PHGDH knockdown activates autophagic flux to suppress migration and invasion of gastric cancer cells

Introduction

Globally, gastric cancer (GC) represents a highly prevalent malignancy of the gastrointestinal tract and persists as a significant driver of cancer mortality across diverse socioeconomic settings (1,2). As a major contributor to global cancer deaths, GC demonstrates particularly high incidence rates in East Asia and Latin America, posing a persistent threat to public health (3).

Despite all the advancements in treatment modalities such as surgical resection, cytotoxic drugs, molecularly targeted agents, and immunotherapy, GC patients have a poor prognosis. This is primarily because GC is often asymptomatic in its early stage, and even though there have been significant advances in professional development and training programs, the implementation of screening programs are variable; for example, while there are screening programs with proven success, such as national screening programs in Japan with high participation rates, adherence is variable around the world (4). Detection of GC using the current biomarkers is poor, and has low sensitivity and specificity, which necessitates effective molecular targets to easily detect and treat GC (5,6). Therefore, understanding the significant molecular determinants of GC pathogenesis and progression will one day ultimately improve patient care.

Metabolic reprogramming has been recognized as one of the hallmark features of cancer biology, particularly in tumors with high proliferative capacity and those adapting to hypoxic and oxidative stress conditions (7,8). To sustain rapid growth and survival, cancer cells frequently upregulate specific metabolic pathways to supply essential biomolecules, such as nucleotides, lipids, amino acids, and antioxidants (9,10). Among these pathways, the serine synthesis pathway (SSP), a key glycolytic branch, plays a pivotal role not only in generating serine and glycine but also in fueling glutathione (GSH) biosynthesis, purine metabolism, and the one-carbon cycle (11,12). As the flux-controlling enzyme in the serine biosynthesis pathway, phosphoglycerate dehydrogenase (PHGDH) mediates the oxidation of 3-phosphoglycerate to generate 3-phosphohydroxypyruvate. This reaction constitutes the committed step for de novo serine production.

In recent years, PHGDH has been shown to play a significant role in stimulating glutamine metabolism via the SSP for serine production; inhibition of PHGDH by PKCζ abrogates cancer cell adaptation to nutrient stress (13). PHGDH is aberrantly upregulated in a range of solid tumors, including breast (14), melanoma (15), non-small cell lung (16), and colorectal cancers (17), with higher expression levels linked to further disease progression, metastatic disease, and poor prognosis. In addition to its Metabolic role, PHGDH may possess other pro-malignant properties, which may enhance both oxidative stress and autophagy regulation-related malignancy (18,19).

Nevertheless, the relationship between PHGDH, oxidative stress, and autophagy within the context of GC has not been well described. We present this article in accordance with the MDAR reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1817/rc).

Methods

Materials and reagents

The primary antibodies against PHGDH (14719-1-AP) were sourced from Proteintech (Wuhan, China). The LC3B antibody (NB100-2220) was obtained from Novus Biologicals (Centennial, Colorado, USA). Antibodies against p62 (8025), mTOR (2983), phospho-mTOR (Ser2448, 5536), ULK1 (8504), HIF-1α (14179), Atg13 (13273), FIP200 (12436), Atg101 (13492), phospho-ULK1 (Ser757, 14202), phospho-ULK1 (Ser555, 5869), and GAPDH (2118) were all supplied by Cell Signaling Technology (CST, Danvers, Massachusetts, USA). The HRP-conjugated secondary antibody (A0208) was obtained from Beyotime Biotechnology (Shanghai, China). PHGDH and LC3B antibodies were diluted at 1:1500. All other primary antibodies were used at 1:1,000. The enhanced chemiluminescence (ECL) detection kit (PK10003) was from Proteintech (Wuhan, China). For protein isolation, RIPA buffer supplemented with protease inhibitors (PMSF and cocktail) and phosphatase inhibitor MA0171 (Meilunbio, Dalian, China) was utilized. Protein quantification was performed with the Enhanced BCA Assay Kit (P0010, Beyotime Biotechnology, Shanghai, China). Autophagy inhibitors, including 3-MA (S2767) and CQ (S4157), were purchased from Selleck Chemicals (Houston, Texas, USA). Polyvinylidene fluoride (PVDF) membranes were purchased from Millipore (IPVH00010, Burlington, Massachusetts, USA).

Immunohistochemistry (IHC)

Tissue samples of gastric carcinoma and adjacent non-cancerous mucosa (n=7 pairs) were prospectively obtained during the surgical resection at The First Affiliated Hospital of Henan University. All samples were collected from gastrectomy specimens prior to any chemotherapy or radiotherapy, and no patients received neoadjuvant treatment. Clinicopathological data, including age, sex, tumor stage (AJCC 8th edition), histological grade, and treatment history, are detailed in Table S1. All specimens underwent standardized processing: initial fixation in 10% neutral formalin for 24–48 hours, and the samples were subsequently embedded in paraffin and sliced into 5-µm sections using a microtome. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Basic Medical Sciences, Henan University (No. HUSOM2020-027), and all participants provided written informed consent.

Before immunohistochemical staining, tissue sections were baked at 60 ℃ for 30 minutes, deparaffinized in xylene, and rehydrated through graded ethanol solutions. Heat-induced antigen retrieval was achieved in citrate buffer at 120 ℃ for 20 minutes. To block endogenous peroxidase activity, tissue sections were treated with 3% H₂O₂, followed by a 30-minute blocking step using 5% bovine serum albumin. The sections were then probed with anti-PHGDH primary antibody (1:200) at 4 ℃ overnight, washed, and exposed to secondary antibody (1:5,000) at ambient temperature for 2 hours. Chromogenic development employed DAB substrate with hematoxylin counterstaining. Positive immunoreactivity was defined as brown-yellow deposition localized to cytoplasmic or nuclear compartments. For quality control, normal gastric mucosa served as the negative control, and human liver tissue with known PHGDH expression served as the positive control.

