/µL) was mixed with DMSO or different concentrations of compd 25 − 4 for 30 min. After the incubation period, a saturated amount of substrates (1.6 mM GAP, 1 mM β-NAD, 1 mM ADP and 20 ng/µL GAPDH) was added, and the reaction was conducted for 1 h in buffer (20 mM Tris, 100 mM NaCl, 0.1 mM MgSO₄, 10 mM Na₂HPO₄, 2 mM DTT, pH 8.6). NADH can be detected at the absorbance wavelength of 339 nm.

Measurement of glucose concentration and lactate production

Glucose concentration in liquid medium was measured by an assay kit (Applygen Technologies, Beijing, E1010) based on the glucose oxidase (GOD) method. For lactate production measurement, the culture medium from cells (2 × 10⁶) was collected, and the lactate levels were determined using a lactate colorimetric assay kit (#G0816W, Suzhou Grace Biotechnology Co., Ltd, China) following the manufacturer’s protocol.

Cell proliferation assay

Cells were cultured in 96-well culture plates for 12 h before the addition of compounds at various concentrations. Cell proliferation was determined after treatment with compounds for 3 days. Subsequently, Cell Counting-Lite 2.0 (#DD1101-02, Vazyme, Nanjing, China) was added to evaluate the cell viability, and luminescence was measured in a multilabel reader (Envision, PerkinElmer, MA, USA). GI₅₀ values were calculated using Prism 8.0 (GraphPad Software, San Diego, CA, USA).

The cellular thermal shift assay (CETSA)

Cells were treated with either compd 25 − 4 1 µmol/L or DMSO and incubated for 8 h. The samples were then collected, lysed, and transferred to PCR tubes for a 3-minute incubation at various temperatures. Centrifugation was used to isolate the insoluble proteins, and Western blotting was subsequently performed to evaluate the soluble protein content.

Colony formation

Cells were plated in 6-well plates and treated with either compd 25 − 4 or DMSO, and cultured for a specified number of days. For colony visualization, the colonies were fixed with 4% paraformaldehydel for 15 min and then stained with 0.5% Crystal Violet Staining Solution for 15 min.

Plasmid and lentivirus production

The plasmids pLVX-shPGK1, pLenti-AKT1S1, and pLenti-AKT1S1(T246A) were purchased from PPL (Public Protein/Plasmid Library, China). Virus particles were harvested 48 h after the transfection of HEK293T cells with the expression vector together with packaging plasmids psPAX2 and pMD2.G using Lipofectamine 1000 reagent.

Co-immunoprecipitation (Co-IP)

Cell extracts were incubated with anti-PGK1 antibody or control IgG together with protein Protein A/G Magnetic Beads at 4 °C overnight, and immunoprecipitates were then subjected to Western blotting assay.

Real-time RT-PCR

The assay was performed as previously reported [23]. After drug treatment, total RNA was extracted using the SPARKeasy Cell RNA Kit (AC0205, Shandong Sparkjade Biotechnology Co., Ltd.). cDNA samples were prepared by reverse transcription with Hifair^® V One-Step RT-gDNA Digestion SuperMix for qPCR (11142ES60, Yeason, China), and mRNA levels were measured by Hieff^® qPCR SYBR Green Master Mix (11201ES08, Yeason, China).

The PCR Primers used for real-time PCR were as follows:

PGK1: CAAGGTTAAAGCCGAGCCAG; GTCTGCAACTTTAGCTCCGC.

EGFR: AACACCCTGGTCTGGAAGTACG; TCGTTGGACAGCCTTCAAGACC.

GADPH: CCAAGGAGTAAGACCCCTGG; TGGTTGAGCACAGGGTACTT.

RPPA analysis

Reverse phase protein array (RPPA) was conducted by Fynn Biotechnology at the Mills Institute for Personalized Cancer Care (Jinan, Shandong Province, China) to quantitatively assess changes in 306 cancer-related proteins and phosphoproteins in KYSE450 cells following 25 − 4 treatment, followed by KEGG pathway analysis.

