Non-canonical function of PHGDH promotes HCC metastasis by interacting with METTL3

Abstract

Purpose

Phosphoglycerate dehydrogenase (PHGDH), a pivotal enzyme in serine synthesis, plays a key role in the malignant progression of tumors through both its metabolic activity and moonlight functions. This study aims to elucidate the non-canonical function of PHGDH in promoting hepatocellular carcinoma (HCC) metastasis through its interaction with methyltransferase-like 3 (METTL3), potentially uncovering a novel therapeutic target.

Methods

Western blot was used to study PHGDH expression changes under anoikis and cellular functional assays were employed to assess its role in HCC metastasis. PHGDH-METTL3 interactions were explored using GST pull-down, Co-immunoprecipitation and immunofluorescence assays. Protein stability and ubiquitination assays were performed to understand PHGDH’s impact on METTL3. Flow cytometry, cellular assays and nude mice model were used to confirm PHGDH's effects on anoikis resistance and HCC metastasis in vitro and in vivo.

Results

PHGDH is upregulated under anoikis conditions, thereby enhancing the metastatic potential of HCC cells. By interacting with METTL3, PHGDH prevents its ubiquitin-dependent degradation, resulting in higher METTL3 protein levels. This interaction upregulates epithelial-mesenchymal transition related genes, contributing to anoikis resistance and HCC metastasis. Nude mice model confirms that PHGDH’s interaction with METTL3 is crucial for driving HCC metastasis.

Conclusion

Our research presents the first evidence that PHGDH promotes HCC metastasis by interacting with METTL3. The PHGDH-METTL3 axis may serve as a potential clinical therapeutic target, offering new insights into the multifaceted roles of PHGDH in HCC metastasis.

Introduction

Metabolic reprogramming is a fundamental hallmark of cancer. Despite abundant oxygen, tumor cells often preferentially engage in utilizing aerobic glycolysis, a phenomenon known as the Warburg effect, which supports the production of both energy and biomolecules essential for tumor growth [1, 2]. Metabolic enzymes regulate metabolic fluxes through their classical enzymatic activities to meet the needs of tumor cells [3]. However, a burgeoning body of research has unveiled that these enzymes can also manifest non-canonical activities in response to aberrant signaling within the tumor microenvironment, thereby exerting a significant influence on the metabolic plasticity of cancer cells [4–6].

Phosphoglycerate dehydrogenase (PHGDH), the rate-limiting enzyme in the serine biosynthesis pathway, has emerged as a metabolic enzyme frequently deregulated in tumors [7–9]. PHGDH exhibits common overexpression in various cancers, leading to increased cell proliferation and migration capabilities [10–13]. PHGDH has been linked to the pathogenesis of cancer beyond its classical enzymatic function. For instance, PHGDH promotes the assembly of the eIF4F complex on the 5’ mRNA structure through its interaction with eIF4A1 and eIF4E, thereby enhancing the expression of proteins pivotal to pancreatic cancer progression [14]. In breast cancer, downregulation of PHGDH expression has been correlated with disruptions in its interaction with phosphofructokinase and the activation of the hexosamine pathway, thereby promoting tumor metastasis [15]. Intriguingly, nuclear-localized PHGDH forms a complex with nuclear cMYC, collaboratively driving the gene expression necessary for the recruitment of neutrophils and the establishment of the tumor microenvironment [16]. Additionally, the glucose deprivation-induced nuclear translocation of PHGDH has been shown to augment tumor growth [17]. Notably, PHGDH’ mitochondrial localization and it direct interaction with adenosine nucleotide translocator 2 (ANT2) have been implicated in the promotion of mitochondrial DNA-encoded proteins translation in hepatocellular carcinoma (HCC) cells [18]. In summary, the moonlight functions of PHGDH play a critical role in driving the malignant progression of tumors.

HCC metastasis is a major cause of death in most patients, with metastatic tumor cells dynamically and selectively adjusting their metabolism at each step of the metastatic cascade to adapt to the various tumor microenvironments [19, 20]. While PHGDH’s enzymatic activity is known to contribute to the metastatic process in HCC, the potential impact of its non-enzymatic functions on metastasis is still relatively unclear [21, 22]. Our research provides a comprehensive analysis of the molecular mechanisms by which PHGDH’s moonlight functions contribute to HCC metastasis. Moreover, we report for the first time that increased PHGDH expression in anoikis environments is a potent promoter of metastatic dissemination.