Cell lines and culture conditions

Four human gastric adenocarcinoma cell models—AGS (CRL-1739), MKN28 (TKG-0189), SGC-7901 (TKG-0490), and MGC803 (SCSP-5027)—were commercially sourced from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). AGS and MGC803 cells were propagated in RPMI-1640 basal medium (Cat. #11875093, Gibco, Waltham, Massachusetts, USA), while MKN28 and SGC-7901 cells were propagated in high-glucose DMEM medium (Cat. #11965092, Gibco, Waltham, Massachusetts, USA). All media contained 10% (v/v) heat-treated fetal bovine serum (FBS; Cat. #10099141, Gibco, Waltham, Massachusetts, USA), along with penicillin-streptomycin antibiotics at final concentrations of 100 U/mL and 100 µg/mL (Cat. #15140122, Gibco, Waltham, Massachusetts, USA). Cells were incubated under standard conditions at 37 ℃ in a humidified incubator with 5% CO₂. Medium was replaced every two days until confluence was achieved.

Western blot analysis

Proteins were isolated from GC cells subjected to PHGDH knockdown, overexpression, or treatment with autophagy inhibitors, as well as from untreated patient tumor tissues. For lysis procedures, all samples underwent homogenization in ice-cold RIPA lysis buffer containing dual-inhibitor cocktails. Homogenates were incubated at 4 ℃ for 30 minutes with periodic vortexing, followed by centrifugation at 12,000 ×g for 20 minutes at 4 ℃. The resulting supernatants were collected for subsequent analysis. Protein concentration was performed with an Enhanced BCA Protein Assay Kit. Equivalent protein quantities (20–50 µg/lane) were resolved on either 6% or 10% polyacrylamide gels through SDS-PAGE electrophoresis, then electroblotted onto PVDF membranes. Following overnight blocking at 4 ℃ with 5% skim milk, the membranes were probed with primary antibodies for 16–18 hours at 4 ℃. Following TBST washes, HRP-conjugated secondary antibodies were applied at 37 ℃ for 1 h, and chemiluminescent signals were subsequently revealed via an ECL detection kit.

Quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA was extracted and purified using TRIzol reagent (Cat. #15596026, Invitrogen, Waltham, Massachusetts, USA). The extracted RNA was then subjected to reverse transcription with the RevertAid First Strand cDNA Synthesis Kit (Cat. #K1622, Invitrogen, Waltham, Massachusetts, USA) to generate complementary DNA. qRT-PCR was conducted on a LightCycler 480 II system (Roche, Basel, Switzerland) with SYBR Premix Ex Taq (Takara, Kusatsu, Shiga, Japan) in a final reaction volume of 20 µL. The amplification protocol consisted of an initial denaturation step at 95 ℃ for 30 s, followed by 40 cycles of denaturation (95 ℃, 15 s), annealing (58 ℃, 10 s), and extension (72 ℃, 20 s), with a final elongation (72 ℃, 5 min). GAPDH was used as an endogenous reference gene, and relative quantification of target gene expression was calculated via the 2^−ΔΔCt approach. Primer sequences: PHGDH F: 5′-TTCAGTCACATGCTGCTTCC-3′; PHGDH R: 5′-GGCTGCTGTCCTACCAGACT-3′; GAPDH F: 5′-GTGAAGGTCGGAGTCAACG-3′; GAPDH R: 5′-TGAGGTCAATGAAGGGGTC-3′.

siRNA transfection and stable cell line construction

Specific siRNA targeting PHGDH (sense: 5′-GGGAGGAAAUUGCUGUUCATT-3′; antisense: 5′-UGAACAGCAAUUUCCUCCCTT-3′) was used for the transient knockdown of PHGDH. The si-NC was a scrambled small interfering RNA that had no known target in the human genome to act as a negative control, and the “Blank” refers to cells that were not transfected. As a measure of specificity and possible off-target effects, two independent siRNA constructs (si-PHGDH#1 and si-PHGDH#2) were tested without varied results. Stable PHGDH knockdown was accomplished through the use of lentiviral vectors with two unique short hairpin RNA (shRNA) sequences (shPHGDH#1 and shPHGDH#2). Stable overexpression was accomplished through the use of the full-length PHGDH cDNA within the PCDH-CMV-MCS-EF1-puro vector. Lentiviral particles were produced in 293T cells and used to infect target cells, which were selected for puromycin resistance after transduction. All successful knockdown or overexpressions was verified by Western blot.

Colony formation and cell proliferation assays

MGC803 and AGS cells transfected with si-PHGDH or stably overexpressing PHGDH were plated in 96-well plates at densities of 8×10³ and 5×10³ cells/well. Following a 48-hour incubation period, metabolic viability was assessed using the MTS assay (Promega, Madison, Wisconsin, USA). After allowing the reaction to proceed for 2 hours at 37 ℃, optical density values were recorded at 490 nm. For analysis of clonogenic capacity, genetically modified cells were distributed in 6-well plates at 600 cells/well and maintained in culture for 10 days with medium refreshed every 3 days. Subsequently, cell colonies were immobilized with 4% PFA for 10 minutes and then subjected to staining with 0.1% crystal violet for 20 minutes. Only cellular aggregates consisting of ≥50 individual cells were considered viable colonies and enumerated through microscopic examination.