Tumor esophageal cancer organoid culture

Human patient samples were obtained with approval from the 901st Hospital of the Joint Logistics Support Force of the People’s Liberation Army, Anhui, China. All studies involving human specimens were performed in accordance with ethical guidelines, and informed consent were obtained for the use of these samples. Tumor samples underwent multiple washes with washing solution (#PRS-TCR-1) (Hefei PreceDo pharmaceuticals Co. Ltd., Anhui, China) before being minced into small pieces using a scalpel and incubated with digestive enzyme (#PRS-TDE-2) (Hefei PreceDo pharmaceuticals Co. Ltd., Anhui, China) for 1–2 h at 37 °C. Following incubation, the mixture was filtered through a 70-µm cell strainer to remove large undigested fragments. The cell suspension was centrifuged at 400×g for 5 min, and the cell pellet was resuspended in PBS. The centrifugation step was repeated twice to remove debris and residual digestive enzyme. Matrigel was thawed and kept on ice, while a 24-well plate was pre-warmed to 37 °C. 2000 cells were plated per well in a mixture of 50 µL 1:1 Matrigel and esophageal carcinoma organoid medium (#PRS-ECM-3D) (Hefei PreceDo pharmaceuticals Co. Ltd., Anhui, China). The droplets were incubated at 37 °C for 30 min, after which 500 µL of esophageal carcinoma organoid medium was added to each well. The medium was changed every three days.

In vivo experiments

Five-week-old female Balb/c nude mice were purchased from GemPharmatech Co., Ltd. (Nanjing, China). All animals were housed in a specific pathogen-free facility and used according to the animal care regulations of the Hefei Institutes of Physical Science, Chinese Academy of Sciences. In vivo studies were approved by the Hefei Institutes of Physical Science Ethics Committee, Chinese Academy of Sciences (IACUC-2023-PD-147). Xenograft tumor models were used to evaluate the efficacy of compd 25 − 4. Tumor volumes were determined by caliper measurement every 2 days and calculated using the formula:

where width (W) is defined as the smaller of the two measurements and length (L) is defined as the larger measurement. Tumor tissues were collected, followed by photography and measurement.

Statistical analysis

Statistical significance was determined based on P values using the appropriate statistical tests. Error bars are presented as means ± standard deviation (SD). Statistical analyses were performed using Prism software (GraphPad 8.0).

Results

PGK1 is a potential therapeutic target for ESCC

PGK1 is upregulated in various types of tumors as a metabolic enzyme involved in the glycolytic pathway. To investigate the association of PGK1 with esophageal cancer, we analyzed the expression of PGK1 in esophageal tumor tissues and its correlation with patient survival. Data from The Cancer Genome Atlas (TCGA) database showed that PGK1 is highly expressed in esophageal cancer, similar to EGFR overexpression (Fig. 1A and B). Additionally, the Gene Expression Profiling Interactive Analysis (GEPIA) database revealed that PGK1 is negatively correlated with the overall survival rate of esophageal cancer patients (Fig. 1C). To further confirm the role of PGK1 in ESCC, we used lentivirus-based small-hairpin RNA (shRNA) to knock down PGK1 in KYSE450 and KYSE520 cells, and observed a significant decrease in cell proliferation and colony formation in both cell lines (Fig. 1D-F). Also, PGK1 knockdown in KYSE450 cells led to a notable decrease in their capacity to form tumors in vivo (Fig. 1G and H). These data suggested that PGK1 is critical for ESCC cell proliferation and tumor development.

Fig. 1.

The role of PGK1 in the tumorigenesis of ESCC. (A) The expression levels of PGK1 in esophageal cancers using The Cancer Genome Atlas (TCGA) data. EGFR expression levels serve as the positive control, while Alpha-Fetoprotein (AFP) expression levels are used as the negative control. (B) Effect of different expression levels of PGK1 between normal esophageal tissues and ESCC tissues using TCGA data. ****P < 0.0001. P values were calculated by Student’s t-test. (C) Effect of different expression levels of PGK1 on survival in esophageal cancer using the Gene Expression Profiling Interactive Analysis (GEPIA) web server. (D) The lentivirus-based small-hairpin RNA (shRNA) was used to knock down PGK1 in KYSE450 and KYSE520 cells. After 48 h, the transfection efficiency was examined by Western blotting. (E) The proliferation of KYSE450 and KYSE520 cells with PGK1 knocked down. The experiment was repeated three times. **p < 0.01; ***p < 0.001; ****p < 0.0001 vs. shSCR Group. ns, not significant. Statistical significance was determined by two-way ANOVA. (F) The colony formation capacities of KYSE450 and KYSE520 cells upon PGK1 knockdown. Quantification is shown in the right panel. ****P < 0.0001 vs. shSCR Group. Statistical significance was determined by one-way ANOVA. (G-H) KYSE450 cells infected with scrambled shRNA or shPGK1 were inoculated into the flanks of nude mice (n = 4 per group). Tumor growth (G) and body weight (H) in each group were measured during the experiment. *P < 0.05; **P < 0.01 vs. shSCR Group. Statistical significance was determined by two-way ANOVA