Results

PHGDH is closely associated with anoikis resistance in HCC

Anoikis, a distinct form of programmed cell death, is intimately linked to the metastatic potential of tumor cells [23, 24]. This study embarked on an exploration of the nexus between PHGDH and anoikis resistance. HCC cells were cultured in low-adhesion plates to simulate anoikis-inducing conditions. Western blot analysis revealed a significant upregulation of PHGDH expression under such conditions (Fig. 1A). Moreover, analysis of The Cancer Genome Atlas (TCGA) database revealed a significant correlation between elevated PHGDH expression and increased expression levels of genes associated with anoikis (Fig. 1B). The impact of PHGDH on the metastatic capacity of HCC cells was further assessed through wound healing and transwell assays, which showed that PHGDH overexpression markedly enhanced the metastatic potential of HCC cells (Fig. 1C-K) (n = 3 independent experiments). In summary, these findings underscore the elevated expression of PHGDH in the context of anoikis and its pivotal role in augmenting the metastatic potential of HCC cells.

Fig. 1.

PHGDH is closely associated with anoikis resistance. A Western blot analysis of PHGDH expression in PLC/PRF/5 and Huh7 cells under anoikis conditions (cultured in low-adhesion plates for 24 h). Meanwhile, flow cytometry was used to detect the effect on the apoptosis rate in PLC/PRF/5 and Huh7 cells. B Relative expression analysis of PHGDH and anoikis-related genes in the The Cancer Genome Atlas (TCGA) database using the log-rank test. C-F. Wound healing assay was performed in MHCC-97 H (C) and SK-Hpe1 (E) cells infected with Flag labeled adenovirus AdPHGDH/AdGFP to overexpress PHGDH, as well as in Huh7 (D) and PLC/PRF/5 (F) cells treated with shRNA lentiviral vector to achieve PHGDH knockdown (n = 3 independent experiments), scale bar = 200 μm. G-J Transwell assay was performed in MHCC-97 H (G) and SK-Hpe1 (H) cells infected with Flag labeled adenovirus AdPHGDH/AdGFP to overexpress PHGDH, as well as on PLC/PRF/5 (I) and Huh7 (J) cells treated with shRNA lentiviral vector to achieve PHGDH knockdown (n = 3 independent experiments), scale bar = 100 μm. K Western blot analysis was conducted to detect the protein expression levels of PHGDH in (C-J). P-values were derived from one-way ANOVA followed by Tukey’s test in (D, F, H, J) and unpaired two-tailed Student’s t-test in (C, E, G, I). Data are represented as the mean ± SD, *P < 0.05, **P < 0.01, ***P < 0.001

PHGDH interacts with METTL3 in hepatoma cells

Metabolic enzymes can influence tumor malignancy through protein-protein interactions [4, 5]. To investigate whether PHGDH is involved in the metastatic process of HCC via its interaction with metastasis-related proteins, a combined analysis of the PHGDH interaction protein database and cell adhesion-related genes was conducted. This analysis pointed to a potential interaction between PHGDH and cell adhesion-related proteins such as methyltransferase-like 3 (METTL3) (Fig. 2A). METTL3 is an RNA methyltransferase responsible for catalyzing the addition of N6-methyladenosine (m6A) modifications to mRNA. Additionally, METTL3 is also involved in tumor metastasis [25–27]. The interaction between PHGDH and METTL3 protein in HCC cells was validated through GST pull-down assay. His-tagged PHGDH protein was effectively captured by GST-tagged METTL3 protein, confirming their direct binding under in vitro conditions (Fig. 2B). To further validate this interaction, Co-immunoprecipitation (Co-IP) experiments were conducted with both endogenous and exogenous protein sources, which reinforced the in vivo interaction between PHGDH and METTL3 in HCC cells (Fig. 2C, D). Additionally, immunofluorescence experiments confirmed the co-localization of PHGDH and METTL3 in HCC cells (Fig. 2E). To precisely identify the interacting domains between PHGDH and METTL3, truncated mutants of both proteins were constructed, followed by truncation IP experiments. These assays revealed that PHGDH interacts with the C-terminal region of METTL3 via its RD domain (Fig. 2F, G). In summary, these results demonstrate that PHGDH interacts with METTL3 in HCC cells.

Fig. 2.