Evaluation of cell migration and invasion capacity

To assess metastatic potential, MGC803 cells with siRNA or stable shRNA-mediated PHGDH knockdown (si-/sh-PHGDH) and AGS cells with stable PHGDH overexpression (PHGDH-OE) were examined. Polycarbonate membrane inserts (8 µm pores; EMD Millipore, Burlington, Massachusetts, USA) were positioned in 24-well plates for both experimental approaches. For invasion assays, the filter’s apical side received a 60 µL coating of Matrigel (BD Biosciences, Franklin Lakes, New Jersey, USA), while migration assays were conducted without Matrigel.

Single-cell suspensions of MGC803 (5×10⁵) or AGS (2×10⁵) cells were prepared in 200 µL serum-free RPMI-1640 medium, and carefully loaded into the apical compartments. The basal compartments received 800 µL of complete RPMI-1640 medium containing 10% FBS as a chemotactic stimulus. Following a 24-hour incubation period under physiological conditions (37 ℃, 5% CO₂), non-migratory cells retained on the upper chamber were mechanically eliminated. The transmigrated cells attached to the lower membrane surface were subsequently immobilized, subjected to staining, and quantitatively analyzed using microscopic examination.

Cell migration assessment

To evaluate directional cell migration, confluent monolayers were established by culturing genetically modified MGC803 cells with stable PHGDH knockdown (sh-PHGDH) and AGS cells with PHGDH overexpression (PHGDH-OE) in 6-well plates under standard incubation conditions (37 ℃, 5% CO₂) until a confluent monolayer formed. To create reproducible wound margins, a 200 µL pipette tip (sterilized) was drawn across the confluent monolayer in straight paths. Following wound generation, non-adherent cells were eliminated through three successive PBS washes, after which serum-free medium was added to each well to minimize proliferation effects. Images of the scratch area were documented at 0, 12, 24, and 36 h using an inverted phase-contrast optical system. Quantitative analysis was performed by measuring the wound gap width with ImageJ software (NIH, USA), and the relative wound closure percentage was determined to evaluate cellular migratory capacity.

Cell cycle and apoptosis analysis

MGC803 cells (2×10⁶/well) were plated in 6-well plates and allowed to adhere for 24 hours prior to siRNA transfection. Post-transfection samples were collected at 24 h and 48 h intervals for subsequent cell cycle and apoptosis analysis. For cell cycle detection, harvested cells were immobilized in chilled 70% ethanol and maintained at 4 ℃ overnight (20 hours) under light-protected conditions. Following PBS washes, cellular DNA was labeled with PI using a commercial kit (MedChemExpress, Monmouth Junction, New Jerse, USA). Apoptotic evaluation was performed using dual staining with fluorescein isothiocyanate-conjugated Annexin V and PI (Yeasen, Shanghai, China). All flow cytometric measurements were conducted on an Attune NxT flow cytometer (Thermo Fisher Scientific), with subsequent data analysis performed using FlowJo software (BD Biosciences).

Transfection and evaluation of fluorescent dots

MGC803 cells (2×10⁶/well) were plated in 6-well plates and allowed to adhere overnight. After transfection with si-PHGDH or si-NC for 24 h, cells were infected with mRFP-GFP-LC3 reporter adenovirus (HanBio, Shanghai, China) for another 24 hours. After enzymatic dissociation, the transfected cells were then digested and reseeded onto coverslips in 24-well plates. Once attached, cellular autophagy was modulated through 24-hour treatment with either 3-MA or CQ. Live-cell imaging was performed at physiological temperature (37 ℃) using a laser scanning confocal microscope. Yellow and red LC3 puncta were visualized and quantified using ImageJ software.

Measurement of cellular ROS

Cellular oxidative stress was evaluated using a commercially available ROS detection kit (Beyotime Biotechnology, Shanghai, China). Briefly, MGC803 cells were collected after si-PHGDH transfection for 24 or 48 h, then harvested and exposed to 10 µM DCFH-DA probe at 37 ℃ for 30 minutes under light-protected conditions. Cells were rinsed with serum-free medium to clear excess probe, followed by flow cytometric quantification (Attune NxT) and FlowJo-based data processing (version X).

Proteomics sample preparation and mass spectrometry analysis

After PHGDH interference (shRNA transfection), GC cells were collected and lysed in RIPA lysis buffer supplemented with dual protease/phosphatase inhibitor cocktails. Protein concentrations were measured via a BCA assay (Beyotime, Shanghai, China). For proteolytic digestion, protein extracts underwent reduction with 8 M urea and 10 mM DTT (37 ℃, 2 h), then alkylated using 50 mM iodoacetamide (15 min at 25 ℃ in darkness). The proteins were concentrated using centrifugal ultrafiltration devices (Nanosep 10K) and washed multiple times with 8 M urea and 25 mM NH₄HCO₃. Subsequently, Trypsin digestion (Promega, 1:100 w/w) was carried out at 37 ℃ overnight (12 h). The resulting peptides were labeled using a TMT labeling kit (Thermo Scientific, Waltham, Massachusetts, USA). TMT-labeled peptides were combined in equimolar ratios and fractionated via high-pH reversed-phase liquid chromatography.

The samples were subjected to mass spectrometric characterization on an Orbitrap Fusion Lumos Tribrid mass spectrometer interfaced with a Vanquish Neo UHPLC system. Chromatographic separation employed a two-column setup: (I) trapping column (75 µm × 2 cm, PepMap100) and (II) analytical column (75 µm × 20 cm, Reprosil-Pur C18-AQ, 1.9 µm). The mobile phase consisted of 0.1% formic acid in water (A) and 80% acetonitrile with 0.1% formic acid (B), eluted at 400 nL/min with a linear gradient. The mass spectrometer operated in data-dependent acquisition (DDA) mode, acquiring MS1 spectra in the 380–980 m/z range at a resolution of 120,000, followed by HCD-MS/MS (30,000 resolution) of the most intense precursors. Database searching against the UniProt human proteome (Proteome Discoverer 2.4) enabled protein identification and TMT-based quantification.