Compd 25 − 4 is a highly potent PGK1 inhibitor

To further explore PGK1’s role in ESCC, a highly potent PGK1 inhibitor is indispensable. However, the potencies of existing PGK1 inhibitors are limited with only micromolar activities. To further improve the activity of NG52, we took a structure-based drug design strategy and developed a novel PGK1 inhibitor, compd 25 − 4 (Fig. 2A). The glycolytic enzyme activity assay showed that compd 25 − 4 potently inhibited the activity of PGK1 with an IC₅₀ of 41 nM (Fig. 2B). Compared with NG52, compd 25 − 4 exhibits over 243-fold greater inhibitory activity. Given that the parental compound NG52 is a known yeast cell cycle-regulating kinase inhibitor, we assessed the in vitro activity of compd 25 − 4 on cyclin-dependent kinases (CDKs). The results indicated that compd 25 − 4 had weaker activity against the CDK kinases compared to PGK1 (Supporting Information Table S1).

Fig. 2.

Identification of compd 25 − 4 as a novel PGK1 inhibitor. (A) Chemical structures of Compd 25 − 4 and NG52. (B) Compd 25 − 4 IC₅₀ determination with isolated PGK1 protein using the glycolytic enzyme activity Assay. (C) Docking model of the compd 25 − 4/PGK1 complex. Compd 25 − 4 and the surrounding residues of PGK1 were shown as sticks. (D) The effect of compd 25 − 4 on the stability of PGK1 protein in a temperature-dependent manner was investigated using KYSE450 and KYSE520 cell lysates. (E) KYSE450 and KYSE520 cells were transfected with the PGK1 overexpressing vector or empty vector (NC, negative control). The transfection efficiency was examined by Western blotting. (F) Cells with PGK1 overexpression were treated with indicated concentrations of compd 25 − 4 for 72 h. Cell viability was determined by the Cell Titer-Glo Assay. The experiment was repeated three times. ****P < 0.0001. Statistical significance was determined by two-way ANOVA

To better understand the binding mode of compd 25 − 4 with PGK1, we then applied computer-aided structural analysis by docking the compd 25 − 4 onto the molecular structure of PGK1 kinase (PDB ID: 5NP8) using AutoDock. The results showed that the two amino groups of the compound formed hydrogen bonds with Val340 and Gly313, while the cyclohexanol extended into the binding pocket, forming a hydrogen bond with Asp285. Moreover, the phenyl ring formed a Pi-Pi stacking interaction with Phe292, which is crucial for the recognition and binding of the compd 25 − 4 to PGK1 (Fig. 2C). To verify the on-target effect of compd 25 − 4, we conducted the cellular thermal shift assays (CETSA) on KYSE450 and KYSE520 cells to investigate the direct interaction of compd 25 − 4 with PGK1. The results showed that compd 25 − 4 effectively improved the thermal stability of PGK1 protein, providing additional evidence for the direct binding between compd 25 − 4 and PGK1 in cells (Fig. 2D). Additionally, to further explore the on-target effect of compd 25 − 4, we constructed cancer cell lines overexpressing PGK1 and treated them with compd 25 − 4 to investigate the dependence of esophageal cancer cell survival on PGK1. The findings showed that the anti-proliferative effect of compd 25 − 4 is diminished when PGK1 is overexpressed, indicating that the inhibition ability of compd 25 − 4 is dependent on PGK1 targeting (Fig. 2E and F, Supporting Information Fig. S9A). Together, the above studies demonstrated that compd 25 − 4 is a highly potent PGK1 inhibitor.