PHGDH interacts with METTL3. A Venn diagram analysis of PHGDH-interacting proteins database and cell-cell adhesion-related genes. B The fusion protein GST-METTL3 was mixed with the recombinant His-PHGDH in reaction buffer and incubated at a temperature of 37 degrees Celsius for a duration of 4 h. Subsequent detection of the samples was performed using Western blot. C Endogenous Co-immunoprecipitation analysis was performed to investigate the interaction between PHGDH and METTL3 in PLC/PRF/5 and Huh7 cells. D Exogenous Co-immunoprecipitation analysis was conducted in HEK293 cells transfected with PHGDH-HA and METTL3-Flag plasmids. E Representative immunofluorescence image for PHGDH and METTL3 in PLC/PRF/5 and Huh7 cells. Scale bar = 10 μm. F Truncated IP was performed to detect the interaction between PHGDH and truncated mutants of METTL3 (ΔN and ΔC) in HEK293 cells transfected with PHGDH-HA and METTL3 truncated mutants. G Truncated IP was conducted to assess the interaction between METTL3 and truncated mutants of PHGDH (ΔSBD1, ΔSBD1 + NBD, ΔSBD2 + RD, ΔRD) in HEK293 cells

PHGDH enhances METTL3 protein stability by inhibiting its ubiquitination-mediated degradation

The impact of the interaction between PHGDH and METTL3 was further investigated. It was observed that PHGDH overexpression led to an increase in METTL3 protein levels, whereas the converse effect was observed upon PHGDH knockdown (Fig. 3A, B). Given that protein-protein interactions can affect protein stability, we examined whether PHGDH influences the stability of METTL3 through their interaction. The results showed that PHGDH significantly enhanced METTL3 protein stability, whereas its knockout destabilized METTL3 protein in HCC cells (Fig. 3C-F). Protein degradation primarily occurs through two major pathways: the lysosomal degradation pathway and the ubiquitin-proteasome degradation pathway [28, 29]. To determine the pathway through which PHGDH regulates METTL3 stability, cells were treated with the proteasome inhibitor MG132 or lysosome inhibitor chloroquine. Notably, MG132 treatment significantly restored the reduction in METTL3 levels caused by PHGDH knockdown, whereas chloroquine treatment had no such effect (Fig. 3G). These findings indicate that PHGDH modulates METTL3 degradation via the ubiquitin-proteasome pathway. Furthermore, the effect of PHGDH on METTL3 ubiquitination levels was assessed by ubiquitination IP assays, which revealed that wild-type PHGDH overexpression markedly attenuated METTL3 ubiquitination, while the PHGDH-RD mutant lacked this inhibitory effect (Fig. 3H).

Fig. 3.

PHGDH enhances the stability of METTL3 protein by inhibiting its ubiquitination-mediated degradation. A Western blot analysis of METTL3 expression in MHCC-97 H and SK-Hpe1 cells infected with Flag labeled adenovirus AdPHGDH/AdGFP to overexpress PHGDH. B Western blot analysis of METTL3 expression in PLC/PRF5 and Huh7 cells treated with shRNA lentiviral vector to achieve PHGDH knockdown. C-F. Half-life analysis of METTL3 in PHGDH-overexpress MHCC-97 H (C) and SK-Hep1 (D), as well as PHGDH knockdown PLC/PRF/5 (E) and Huh7 (F) cells. G Western blot analysis of METTL3 expression in PHGDH knockdown PLC/PRF/5 and Huh7 cells treated with MG132 (10µM) or chloroquine (20µM) for 12 h. H Western blot analysis of METTL3 ubiquitination levels in MHCC-97 H cells transfected with METTL3-Flag and PHGDH-HA/PHGDH-ΔRD-HA. I. Western blot analysis to detect the impact of several E3 ubiquitin ligases on METTL3 protein levels in MHCC-97 H cell transfected with the corresponding recombinant plasmids. J Co-IP experiment to detect the interaction between PHGDH and SPOP in PLC/PRF/5 cells. K Western blot analysis of METTL3 ubiquitination levels in MHCC-97 H cells transfected with SPOP-Flag and PHGDH-HA

To determine the specific E3 ubiquitin ligase through which PHGDH influences the ubiquitination and subsequent degradation of METTL3, we evaluated the impact of various E3 ubiquitin ligases on METTL3 protein levels. Notably, we identified that speckle-type POZ protein (SPOP) markedly reduces METTL3 protein levels (Fig. 3I). This finding prompted us to hypothesize a role for PHGDH in regulating METTL3 ubiquitination and degradation through SPOP. Co-IP experiments substantiated the interaction between PHGDH and SPOP in HCC cells (Fig. 3J). Moreover, overexpression of SPOP enhanced the ubiquitination levels of METTL3, whereas overexpression of PHGDH showed the opposite effect (Fig. 3K). In conclusion, these results indicate that PHGDH increases METTL3 protein stability by inhibiting the ubiquitination and degradation of METTL3 through SPOP.