Statistical analysis

Quantitative data are presented as mean ± SD (n≥3 biological replicates) from at least three independent experiments. Differences between groups were assessed using Student’s t-test or one-way analysis of variance (ANOVA). Statistical significance thresholds were established as follows: *P<0.05 (significant), **P<0.01 (highly significant), and ***P<0.001 (extremely significant).

Results

PHGDH upregulation correlates with cellular differentiation and tumor stage in GC

To evaluate PHGDH expression levels in GC, we first evaluated its protein levels in tumor samples at different clinical stages. The results of the Western blotting showed PHGDH protein expression levels were elevated in tumor samples when compared to matched normal mucosa, the greatest of which were in Stage III and Stage IV (Figure 1A). In a matching fashion, qRT-PCR showed that PHGDH mRNA levels were both significantly elevated in the tumor samples and that these levels followed similar stage-dependent trends (Figure 1B). Immunohistochemical analysis (IHC) further supported this aim by performing stainability analysis, which revealed that PHGDH presence was significantly higher in tumors of higher clinical stages (Figure 1C). Densitometric quantification of the Western blot bands was performed using ImageJ (Figure 1D), and mean PHGDH/GAPDH ratios ± SD are reported as an error bar (Figure 1E). It must be noted again that due to the limited number of hospital-based clinical specimens (n=7 pairs), these findings should be interpreted as preliminary evidence.

Figure 1 Expression of PHGDH in GC tissues and cell lines. (A) Detection of PHGDH protein levels by Western blot in GC tissues from different stages, along with corresponding adjacent noncancerous tissues. (B) Analysis of PHGDH mRNA levels in GC specimens and corresponding adjacent noncancerous tissues across different stages. (C) Representative immunohistochemistry images showing PHGDH levels in GC tissues and paired noncancerous tissues (scale bar =50 µm). (D) Western blot analysis of PHGDH levels across GC cell lines with different degrees of differentiation. (E) Quantification of PHGDH protein levels in (D) by densitometry analysis. Data were presented as mean ± SD from three independent experiments (n=3). Statistical significance in (B) was assessed using Student’s t-test. Statistical analysis in (E) was performed using one-way ANOVA along with Tukey’s post-hoc test. *, P<0.05; **, P<0.01; ***, P<0.001. ANOVA, analysis of variance; GC, gastric cancer; PHGDH, phosphoglycerate dehydrogenase; SD, standard deviation.

We then measured PHGDH levels in gastric cell lines of differing differentiation statuses (AGS, MKN-28, SGC-7901, and MGC803). The Western blotting showed differences in PHGDH expression, with the highest levels in poorly differentiated MGC803 cells and the lowest levels in AGS (Figure 1D,1E), a trend that reflected progression in differentiation. Supporting the weight of the particles, the qRT-PCR analysis indicated that the MGC803 cells had the highest PHGDH mRNA levels and the lowest in AGS. With these results established, knockdown studies were conducted in MGC803 cells and overexpression studies in AGS cells for functional significance.

Knockdown of PHGDH suppresses cellular proliferation and metastatic potential in MGC803 cells

To knock down PHGDH expression, MGC803 cells that had relatively high levels of endogenous PHGDH were transfected with a PHGDH-specific siRNA (si-PHGDH). The extent to which PHGDH was knocked down was measured by qRT-PCR and Western blotting (Figure 2A,2B), with both transcripts and proteins showing significantly reduced levels after knockdown. Colony-formation and MTS (metabolic activity) assays demonstrated that knockdown significantly decreased both the total number of colonies and metabolic activity, suggesting that MGC803 proliferation was inhibited relative to controls (Figure 2C,2D). The y-axis is labeled, “Cell viability (OD490 nm, MTS assay)”. All data are shown as mean ± SD of n=3 independent experiments. In summary, PHGDH silencing significantly inhibited both migration (Figure 2E,2F) and invasion (Figure 2G,2H) when compared to controls. Taken together, these data demonstrate that PHGDH drives GC cell growth and the process of metastasis.

Figure 2 Silencing PHGDH inhibited the growth, migratory capacity, and invasive potential of MGC803 cells. (A) Western blot analysis of PHGDH protein levels in MGC803 cells after siRNA-mediated knockdown. (B) RT-PCR assessed PHGDH mRNA expression after knockdown. (C) Colony formation assay examining the colony-forming ability of MGC803 cells after silencing PHGDH. (D) A cell growth curve assay was conducted to assess the proliferative ability of MGC803 cells after PHGDH knockdown. (E) Transwell migration assay and (F) matrigel invasion assay examining MGC803 cells after transient PHGDH knockdown. (G,H) Quantification of migrated and invaded cells, respectively. Data were presented as mean ± SD from three independent experiments (n=3). Statistical analysis was performed using one-way ANOVA along with Tukey’s post-hoc test. ***, P<0.001. ANOVA, analysis of variance; NC, negative control; PHGDH, phosphoglycerate dehydrogenase; RT-PCR, reverse transcription-polymerase chain reaction; SD, standard deviation.