Compd 25 − 4 more effectively suppresses the growth of EGFR-positive ESCC cells

Next, we investigated the effect of compd 25 − 4 on ESCC cell proliferation in a panel of ESCC cell lines (Supporting Information Table S2). The results demonstrated that compd 25 − 4 effectively inhibited both the cell growth and colony formation in ESCC cancer cell lines (Fig. 3A and B, Supporting Information Fig. S2). Notably, compd 25 − 4 exhibited over fivefold greater potency against EGFR-positive cells compared to EGFR-negative cells. Consistently, we observed a similar anti-proliferation effect of compd 25 − 4 on primary ESCC cells (Fig. 3C and D), in which more significant inhibition of EGFR-positive patient-derived cell (PDC) proliferation was observed with compd 25 − 4 treatment. Given that patient-derived organoids (PDOs) are useful models to mimic the biological characteristics of the primary tumors, we next examined the effects of compd 25 − 4 in primary EGFR-positive ESCC PDO in vitro to investigate the anti-tumor effect of this compound. We observed a reduction in both number and size of organoids formations following compd 25 − 4 treatment (Fig. 3E). In addition, we detected substantial growth inhibition in PDOs treated with compd 25 − 4 (Fig. 3F). To further evaluate the in vivo inhibitory activity of compd 25 − 4, we constructed tumor xenograft models derived from KYSE450 cells. Significant tumor growth inhibition was observed in mice treated with compd 25 − 4 in a dose-dependent manner with no noticeable change in body weight (Fig. 3G, Supporting Information Fig. S3). At a dose of 200 mg/kg, compd 25 − 4 achieved a tumor growth inhibition rate (TGI) of 73.3% (Fig. 3H and I). In conclusion, these findings suggest that compd 25 − 4 inhibits the proliferation of ESCC by targeting PGK1, and that EGFR-positive ESCC cells are more susceptible to the effects of compd 25 − 4.

Fig. 3.

The anti-tumor effect of compd 25 − 4 on EGFR-positive ESCC. (A) Antiproliferative effects of compd 25 − 4 against OE19, OE33, OE21, KYSE450, KYSE30, and KYSE520. Statistical analysis is shown in the right panel. ****p < 0.0001. Statistical significance was determined by Student’s t-test. (B) Immunoblotting analysis of PGK1 and EGFR expression using a panel of esophageal cancer cells. (C) Antiproliferative effects of compd 25 − 4 against primary esophageal cancer cells. The maximum concentration of compd 25 − 4 was set at 10 µmol/L. Statistical analysis is shown in the right panel. ****p < 0.0001. Statistical significance was determined by Student’s t-test. (D) Immunoblotting analysis of PGK1 and EGFR expression using a panel of primary esophageal cancer cells. (E) Effects of compd 25 − 4 treatment on ESCC PDO models. Quantifications are shown in the lower panel. *p < 0.05; **p < 0.01; ****p < 0.0001. ns, not significant. Statistical significance was determined by one-way ANOVA. (F) Antiproliferative effects of compd 25 − 4 on the esophageal cancer organoid. (G) KYSE450-bearing xenograft model was treated with compd 25 − 4 at 50, 100, and 200 mg/kg dosages or vehicle. The tumor sizes were measured after compd 25 − 4 treatment. Each group contained five animals. (H) Representative photographs of tumors in each group after 50, 100, and 200 mg/kg compd 25 − 4 or vehicle treatment. (I) Comparisons of the final tumor weight in each group after the 16-day treatment period. *p < 0.05; **p < 0.01. ns, not significant. Statistical significance was determined by one-way ANOVA

Compd 25 − 4 induces EGFR degradation in EGFR-positive ESCC cells

Tumor cells often rely on glycolysis for energy production, and one of the known functions of PGK1 kinase is glycolysis regulation. Therefore, we explored whether the selective inhibition of compd 25 − 4 on EGFR-positive ESCC cells is mediated through differences in metabolism. We found that compd 25 − 4 similarly inhibited glucose consumption and lactate production in a dose-dependent manner in both of EGFR-positive (KYSE450 and KYSE520) and EGFR-negative cells (OE19 and OE33) (Fig. 4A and B). However, when we supplemented the downstream metabolite of PGK1, 3-PG or phosphoenolpyruvate (PEP), together with compd 25 − 4 to treat ESCC cells, 3-PG and PEP were able to almost completely rescue the anti-proliferation effect of compd 25 − 4 in EGFR-negative cells, but could not fully rescue this effect in EGFR-positive cells (Fig. 4C). These findings suggested that the inhibition on glycolytic metabolism by compd 25 − 4 alone cannot fully explain its growth inhibition effect on EGFR-positive ESCC cells, which may account for the selective inhibition of these cells.