PHGDH inhibits anoikis and promotes HCC metastasis via METTL3 independent of enzymatic activity

Resistance to anoikis is a hallmark of tumor metastasis, enabling tumor cells to spread to distant organs via the circulatory system [23, 24]. METTL3 is closely associated with cell migration and can promote tumor metastasis by facilitating the epithelial-mesenchymal transition (EMT). We further investigated the potential of PHGDH to modulate EMT through METTL3. Western blot analysis revealed that PHGDH overexpression increased Snail protein levels and decreased E-cadherin expression, while METTL3 knockdown reversed these effects (Fig. 4A). This implies that PHGDH influences the EMT process through its interaction with METTL3. Furthermore, flow cytometry was used to evaluate changes in anoikis-induced apoptosis. It was observed that PHGDH overexpression significantly reduced HCC cell apoptosis in an anoikis environment, an effect reversed by METTL3 knockdown (Fig. 4B). Additionally, wound healing and transwell assays demonstrated that METTL3 knockdown reduced the PHGDH-induced enhancement of HCC cell metastasis (Fig. 4C-F) (n = 3 independent experiments).

Fig. 4.

PHGDH inhibits anoikis and promotes HCC metastasis via METTL3 independent of enzymatic activity. A Western blot analysis of E-cadherin and Snail expression in PHGDH-overexpress MHCC-97 H and SK-Hep1 cells. B Flow cytometry was used to detect the effect on the apoptosis rate in PHGDH overexpress MHCC-97 H cells treated with METTL3 shRNA lentiviral vector to achieve METTL3 knockdown. Cells cultured or not cultured under anoikis conditions. C, D Wound healing assay was performed in PHGDH overexpress MHCC-97 H (C) and SK-Hpe1 (D) cells treated with METTL3 shRNA lentiviral vector to achieve METTL3 knockdown (n = 3 independent experiments), scale bar = 200 μm. E, F Transwell assay was performed in PHGDH overexpress MHCC-97 H (E) and SK-Hpe1 (F) cells treated with METTL3 shRNA lentiviral vector (n = 3 independent experiments), scale bar = 100 μm. G Co-IP experiment to detect the interaction between PHGDH-V425M and METTL3 in MHCC-97 H cells transfected with the HA-PHGDH-V425M and Flag-METTL3. H Western blot analysis of E-cadherin and Snail expression in MHCC-97 H cells transfected with the HA-PHGDH-V425M. Subsequently, shRNA was used to achieve METTL3 knockdown. I Flow cytometry was used to detect the apoptosis rates of MHCC-97 H cells transfected with PHGDH-WT and V425M under anoikis conditions. P-values were derived from one-way ANOVA followed by Tukey’s test. Data are represented as the mean ± SD, **P < 0.01, ***P < 0.001

To elucidate the contribution of PHGDH’s enzymatic activity to its interaction with METTL3 and its role in HCC metastasis, we investigated the effects of a catalytically inactive PHGDH mutant (PHGDH V425M) [30, 31] on the interaction between PHGDH and METTL3. Our results indicated that the PHGDH V425M mutant maintained the ability to interact with METTL3 (Fig. 4G). Moreover, we evaluated the influence of the PHGDH V425M on the expression levels of METTL3 downstream molecules. The data revealed that the PHGDH V425M exerted effects on the expression of METTL3’s downstream molecules in a manner akin to that of the wild-type PHGDH (Fig. 4H). Notably, PHGDH V425M mutant exhibited a partial impact on the apoptosis rate, suggesting that PHGDH may exert its effects on anoikis via a dual mechanism involving both its canonical enzymatic function and additional non-canonical roles (Fig. 4I). Above all, these results indicate that PHGDH regulates downstream molecules of METTL3 and promotes HCC metastasis in a manner that is independent of its enzymatic activity, highlighting the multifaceted nature of PHGDH’s role in HCC.