PHGDH knockdown induces apoptosis but does not cause significant cell cycle arrest in MGC803 cells

Annexin V/7-AAD staining indicated that si-PHGDH significantly increased both early and late apoptotic fractions at 24 and 48 h (Figure 3A,3B), demonstrating that apoptosis is a major mechanism of growth inhibition. Flow-cytometric cell-cycle analysis did not show any significant differences in the percentage of G1, S, or G2/M phase of the cell cycle after treatment between the si-PHGDH and si-NC groups (Figure 3C-3E). Together, these results suggest that the antiproliferative effects associated with knockdown of PHGDH occur through apoptosis rather than cell-cycle arrest.

Figure 3 Apoptosis and cell cycle progression analysis of MGC803 cells following transient PHGDH knockdown. (A) Apoptosis of MGC803 cells at 24 and 48 h using Annexin V/7-AAD double staining followed by flow cytometry. (B) Quantification of apoptotic cells shown in (A). (C) PI staining-based flow cytometry for cycle distribution at 24 and 48 h after PHGDH knockdown was analyzed by flow cytometry using PI single staining. (D,E) Phase-specific statistical analysis at 24 and 48 h (E) based on (C). Data were presented as mean ± SD from three independent experiments (n=3). Statistical analysis was performed using one-way ANOVA along with Tukey’s post-hoc test. **, P<0.01; ***, P<0.001. ANOVA, analysis of variance; NC, negative control; PHGDH, phosphoglycerate dehydrogenase; PI, propidium iodide; SD, standard deviation.

Stable knockdown of PHGDH inhibits the migratory capacity of MGC803 cells

Stable PHGDH-silenced MGC803 cell lines (shPHGDH1 and shPHGDH4) were established. Both qRT-PCR and Western blot confirmed significant downregulation of PHGDH (Figure 4A,4B).

Figure 4 Stable PHGDH knockdown inhibited the migration of MGC803 cells. (A) RT-PCR detected PHGDH levels in stably transduced MGC803 clones. (B) Western blot detection of PHGDH protein suppression in stable knockdown clones. (C) Transwell migration assay assessing cell migration following stable PHGDH knockdown. (D) Quantification of migrated cells is shown in (C). (E) A wound healing assay was performed at 0, 12, and 24 h. (F) Quantification of wound closure percentages is shown in (E) at the indicated time points. Data were presented as mean ± SD from three independent experiments (n=3). Statistical analysis was performed using one-way ANOVA along with Tukey’s post-hoc test. ***, P<0.001. ANOVA, analysis of variance; NC, negative control; PHGDH, phosphoglycerate dehydrogenase; RT-PCR, reverse transcription-polymerase chain reaction; SD, standard deviation.

Transwell and wound-healing assays showed that stable PHGDH knockdown led to significantly fewer migrated cells and slower wound closure (Figure 4C-4F), confirming reduced motility.

PHGDH overexpression enhances the metastatic potential of AGS cells

Two PHGDH-overexpressing clones (PHGDH1-OE, PHGDH2-OE) were validated by Western blot (Figure 5A).

Figure 5 Cell expansion, movement, and penetration assays of AGS cells following stable PHGDH overexpression. (A) Western blot verification of PHGDH protein levels in AGS cells with stable PHGDH overexpression. (B) Cell proliferation was assessed using an MTS assay at designated time points. (C) Wound healing assay evaluating cell migration at 0, 12, 24, and 36 h. (D) Quantification of wound closure percentages shown in (C). (E) Cell migration was analyzed using a Transwell assay. (F) Quantification of migrated cells is shown in (E). (G) Matrigel invasion assay assessing invasive capacity. (H) Quantification of invaded cells is shown in (G). Data were presented as mean ± SD from three independent experiments (n=3). Statistical analysis was performed using one-way ANOVA along with Tukey’s post-hoc test. *, P<0.05; **, P<0.01; ***, P<0.001. ANOVA, analysis of variance; NC, negative control; PHGDH, phosphoglycerate dehydrogenase; SD, standard deviation.

Cell viability (OD490 nm, MTS assay) did not differ significantly from controls within 72 h (Figure 5B; mean ± SD, n=3).

However, PHGDH-OE significantly enhanced wound-healing and Transwell migration/invasion rates (Figure 5C-5H), confirming its pro-metastatic role.

Error bars have been added to Figure 5B, and statistical tests are based on three independent replicates.

PHGDH knockdown enhances autophagic flux in MGC803 cells

The analysis of the proteomic profile in MGC803 cells that have been silenced for PHGDH revealed pathways involving autophagy to be enriched, which included Autophagy-animal, PI3K-Akt, and Endocytosis (ref. Figure 6A). Western blot confirmed an increase in LC3-II and a moderate decrease in p62 (ref. Figure 6B); both results indicated activation of autophagy in cells. The change in p62 was modest, but the decrease in p62 was consistent in all replicates, suggesting the activation of the autophagic flux instead of the accumulation of substrate. Three-MA was the treatment that decreased levels of LC3-II (ref. Figure 6C), and CQ increased levels of both LC3-II and p62 (ref. Figure 6D), indicating that turnover of autophagosomes was occurring. mRFP-GFP-LC3 imaging showed increased red puncta in response to si-PHGDH, suggesting increased autolysosome formation (ref. Figure 6E). Data from replicate (n=3) data indicated there was reproducibility in the results. At the initiation level, FIP200 and ATG101 were increased after silencing of PHGDH with minor changes to ULK1 and ATG13 (ref. Figure 6F).