Fig. 4.

The inhibition of glycolytic metabolism by compd 25 − 4 alone cannot fully account for its growth inhibition effect. (A) The effect of compd 25 − 4 on the glucose consumption in KYSE450, KYSE520, OE19, and OE33 cells. *p < 0.05; ***p < 0.001; ****p < 0.0001. ns, not significant. Statistical significance was determined by one-way ANOVA. (B) The effect of compd 25 − 4 on the lactate production in KYSE450, KYSE520, OE19, and OE33 cells. **p < 0.01; ***p < 0.001; ****p < 0.0001. ns, not significant. Statistical significance was determined by one-way ANOVA. (C) Cells were treated with compd 25 − 4 in the presence of 3-PG or PEP for 72 h, and cell viability was determined by the Cell Titer-Glo Assay. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. Statistical significance was determined by one-way ANOVA

To investigate the mechanism underlying the growth inhibition induced by compound 25 − 4 in EGFR-positive ESCC cells, we employed a reverse phase protein array (RPPA) to quantitatively assess changes in 306 cancer-related proteins and phosphoproteins in KYSE450 cells following compd 25 − 4 treatment. KEGG pathway analysis revealed significant enrichment of the ErbB, PI3K-AKT, AMPK, and mTOR signaling pathways, as well as the autophagy pathway (Fig. 5A). Given that EGFR is the upstream regulator of these identified pathways and serves as one of the primary drivers in ESCC pathogenesis, we hypothesized that 25 − 4 likely exerts its effects through modulating EGFR-mediated signaling. Therefore, we examined the EGFR-mediated signaling pathway upon compd 25 − 4 treatment. We found that the protein levels of EGFR in KYSE450 and KYSE520 cells decreased in a dose- and time-dependent manner following compd 25 − 4 treatment (Fig. 5B-C). Since the EGFR-driven pro-survival pathway is crucial for tumor progression, we evaluated the impact of compd 25 − 4 on downstream EGFR signaling pathways and found that compd 25 − 4 significantly downregulated the phosphorylation levels of ERK in a dose-dependent manner (Supporting Information Fig. S4A). To exclude the possibility of a direct effect of compd 25 − 4 on EGFR protein activity, we conducted EGFR kinase activity assay, and the results showed that compd 25 − 4 has no effect on EGFR kinase activity (Fig. S4B). Similarly, in esophageal cancer cells with PGK1 knockdown via shRNA, the protein level of EGFR and the phosphorylation levels of ERK were also reduced (Fig. 5D and E, Fig. S4C). Additionally, to further explore the mechanism of compd 25 − 4, we constructed cancer cell lines with EGFR knockdown and treated them with compd 25 − 4. The findings showed that the anti-proliferative effect of compd 25 − 4 was reduced when EGFR was knocked down (Supporting Information Fig. S5A-B), indicating that the inhibition ability of compd 25 − 4 is dependent on EGFR levels. Altogether, these results demonstrate that compd 25 − 4 can selectively inhibit EGFR-positive ESCC cells by decreasing EGFR protein levels and downregulating EGFR-mediated signaling pathway.

Fig. 5.