PHGDH promotes HCC metastasis via METTL3

To evaluate the roles of PHGDH and METTL3 in metastasis, we developed HCC lung metastasis model in nude mice via tail vein injection of human HCC cell lines (Fig. 5A). Images and H&E staining of lung metastatic tissues demonstrated that PHGDH overexpression led to a significant increase in the number of metastatic nodules in the lungs of nude mice, indicating that PHGDH enhances the metastatic potential of HCC. In contrast, overexpression of the PHGDH-RD mutant did not result in a significant increase in lung metastasis nodules. Furthermore, METTL3 knockdown in PHGDH-overexpressing cells significantly reduced the number of lung metastatic nodules, implicating METTL3 as a mediator of PHGDH-induced HCC metastasis (Fig. 5B-D). Analysis of TCGA database revealed a strong correlation between elevated PHGDH expression and adverse prognosis in HCC, with high METTL3 expression showing a similar trend. Notably, the co-overexpression of PHGDH and METTL3 correlated with the worst prognosis in HCC patients (Fig. 5E). Additionally, the expression levels of anoikis-related genes were relatively higher when METTL3 expression was high (Fig. 5F). These findings corroborate our in vitro results, demonstrating that PHGDH enhances HCC metastasis by modulating anoikis and EMT through METTL3.

Fig. 5.

PHGDH promotes HCC metastasis via METTL3. A Schematic diagram of a model of lung metastasis injected with tail vein injection in nude mice. B Representative lung imagine of a nude mouse lung metastasis model (n = 6 per group). C Hematoxylin-eosin (HE) staining of nude mouse lung metastases (n = 6 per group), scale bar = 1000 μm. D Quantitative analysis of tumor nodules in nude mouse lung metastasis model (n = 6 per group). E Survival analysis based on the Cancer Genome Atlas (TCGA)-HCC database was used to evaluate the prognostic survival rate of HCC patients, for PHGDH, METTL3, or both PHGDH and METTL3. The Log-Rank Test was used for analyzing differences in prognosis. F Relative expression analysis of METTL3 and anoikis-related genes in the TCGA-LIHC database using the log-rank test. P-values were derived from two-way ANOVA followed by Tukey’s test in (D). Data are represented as the mean ± SD, **P < 0.01, ***P < 0.001

Discussion

Metabolic reprogramming is an adaptive response of tumor cells to abnormal microenvironmental signals. The modulation of metabolic enzyme activity not only satisfies the demands for biosynthesis and energy production but also plays a significant role in tumor progression [2, 32, 33]. The regulatory functions of metabolic enzymes extend beyond their classical enzymatic roles to encompass non-enzymatic activities. Here, our study sheds new light on the moonlight activity of the key metabolic enzyme PHGDH in the serine synthesis pathway. We discovered that PHGDH interacts with METTL3, inhibiting its ubiquitination and subsequent degradation, thereby increasing METTL3 protein levels. Stabilization of METTL3 subsequently enhances the expression of EMT-related molecules, ultimately suppressing anoikis and promoting HCC metastasis. Our findings provide new insights into the multifunctional roles of PHGDH in HCC metastasis and its potential key role in anoikis resistance.

PHGDH, a pivotal metabolic enzyme in the serine synthesis pathway, is frequently overexpressed in various tumors and contributes to malignant progression through multiple mechanisms [11, 34–36]. Beyond its canonical role in serine synthesis, which promote tumor growth, PHGDH also exhibits moonlighting functions, such as alterations in subcellular localization and engagement in protein-protein interactions [15–17, 37]. Previous studies have demonstrated that PHGDH can facilitate HCC metastasis by elevating serine levels through its canonical enzymatic activity [21, 22]. However, whether PHGDH influences HCC metastasis through its moonlighting functions remains unclear. Our research found that PHGDH expression was significantly upregulated under anoikis conditions, where it interacted with METTL3 to affect the expression of downstream EMT molecules, thereby inhibiting anoikis and promoting the progression of HCC cell metastasis. This is the first study to demonstrate that PHGDH drives HCC metastasis through protein-protein interactions, underscoring its critical role in HCC progression. Targeting the PHGDH-METTL3 axis specifically could be a viable therapeutic approach for a broad spectrum of HCC patients, not limited to those resistant to certain chemotherapeutic drugs. Meanwhile, we founded that the enzymatic activity of PHGDH is dispensable for its interaction with METTL3 and for the regulation of downstream molecules. However, the loss of PHGDH’s enzymatic activity partially attenuates its ability to suppress anoikis, indicating that both its catalytic and non-canonical functions of PHGDH may contribute to the regulation of anoikis. Additionally, it is still unclear whether PHGDH affects the RNA modification function of METTL3 and thereby regulates its downstream target genes. Further analysis is needed to elucidate the impact of the interaction between PHGDH and METTL3 on the gene-regulating function of METTL3. Furthermore, it is imperative to conduct further experiments in more advanced animal models, such as gene knockout mice for PHGDH and METTL3, to confirm the role of PHGDH in promoting HCC metastasis through its interaction with METTL3. Additionally, future studies can utilize animal models such as Patient-Derived Xenograft (PDX) to more accurately reflect the human HCC metastatic process.