Figure 6 Expression analysis of autophagy-related proteins and pathway profiling after PHGDH knockdown in gastric cancer cells. (A) KEGG pathway enrichment analysis based on proteomic data from PHGDH-knockdown cells. (B) Immunoblotting was performed to quantify LC3 conversion and P62 degradation after transient PHGDH knockdown. (C) Western blot detection of LC3, P62, mTOR, and p-mTOR protein levels following PHGDH knockdown and treatment with or without 3-MA (12.5 µM). (D) Western blot detection of LC3, P62, mTOR, and p-mTOR protein levels after PHGDH knockdown and treatment with or without CQ (12.5 µM). (E) Fluorescence signals of autophagosomes and autophagic flux were observed using an mRFP-GFP-LC3 dual-labeled adenoviral system following PHGDH knockdown. (F) Western blot assessment of autophagy initiation-related proteins ULK1, p-ULK1, FIP200, ATG13, and ATG101 post-PHGDH knockdown. All Western blot experiments were performed in triplicate (n=3) and yielded consistent results. KEGG, Kyoto Encyclopedia of Genes and Genomes; NC, negative control; PHGDH, phosphoglycerate dehydrogenase.

PHGDH knockdown increased ROS production and HIF-1α expression in MGC803 cells

With limited involvement of ULK1 phosphorylation, we examined pathways activated by oxidative stress. Flow cytometry determinations revealed that levels of ROS (mean fluorescence intensity, MFI) increased significantly after PHGDH knockdown from 143 at 24 h to 183 at 48 h versus 61.6 in si-NC (Figure 7A,7B).

Figure 7 ROS accumulation and HIF-1α upregulation following PHGDH knockdown in gastric cancer cells. (A) Flow cytometry for intracellular ROS levels at 24 h and 48 h after transient PHGDH knockdown in MGC803 cells. (B) Quantification of ROS levels based on the MFI from (A). (C) Western blot for HIF-1α levels in MGC803 cells with stable PHGDH knockdown. (D) Densitometric analysis of HIF-1α protein levels to GAPDH from (C). Data were presented as mean ± SD from three independent experiments (n=3). Statistical analysis was performed using one-way ANOVA along with Tukey’s post-hoc test. *, P<0.05; ***, P<0.001. ANOVA, analysis of variance; MFI, mean fluorescence intensity; NC, negative control; PHGDH, phosphoglycerate dehydrogenase; SD, standard deviation.

Western blot analysis demonstrated increased protein levels of HIF-1α in shPHGDH1 and shPHGDH4 cells compared to controls (Figure 7C,7D).

Overall results suggest that PHGDH knockdown increases ROS and activates HIF-1α, which may induce autophagy through oxidative-stress-mediated consequences.

Discussion

Our data revealed that PHGDH expression is progressively upregulated in GC tissues in a stage-dependent pattern, and at the cellular level, PHGDH modulates autophagic flux through regulation of the ROS-HIF-1α axis and the Mtor-ULK1 pathway, thereby promoting cell migration and resistance to apoptosis. Beyond its canonical metabolic functions in GC, PHGDH emerges as a nodal regulator that orchestrates crosstalk between nutrient utilization pathways, stress adaptation mechanisms, and autophagic flux control.

As the gateway enzyme of the serine biosynthesis cascade, PHGDH governs tumor metabolic flux by controlling the provision of one-carbon donors essential for nucleotide synthesis and redox homeostasis, thereby supporting the biosynthesis of nucleotides, NADPH, and glutathione to enhance tumor cells' metabolic adaptability (20-23). In our study, we observed significant PHGDH upregulation in GC specimens, showing a stage-associated expression pattern. This clinicopathological trend, consistent with previous reports such as Xian et al.’s multicenter analysis, suggests that PHGDH upregulation may serve as a potential indicator of poor prognosis, pending validation in larger cohorts (24). Mechanistically, a prior study has demonstrated that PHGDH sustains tumor proliferation under serine-deprived conditions by activating the SSP, while simultaneously maintaining redox homeostasis through NADPH and glutathione, thereby mitigating oxidative stress-induced damage (25). Such functions confer PHGDH-overexpressing cells with greater adaptability to the tumor microenvironment and survival advantages, which likely underlie its close association with GC progression. Functional assays in our study demonstrated that PHGDH knockdown significantly impaired GC cell migration and invasion, whereas overexpression enhanced cell motility and transmembrane capacity, thereby confirming its pro-tumorigenic role. The pro-migratory effects of autophagy in malignancies are mediated through a tripartite mechanism encompassing focal adhesion kinase (FAK) pathway activation, integrin spatial redistribution, and microtubule dynamics fine-tuning, collectively facilitating invasive phenotypes (26,27). This phenomenon may be explained by PHGDH-mediated modulation of oxidative stress and autophagy pathways, thereby facilitating the invasive and migratory potential of tumor cells.

In terms of stress signaling, PHGDH knockdown significantly increased intracellular ROS levels and upregulated HIF-1α protein expression. This observation is consistent with the known role of PHGDH in maintaining redox homeostasis via regulation of glutathione and NADPH biosynthesis. As a central transcription factor in the hypoxia pathway, HIF-1α becomes stabilized under oxidative stress, initiating adaptive responses, including autophagy and glycolysis (28,29). Our findings support a model in which PHGDH modulates ROS accumulation to regulate HIF-1α expression, positioning PHGDH upstream of this pathway. Furthermore, PHGDH suppression enhanced autophagic activity, as evidenced by increased LC3-II levels and decreased P62 expression. Pharmacologic inhibition of PI3K with 3-MA reversed the autophagy induction, while CQ treatment confirmed autophagosome accumulation through impaired lysosomal degradation. These data collectively suggest that PHGDH regulates autophagy in GC cells through a ROS-HIF-1α-dependent pathway.

This concept is further supported by previous studies showing that PDGFR-β can activate autophagy in cancer-associated fibroblasts (CAFs) via the mTOR/FIP200/ATG13 signaling axis, accompanied by HIF-1α upregulation and enhanced lactate export and tumor cell migration (30). In accordance with this notion, the upregulation of FIP200 and ATG101 together with unchanged phosphorylation of ULK1 may indicate that PHGDH could alter the assembly of ULK1 complexes via oxidative stress and HIF-1α signaling in an mTOR-ULK1-independent manner, thereby revealing a novel metabolic-stress-autophagic regulation axis in GC.