Effect of compd 25 − 4 on EGFR degradation in EGFR-positive ESCC. (A) To characterize signaling pathways modulated by compd 25 − 4 in KYSE450 cells, reverse-phase protein array was applied to quantitatively assess changes in 306 cancer-associated proteins and phosphoproteins, followed by KEGG pathway enrichment analysis to identify significantly enriched pathways. (B) KYSE450 and KYSE520 were treated with compd 25 − 4 at 0.3 µmol/L, 1 µmol/L, and 3 µmol/L for 24 h. The EGFR protein levels were measured by Western blotting. (C) KYSE450 and KYSE520 were treated with compd 25 − 4 (1 µmol/L) for the indicated time periods. The EGFR protein levels were measured by Western blotting. (D) The EGFR levels in KYSE450 and KYSE520 cells with PGK1 knocked down were examined by Western blotting. (E) Quantitative results for the corresponding results in Figure D. All experiments were performed in three replicates. *p < 0.05; ***p < 0.001. Statistical significance was determined by one-way ANOVA. (F) The degradation half-time of EGFR after compd 25 − 4 treatment. KYSE450 cells were treated with CHX (40 µg/mL) in the presence or absence of compd 25 − 4 (1 µmol/L) for the indicated time periods. The EGFR protein levels were measured by Western blotting. All experiments were performed in three replicates. Quantifications are shown in the right panel. (G) KYSE450 scramble or PGK1 knockdown cells were treated with cycloheximide (CHX) (40 µg/mL) at indicated intervals and protein stability of EGFR was then analyzed by Western blotting. All experiments were performed in three replicates. Quantifications are shown in the right panel

Compd 25 − 4 reduces EGFR protein levels through the autophagy pathway by targeting PGK1 kinase activity

To investigate how compd 25 − 4 affects EGFR at the protein level, we first investigated whether compd 25 − 4 influences EGFR through transcriptional regulation, which was ruled out by our real-time RT-PCR experiment, in which targeting or downregulating PGK1 did not affect the transcription of EGFR (Supporting Information Fig. S6A-B). We then examined the protein level by assessing the effects of compd 25 − 4 in the presence of the protein synthesis inhibitor cycloheximide (CHX) that blocks nascent protein synthesis. We observed that the EGFR protein levels in KYSE450 and KYSE520 cells decreased upon treatment with compd 25 − 4 compared to the control (Fig. 5F and Fig. S6C), and a similar effect was seen when PGK1 was knocked down in the cells (Fig. 5G and Fig. S6D). Since proteasomal and lysosomal degradation are two key protein degradation pathways, we next explored the mechanism by which compd 25 − 4 induces EGFR degradation through these pathways. We treated cells with either proteasome inhibitor MG132 or autophagy inhibitor chloroquine (CQ), and found that only CQ rescued the protein level of EGFR (Supporting Information Fig. S7A). To further verify the effect of compd 25 − 4 on cellular autophagy, we analyzed the expression of autophagy-related protein LC3-II level by western blotting after compd 25 − 4 treatment. We found that compd 25 − 4 dose-dependently induced an increase in LC3-II expression in KYSE450 and KYSE520 cells (Fig. 6A). To verify the impact on autophagy process, cells were treated with autophagy inhibitor CQ. CQ-induced LC3-II accumulation (autophagy flux) was further increased by compd 25 − 4 treatment (Fig. 6B and Fig. S9B). LC3 puncta formation in mCherry-EGFP-LC3-transfected cells was examined. The average number of the mCherry⁺ EGFP⁺ LC3 puncta (yellow dots) representing autophagosomes in each cell was significantly higher in compd 25-4-treated cells (10.4) compared to control cells (4.2), while the average number of the mCherry⁺ EGFP⁻ LC3 puncta (red-only dots), representing autolysosomes, was also much higher in compd 25-4-treated cells (14.4) than in control cells (9.5). In addition, CQ treatment resulted in more autophagosomes (yellow dots) and autolysosomes (red-only dots) in compd 25-4-treated cells (Fig. 6C), indicating that PGK1 inhibition by compd 25 − 4 promotes autophagosome formation. In addition, LC3-II expression was upregulated when PGK1 was depleted (Fig. 6D). Furthermore, CQ partially rescued the cell growth inhibition caused by compd 25 − 4, indicating that compd 25 − 4 can induce autophagy in esophageal cancer cells (Fig. S7B). Taken together, our results suggest that compd 25 − 4 may degrade EGFR through the autophagy pathway.

Fig. 6.