Anoikis is a specific form of programmed cell apoptosis triggered by detachment from the extracellular matrix, playing a crucial role in tumor metastasis [23, 24]. Although PHGDH has been implicated in metastasis, the connection between PHGDH and anoikis has not been previously reported. Our study identified that PHGDH expression is upregulated under anoikis conditions, and we further discovered that PHGDH overexpression inhibited the apoptosis rate in these conditions. METTL3 is associated with metastasis, and its knockout can reduce resistance to anoikis [38]. our research confirmed a close relationship between METTL3 and anoikis in HCC cells, showing that METTL3 knockout enhances the apoptosis rate under anoikis conditions, which is otherwise reduced by PHGDH overexpression. However, further research is necessary to elucidate the specific mechanisms through which PHGDH affects anoikis and to ascertain whether its overexpression confers enhanced anoikis resistance in sensitive HCC cell lines. Additionally, it is imperative to further dissect the mechanisms by which PHGDH’s enzymatic activity impacts anoikis, in order to broaden our understanding of the multifaceted regulatory roles of PHGDH in this process.

Previous studies have shown that PHGDH can promote tumor proliferation and invasion by interacting with the N-terminal of FOXM1, thereby stabilizing FOXM1 through inhibiting its ubiquitination-induced degradation [39]. Similarly, our study found that PHGDH inhibits METTL3 ubiquitination and degradation through protein-protein interaction, thereby enhancing METTL3 protein levels in HCC. Additionally, we have confirmed that PHGDH interacts with SPOP, thereby attenuating the ubiquitination and degradation of METTL3, a finding that aligns with previous research [40]. However, to fully elucidate the complex mechanisms by which PHGDH regulates ubiquitination and degradation, further comprehensive studies are necessary to explore the intricate interactions among PHGDH, SPOP, and METTL3. Additionally, whether other E3 ubiquitin ligases were involved remains to be confirmed through additional experiments.

PHGDH’s multifaceted role in the development of various tumors is well-established. Our study elucidates a novel mechanism by which PHGDH promotes HCC metastasis through METTL3, thereby expanding the understanding of PHGDH’s critical involvement in HCC metastasis. This research contributes new evidence to the complex repertoire of metabolic enzymes in tumor progression.

Materials and methods

Cell culture and treatment

The HCC cell lines were maintained in Dulbecco’s modified Eagle’s medium at 37℃ in 5% CO₂. The medium was enriched with 10% fetal bovine serum, and 100 units/mL penicillin, 100 µg/mL streptomycin, all sourced from HyClone, USA. Huh7, MHCC-97 H and HEK293 cells were acquired from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China), and PLC/PRF/5 and SK-Hep1 cells were procured from the American Type Culture Collection in Virginia, USA.

Adenoviruses and plasmids

The PHGDH gene was cloned into the pAdTrack-TO4 plasmid, provided by Dr. T-C He from the University of Chicago, to create the adenoviral recombinant AdPHGDH using the AdEasy system (Flag tag was added the to the TO4 vector). The AdEasy system is widely used for creating and packaging recombinant adenoviruses, offering high infectivity, efficient concentration, and safety through E1 gene inactivation [41, 42]. AdGFP, an adenovirus expressing only green fluorescent protein, served as the control. The full-length PHGDH, PHGDH-V425M and truncated mutants (SBD1, SBD2 + RD, SBD1 + NBD, and RD) were subcloned into the pBU-3HA plasmid. METTL3 truncated mutants (ΔN, ΔC) were inserted into the pAdTrack-TO4 plasmid. The shMETTL3 sequence was designed by Life Technologies (https://rnaidesigner.thermofisher.com/rnaiexpress/sort.do) and constructed into the LentiLox 3.7 plasmid, part of the shRNA system, with shCon used as a negative control. The LentiLox 3.7 plasmid was provided by Professor Bing Sun from the Center for Excellence in Molecular Cell Science, Chinese Academy of Sciences.