KEGG pathway analysis uncovered PHGDH’s pleiotropic regulatory roles, with its transcriptional targets overrepresented in cancer-related processes, intracellular pathogen sensing systems, lysosome-dependent degradation pathways, and survival signal amplification circuits. These results suggest that PHGDH may broadly participate in tumor immune response, autophagy, and stress adaptation, thereby facilitating cross-pathway modulation critical for tumor progression. Supporting this, Jin et al. reported that the PHGDH inhibitor NCT-503 synergistically enhanced temozolomide (TMZ)-induced growth suppression and apoptosis in MGMT-positive glioblastoma models, an effect accompanied by robust ROS accumulation that was partially reversed by the ROS scavenger N-acetylcysteine (NAC) (31). These observation underscores PHGDH’s role in coordinating ROS levels and downstream autophagic responses, highlighting its function as a critical regulatory hub in tumor stress adaptation and survival.

The PHGDH overexpression has always been linked with tumor progression and poor prognosis in clinical settings. Thus, it may be used as a prognostic biomarker in GC and is a possible therapeutic target, especially due to the existence of small-molecule inhibitors, including NCT-503 and CBR-5884, that have shown significant antitumor efficacy in preclinical models (32,33). The present study further supports PHGDH inhibition as a compelling therapeutic strategy for GC, especially in high-PHGDH subtypes characterized by aggressive invasion, poor differentiation, and therapeutic resistance.

Despite systematically elucidating the role of PHGDH in regulating migration, invasion, and autophagic activity in GC cells, several mechanistic aspects require further validation. The small number of clinical specimens (n=7) limits definitive statistical inference regarding the association of PHGDH with tumor stage and calling for validation in large independent cohorts and with publicly available transcriptomic datasets (e.g., TCGA-STAD) to further establish clinical relevance. While we only used one siRNA sequence in initial experiments, later studies (i.e., confirmation with a second independent siRNA and shRNA-based stable knockdown) provided an added level of specificity with reduced off-target concern in our findings.

Further studies should seek to establish the direct causal relationship between HIF-1α regulation, induction of FIP200/ATG101, and ULK1 complex formation. In addition, extending to in vivo GC models and multi-omic analyses would further elucidate the systemic metabolic and immunologic effects of PHGDH inhibition. Overall, the findings of this study delineate PHGDH as a critical modulator of oxidative stress and autophagy in GC, with new mechanistic insights into its potential role in tumor aggressiveness and therapeutic vulnerability.

Conclusions

PHGDH, an important enzyme in the serine biosynthetic pathway, is overexpressed in GC tissues and cell lines. Knockdown of PHGDH led to inhibition of cellular proliferation, migration, and invasion, but also activation of autophagy through ROS-HIF-1α signaling. These findings support the idea that PHGDH is involved in the regulation of oxidative stress and autophagy processes that promote GC progression. Although the sample size is small, this study provides early evidence that supports PHGDH as a possible molecular target in the treatment of GC and serves as a basis for future larger clinical studies.

Acknowledgments

We would like to thank the Project of the Natural Science Foundation of Henan province.

Footnote

Reporting Checklist: The authors have completed the MDAR reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1817/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1817/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1817/prf