PGK1 kinase activity inhibition induces EGFR degradation mediated by autophagy pathway. (A) Cells were treated with compd 25 − 4 for 24 h, followed by Western blotting analyses for LC3BI/II levels. The experiment was repeated three times with similar results. The quantification results of the band density LC3-II were labeled below the bands. (B) Cells were treated with compd 25 − 4 in the presence or absence of CQ for 24 h, followed by Western blotting analyses. The experiment was repeated two times with similar results. The quantification results of the band density LC3-II were labeled below the bands. (C) The KYS450 cells transfected with mCherry-EGFP-LC3 were treated with compd 25 − 4 1 µmol/L in the presence or absence of CQ 12.5 µmol/L for 24 h, and representative micrographs for LC3 puncta were shown. Scale bar, 20 μm. The numbers of autophagosomes (yellow dots) and autolysosomes (mCherry dots) per cell were plotted. Quantification of LC3 puncta numbers is shown. n = 20 cells per group. ****p < 0.0001. Statistical significance was determined by two-way ANOVA. (D) LC3BI/II levels were analyzed in PGK1-knockdown KYSE450 and KYSE520 cells compared with control cells by Western blotting. The experiment was repeated three times with similar results. The quantification results of the band density LC3-II were labeled below the bands. (E) Co-IP followed by Western analyses with the indicated antibodies. The experiment was repeated three times with similar results. (F) Kinase assays were performed by incubating PRAS40 with PGK1 in the presence of compd 25 − 4 at concentrations of 0.3 µmol/L and 1 µmol/L, followed by Western blotting analyses using the indicated antibodies. The experiment was repeated three times with similar results. (G) Cells were treated with compd 25 − 4 for 24 h, P-PRAS40 and P-mTOR levels were analyzed by Western blotting. The experiment was repeated three times with similar results. (H) PRAS40 (WT, wild-type) or PRAS40-T246A (Mut, mutant) was overexpressed in scramble or PGK1-knockdown KYSE450 cells, and then Western blotting was used to determine P-PRAS40 and EGFR levels in the cells. The experiment was repeated two times with similar results. (I) KYSE450 and KYSE520 cells with PRAS40 (WT) or PRAS40-T246A (Mut) overexpression were treated with compd 25 − 4 for 24 h, and then Western blotting was used to determine EGFR levels in the cells. The experiment was repeated three times with similar results

Previous studies have reported that PGK1 inhibits autophagy-induced cell death by phosphorylating PRAS40 via its kinase activity in liver cancer cells [16]. And unphosphorylated PRAS40 promotes autophagy by inhibiting mTORC1 autophosphorylation [24–26]. To elucidate the role of PGK1 in autophagy induction in ESCC, co-immunoprecipitation (co-IP) assays were conducted, revealing a direct physical interaction between PGK1 and PRAS40 in KYSE450 cells (Fig. 6E). Subsequent in vitro enzymatic activity assays revealed that PGK1 directly phosphorylates PRAS40, and that compd 25 − 4 effectively inhibits PRAS40 phosphorylation (Fig. 6F and Fig. S9C). Additionally, cellular experiments demonstrated that 25 − 4 effectively suppressed the phosphorylation of PRAS40 and mTOR (Fig. 6G). Consistently, the levels of P-PRAS40 and P-mTOR decreased in KYSE450 and KYSE520 cells following knocking down of PGK1 (Supporting Information Fig. S8A). Furthermore, we investigated the role of PRAS40 in maintaining EGFR protein stability. After PGK1 knockdown, both the phosphorylation of PRAS40 and the protein level of EGFR were significantly reduced compared to control cells. Conversely, overexpression of PRAS40 restored both the phosphorylation of PRAS40 and the protein level of EGFR. However, overexpression of PRAS40 T246A mutant (a phosphorylation-null mutation) did not show any change in the cellular signaling pathway affected by PGK1 (Fig. 6H, Fig. S8B and Fig. S9D). Similar results were observed with compd 25 − 4 treatment, in which only wild-type PRAS40 reversed the inhibitory effect of compd 25 − 4 on EGFR (Fig. 6I and Fig. S9E). These results suggest that inhibition of PGK1 kinase activity promotes EGFR degradation through the autophagy pathway by suppressing PRAS40.

Discussion

PGK1 is the first ATP-generating enzyme in the glycolytic pathway. In addition to regulating glycolytic metabolism, PGK1 also acts as a protein kinase. Studies have shown that PGK1 is overexpressed in many tumors [12, 27]. However, the majority of existing inhibitors against PGK1 show only micromolar activities and are not potent enough for use as tool compounds in biomedical study or as therapeutic agents. In previous work, we discovered that NG52 has inhibitory activity against PGK1 through high-throughput screening. Since then, NG52 had been widely used as a chemical tool to study the function of PGK1 in various diseases [8, 28]. Given that NG52 was originally developed as a CDK inhibitor and its activity against PGK1 is at the micromolar level, we utilized structural optimization to develop compd 25 − 4, a potent and selective PGK1 inhibitor. Compd 25 − 4 significantly enhances PGK1 inhibition at the nanomolar level and achieves selectivity over CDKs. Therefore, compd 25 − 4 represents a novel and improved tool compound for investigating the role of PGK1 in various diseases.