Western blotting

Cells were washed with PBS and then lysed using cell lysis buffer (P0013, Beyotime Biotechnology, China) containing protease inhibitors (C0001, TargetMol, China, 1:1000 dilution). Subsequently, the lysate was incubated for 15 min and centrifuged at 4 °C. Protein samples were prepared by adding the appropriate protein loading buffer, then separated by SDS-PAGE, and subsequently transferred to PVDF membranes. The membranes were blocked with 5% milk to prevent nonspecific binding and incubated with primary antibodies (PHGDH, 14719-1-AP, Proteintech, 1:1000 dilution; METTL3, ab195352, abcam, 1:1000 dilution; Snail, ab224731, abcam, 1:1000 dilution; E-cadherin, ab314063, abcam, 1:1000 dilution) overnight at 4 °C (12–16 h). On the following day, the membranes were washed with Tris-Buffered Saline with Tween-20 (TBST) and incubated with the corresponding secondary antibodies at room temperature for 1.5 h. After removing excess secondary antibodies, the protein bands were visualized using Clarity™ Western ECL Substrate (Bio-Rad Laboratories).

Immunoprecipitation

Huh7 and PLC/PRF/5 cells, as well as HEK293 cells transfected with HA-PHGDH and Flag-METTL3 plasmids were washed twice with PBS, then the collected cells were lysed with cell lysis buffer (with protease inhibitor added) at 4℃ for 30 min, centrifuge to collect the protein solution and add the corresponding tag antibody (HA-Tag, 3724, Cell Signaling Technology, 1:2000 dilution; Flag, F3040, sigma, 1:4000 dilution), incubating overnight at 4 °C (12–16 h). The next day, pre-washed protein A/G magnetic beads (HY-K0202, MedChemExpress, USA) were added to the protein-antibody mixture and incubate at 4℃ for 4 h. Finally, the magnetic beads 6 times were washed with PBST, add loading buffer to prepare protein samples for subsequent experiments.

Detection of protein ubiquitination

We used the Ni-NTA method, a metal affinity chromatography technique, to detect METTL3 ubiquitination by purifying His-tagged proteins with Ni²⁺ beads, which were then analyzed by Western blot [43]. After culturing for 30 h, HCC cells were treated with 20 µM MG132 (HY-13259, MedChemExpress, USA) for an additional 6 h. The HCC cells were washed twice with PBS, scraped into 1.5 ml EP tubes, and Buffer A with protease inhibitors was added. The HCC cells were sonicated (power 25%, 2 s per pulse, 10 pulses total), and 10% of the lysate was reserved as input. Concurrently, 20 µl of Ni-NTA magnetic beads (HY-K0241, MedChemExpress, USA) were pretreated by washing twice with equilibration buffer. The sonicated protein lysate was added to the beads and incubated with rotation at 4 °C overnight. After incubation, the supernatant was discarded, and the beads were washed once with 500 µl Buffer A, followed by two washes with Buffer A/Buffer Ti (1:3 ratio). The beads were then eluted with Buffer Ti. Finally, 60 µl of loading buffer was added to prepare the protein samples.

Immunofluorescence

After a 25 min fixation with 4% paraformaldehyde, the PLC/PRF/5 and Huh7 cells were permeabilized using 0.2% Triton X-100. Non-specific binding sites were blocked with 5% bovine serum albumin (BSA), the cells were incubated with anti-METTL3 (rabbit polyclonal, 1:500 dilution) and anti-PHGDH (mouse monoclonal, 1:500 dilution) at 4 °C for 12 h. After being rinsed twice with PBS, the cells were labeled with secondary antibodies conjugated to Alexa Fluor 488 or Alexa Fluor 594. The cell nuclei were stained with 4’,6-diamidino-2-phenylindole (DAPI). A laser scanning confocal microscope (Leica Microsystems, Germany) was employed to obtain the immunofluorescence images.