Funding: This study was supported by the Project of the Natural Science Foundation of Henan Province (No. 162300410037).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1817/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Basic Medical Sciences, Henan University (No. HUSOM2020-027), and all participants provided written informed consent.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Global variation in postoperative mortality and complications after cancer surgery: a multicentre, prospective cohort study in 82 countries. Lancet 2021;397:387-97. [Crossref] [PubMed]
3. Sundar R, Nakayama I, Markar SR, et al. Gastric cancer. Lancet 2025;405:2087-102. [Crossref] [PubMed]
4. Li Y, Ren N, Zhang B, et al. Gastric cancer incidence trends in China and Japan from 1990 to 2019: Disentangling age-period-cohort patterns. Cancer 2023;129:98-106. [Crossref] [PubMed]
5. Zhang H, Li Z, Ruan Y, et al. Low expression of hsa_circ_0001811 in gastric cancer and its role in clinical diagnosis. J Clin Lab Anal 2021;35:e23642. [Crossref] [PubMed]
6. Fang T, Wang Y, Yin X, et al. Diagnostic Sensitivity of NLR and PLR in Early Diagnosis of Gastric Cancer. J Immunol Res 2020;2020:9146042. [Crossref] [PubMed]
7. Ocaña MC, Martínez-Poveda B, Quesada AR, et al. Metabolism within the tumor microenvironment and its implication on cancer progression: An ongoing therapeutic target. Med Res Rev 2019;39:70-113. [Crossref] [PubMed]
8. Guerra L, Bonetti L, Brenner D. Metabolic Modulation of Immunity: A New Concept in Cancer Immunotherapy. Cell Rep 2020;32:107848. [Crossref] [PubMed]
9. Jiang Z, Gorenstein NM, Morré DM, et al. Molecular cloning and characterization of a candidate human growth-related and time-keeping constitutive cell surface hydroquinone (NADH) oxidase. Biochemistry 2008;47:14028-38. [Crossref] [PubMed]
10. Li X, Peng X, Li Y, et al. Glutamine addiction in tumor cell: oncogene regulation and clinical treatment. Cell Commun Signal 2024;22:12. [Crossref] [PubMed]
11. Wu X, Xia J, Zhang J, et al. Phosphoglycerate dehydrogenase promotes proliferation and bortezomib resistance through increasing reduced glutathione synthesis in multiple myeloma. Br J Haematol 2020;190:52-66. [Crossref] [PubMed]
12. Cappello A, Tosetti G, Smirnov A, et al. p63 orchestrates serine and one carbon metabolism enzymes expression in head and neck cancer. Biol Direct 2023;18:73. [Crossref] [PubMed]
13. Shu M, Liu Y, Wang J. Protein post-translational modifications in serine synthetic pathway: functions and molecular mechanisms. Cell Commun Signal 2025;23:311. [Crossref] [PubMed]
14. Gromova I, Gromov P, Honma N, et al. High level PHGDH expression in breast is predominantly associated with keratin 5-positive cell lineage independently of malignancy. Mol Oncol 2015;9:1636-54. [Crossref] [PubMed]
15. Li M, Wu C, Yang Y, et al. 3-Phosphoglycerate dehydrogenase: a potential target for cancer treatment. Cell Oncol (Dordr) 2021;44:541-56. [Crossref] [PubMed]
16. Zhu J, Ma J, Wang X, et al. High Expression of PHGDH Predicts Poor Prognosis in Non-Small Cell Lung Cancer. Transl Oncol 2016;9:592-9. [Crossref] [PubMed]
17. Yoon S, Kim JG, Seo AN, et al. Clinical Implication of Serine Metabolism-Associated Enzymes in Colon Cancer. Oncology 2015;89:351-9. [Crossref] [PubMed]
18. Wang K, Luo L, Fu S, et al. PHGDH arginine methylation by PRMT1 promotes serine synthesis and represents a therapeutic vulnerability in hepatocellular carcinoma. Nat Commun 2023;14:1011. [Crossref] [PubMed]
19. Hua L, Song Y, Min J, et al. AKR7A3 modulates the metastasis of pancreatic ductal adenocarcinoma through regulating PHGDH-suppressed autophagy. Cancer Sci 2023;114:3101-13. [Crossref] [PubMed]
20. Pacold ME, Brimacombe KR, Chan SH, et al. A PHGDH inhibitor reveals coordination of serine synthesis and one-carbon unit fate. Nat Chem Biol 2016;12:452-8. [Crossref] [PubMed]
21. Wei L, Lee D, Law CT, et al. Genome-wide CRISPR/Cas9 library screening identified PHGDH as a critical driver for Sorafenib resistance in HCC. Nat Commun 2019;10:4681. [Crossref] [PubMed]
22. Shiota M, Naya M, Yamamoto T, et al. Gold-nanofève surface-enhanced Raman spectroscopy visualizes hypotaurine as a robust anti-oxidant consumed in cancer survival. Nat Commun 2018;9:1561. [Crossref] [PubMed]
23. Zhang W, Yin C, Qi L, et al. RFWD3 Reprograms Nucleotide Metabolism Through PHGDH to Induce Chemoresistance In Osteosarcoma. Adv Sci (Weinh) 2025;12:e2410937. [Crossref] [PubMed]
24. Xian Y, Zhang S, Wang X, et al. Phosphoglycerate dehydrogenase is a novel predictor for poor prognosis in gastric cancer. Onco Targets Ther 2016;9:5553-60. [Crossref] [PubMed]
25. Vandekeere S, Dubois C, Kalucka J, et al. Serine Synthesis via PHGDH Is Essential for Heme Production in Endothelial Cells. Cell Metab 2018;28:573-587.e13. [Crossref] [PubMed]
26. Sandilands E, Serrels B, McEwan DG, et al. Autophagic targeting of Src promotes cancer cell survival following reduced FAK signalling. Nat Cell Biol 2011;14:51-60. [Crossref] [PubMed]
27. Fontana R, Vivo M. Dynamics of p14ARF and Focal Adhesion Kinase-Mediated Autophagy in Cancer. Cancers (Basel) 2018;10:221. [Crossref] [PubMed]
28. Wang Y, Mang X, Li D, et al. Piezoeletric cold atmospheric plasma induces apoptosis and autophagy in human hepatocellular carcinoma cells through blocking glycolysis and AKT/mTOR/HIF-1α pathway. Free Radic Biol Med 2023;208:134-52. [Crossref] [PubMed]
29. Wang JZ, Zhu W, Han J, et al. The role of the HIF-1α/ALYREF/PKM2 axis in glycolysis and tumorigenesis of bladder cancer. Cancer Commun (Lond) 2021;41:560-75. [Crossref] [PubMed]
30. Zhang L, Wang K, Zhang J, et al. PDGFR-β/Cav1-induced autophagy via mTOR/FIP200/ATG13 activation in cancer-associated fibroblasts promotes the malignant progression of breast cancer. J Transl Med 2025;23:784. [Crossref] [PubMed]
31. Jin L, Kiang KM, Cheng SY, et al. Pharmacological inhibition of serine synthesis enhances temozolomide efficacy by decreasing O(6)-methylguanine DNA methyltransferase (MGMT) expression and reactive oxygen species (ROS)-mediated DNA damage in glioblastoma. Lab Invest 2022;102:194-203. [Crossref] [PubMed]
32. Elsaadi S, Steiro I, Abdollahi P, et al. Targeting phosphoglycerate dehydrogenase in multiple myeloma. Exp Hematol Oncol 2021;10:3. [Crossref] [PubMed]
33. Pralea IE, Moldovan RC, Țigu AB, et al. Cellular Responses Induced by NCT-503 Treatment on Triple-Negative Breast Cancer Cell Lines: A Proteomics Approach. Biomedicines 2024;12:1087. [Crossref] [PubMed]