PGK1 is frequently overexpressed in ESCC and is associated with shortened overall survival in patients [20]. PGK1 functions as a protein kinase in mitochondria, using ATP as a phosphate donor to directly phosphorylate threonine 338 of pyruvate dehydrogenase kinase isozyme 1 (PDHK1), thereby limiting mitochondrial pyruvate utilization and increasing the Warburg effect [29]. The localization of PGK1 protein in mitochondria is partially regulated by EGFR kinase activity, as EGFR activation has been reported to induce the mitochondrial translocation of a small fraction of cytoplasmic PGK1. Our research revealed that compd 25 − 4 exhibited stronger inhibition towards EGFR-overexpressing cells compared to those with low EGFR expression, and effectively suppressed the growth of EGFR-positive primary patient cells and ESCC cancer cells. Rescue experiments showed that the anti-proliferation effect of compd 25 − 4 is not entirely mediated through its effects on metabolic pathway, but also through EGFR downregulation via the autophagy pathway, by inhibiting phosphorylation of PRAS40. These results suggest that, as a protein kinase, PGK1 plays a crucial role in maintaining EGFR protein stability in ESCC.

Several limitations of this study should be acknowledged. First, the scarcity of clinical patient samples restricted the number of patient-derived organoid (PDO) models. Second, pharmacokinetic analysis following oral administration of compound 25 − 4 at 10 mg/kg to mice demonstrated a maximum plasma concentration of 939 ± 350 ng/mL, which necessitates further optimization to enhance the compound’s potential for clinical translation. Third, although our findings highlight a critical role for PGK1 in EGFR-positive ESCC, the efficacy of PGK1 inhibition and its underlying mechanisms in other EGFR-positive malignancies remain unexplored. Future investigations addressing these limitations are imperative for advancing compound 25 − 4 towards clinical viability and for elucidating the therapeutic potential of PGK1 in various cancers.

In summary, we discovered in this study that novel PGK1 inhibitor compd 25 − 4 effectively inhibits the proliferation of ESCC cells both in vitro and in vivo. In addition to the known function of PGK1 in glycolytic metabolism, compd 25 − 4 also inhibits the protein kinase activity of PGK1, leading to the suppression of PRAS40 activity and inducing the degradation of EGFR proteins through the autophagy pathway. Our study suggests that targeting PGK1 may serve as a promising treatment approach for EGFR-positive ESCC.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization: H.W., W.W., and Q.L.; Investigation, Methodology, Formal analysis, Validation and Visualization: J.G, Y.C., Z.J., S.Q., J.H., and B.W.; Project administration: H.W. and A.W.; Supervision: W.W., J.L., and Q.L.; Writing - original draft: J.G and H.W.; Writing - review & editing: W.W., J.L., and Q.L. All authors have read and approved the article.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 32171479), the Key Research and Development Program of Anhui Province (Grant No. 2023s07020018), the Major Science and Technology Project of Anhui Province (Grant No. 202303a07020007), and the CASHIPS Director’s Fund (Grant No. BJPY2022A02). We are also grateful for the support of the Youth Innovation Promotion Association of CAS support (No. Y2023125) for H. W.

Data availability

All necessary data is included in this paper and available from the lead corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The in vivo studies were approved by the Hefei Institutes of Physical Science Ethics Committee, Chinese Academy of Sciences (IACUC-2023-PD-147). All animals were housed in a specific pathogen-free facility and used in accordance with the animal care regulations of the Hefei Institutes of Physical Science, Chinese Academy of Sciences. Human patient samples were obtained with approval from the 901st Hospital of the Joint Logistics Support Force of the People’s Liberation Army, Anhui, China (Approval Number: LY2023-01). All studies involving human specimens were performed in accordance with ethical guidelines, and informed consent was obtained for the use of these samples.

Competing interests

The authors declare no competing interests.