Animal studies

BALB/c nude mice (4-week-old, male) were used for the HCC orthotopic metastatic model. 2 × 10⁶ MHCC-97 H cells suspended in PBS were injected into the mice via the tail vein (n = 6 per group). The mice were sacrificed at 12 weeks of age, and lung tissues were used for Hematoxylin-Eosin (HE) staining, HE staining was performed by Chengdu Lilai Biotechnology Co., Ltd.

Flow cytometry

MHCC-97 H cells were seeded into 6-well plates and infected with AdPHGDH/AdGFP and shCon/shMETTL3. After 48 h, the cells were digested and collected into EP tubes, followed by two PBS washes. Cells were then resuspended in 200 µl of 1 × Binding Buffer, mixed thoroughly, and stained with 10 µl of PI dye (FXP018Pro-100, 4 A Biotech, China). The mixture was incubated in the dark at 4 °C for 15 min. Finally, 300 µl of 1 × Binding Buffer was added, and flow cytometry was performed within one hour.

Glutathione S-transferase (GST) pull-down assay

Recombinant GST-tagged METTL3 and His-tagged PHGDH proteins were mixed with glutathione-sepharose beads (GE Healthcare, Piscataway, NJ, USA) and allowed to interact for a duration of 3 h. Post-incubation, the beads underwent washing with a buffer solution (containing 137 mM NaCl, 2 mM KH2PO4, 10 mM Na2HPO4, 2.7 mM KCl and 0.5% Triton X-100). Subsequently, the beads were resuspended with a 2× sample buffer and heated at boiling point for 8 min. The samples were then analyzed using Coomassie Brilliant Blue staining for protein visualization and Western blotting for specific protein detection.

Transwell migration assays

Huh7 and PLC/PRF/5, MHCC-97 H, SK-Hep1 cells were introduced into the upper compartment of Transwell filters containing serum-free medium. Following 24 h of incubation, the cells were washed with PBS and fixed using a 4% paraformaldehyde solution. Subsequently, the migrated cells on the lower surface of the filters were stained with a crystal violet solution (Beyotime Biotechnology), while the non-migrated cells in the upper compartment were carefully removed using a cotton swab. After air drying the filters in an inverted position, the migrated cells were observed and counted under a microscope. The number of migrated cells was quantified in five random microscopic fields for each experiment, and this procedure was repeated in three separate experiments.

Wound healing assay

HCC cells were seeded into a 96-well plate. Once the cells reached confluence, a scratch was made using a Wound Maker (Essen Bioscience, USA). The wells were then washed with medium to remove any dislodged cells and supplemented with complete medium. The cells were incubated for 48 h, with real-time imaging performed using an IncuCyte ZOOM Live-Cell Imaging system (Essen Bioscience) at 0, 12, 24, and 36 h.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 8.0. Experimental results are presented as mean ± standard deviation (SD) from a minimum of three biological replicates. An unpaired or paired 2-tailed Student’s t-test was employed for comparing two groups, one-way or two-way ANOVA for comparing single-factor or two-factor multiple groups, and log-rank test applied to analyze the prognostic implications. p-value < 0.05 was considered statistically significant (*p < 0.05; **p < 0.01; ***p < 0.001).

Author contributions

Bin Cheng and Jing Ma completed the main experiments of this study and wrote the main manuscript text, and Rui Liu completed part of the research work. Ni Tang, Pai Peng and Kai Wang guided the whole process of the research work and revised the article. All authors reviewed the manuscript.

Funding

This work was supported by the Innovative and Entrepreneurial Team of Chongqing Talents Plan, Chongqing Medical Scientific Research Project (Joint project of Chongqing Health Commission and Science and Technology Bureau, 2023DBXM007), the Natural Science Foundation Project of Chongqing (CSTB2024NSCQ-MSX0193), the Senior Medical Talents Program of Chongqing for Young and Middle-aged, and the Kuanren talents and Joint Project of Pinnacle Disciplinary Group, the Second Affiliated Hospital of Chongqing Medical University.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethical approval

The animal experiment protocol for this project was reviewed by the Laboratory Animal Management and Use Committee of Chongqing Medical University (IACUC-CQMU-2023-0186) and complies with animal protection, welfare, and ethical principles.

Competing interests

The authors declare no competing interests.