SHMT2 inhibition triggers mitochondrial apoptosis to suppress lung adenocarcinoma progression

Abstract

Background

Mitochondrial dysfunction has been implicated in various cancers, including non-small cell lung cancer (NSCLC). This study aimed to elucidate the role of mitochondrial dysfunction in NSCLC prognosis and the molecular mechanisms involved, particularly focusing on the Serine hydroxymethyltransferase 2 (SHMT2), a key enzyme in mitochondrial metabolism.

Methods

We analyzed the expression of SHMT2 in NSCLC tissue and its prognostic correlation by using The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus Series (GSE81809) databases. Further, the CRISPR-Cas9 system was employed to establish SHMT2-knockout lung adenocarcinoma (LUAD) cell lines, assessing its impact on cell proliferation and apoptosis through various in vitro assays and in vivo nude mouse xenograft models. Mitochondrial homeostasis was evaluated by assessing the ROS levels, mitochondrial morphology, and membrane potential. Meanwhile, the SHMT2 inhibitor SHIN1 and the ROS scavenger N-Acetylcysteine (NAC) were employed to validate the underlying mechanism.

Results

The results revealed that SHMT2 was significantly overexpressed in LUAD tissues and was associated with poor patient prognosis. SHMT2-knockout as well as SHIN1-treatment significantly inhibited the proliferation of LUAD cells and induced ROS-dependent mitochondrial apoptosis. Mechanistically, SHMT2 deficiency could increase the recruitment of BAX to mitochondria, reducing ΔΨm, and release of Cytochrome C (Cyto C) from mitochondria, thereby activating the caspase cascade and initiating intrinsic mitochondrial apoptosis. NAC treatment can reverse the apoptosis and mitochondrial dysfunction induced by SHMT2 knockout. In vivo experiments further confirmed that SHMT2 knockout significantly inhibited tumor growth, whereas ROS scavenging attenuated its antitumor effects.

Conclusions

Our findings suggest that SHMT2 is vital for regulating LUAD apoptosis by maintaining mitochondrial ROS homeostasis, and its deficiency triggers apoptosis through the ROS-BAX-Cyto C-Caspase signaling. Targeting SHMT2 could thus offer new clinical insights and be a promising strategy for LUAD treatment.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-026-07975-9.

Background

Lung cancer remains the leading cause of cancer-related mortality worldwide, with increasing incidence and mortality rates [1, 2]. Lung cancer can be classified according to the histological types, mainly divided into small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC) [3]. About 85% of the lung cancer cases are non-small cell type, including lung adenocarcinoma (LUAD), lung squamous carcinoma (LUSC), and large cell carcinoma (LCC) [4, 5]. LUAD is the most common subtype of lung cancer, which is characterized by high morbidity and mortality [6]. Currently, the strategies for the treatment of LUAD encompass traditional radiotherapy, chemotherapy, surgical resection and targeted therapy. Nevertheless, none of the above therapies achieved desired therapeutic efficacy especially for the patients with distant metastasis, the survival rate remains at ~ 14.3% [7–10]. Therefore, to discover novel therapeutic targets for LUAD is urgently needed.

In recent years, mitochondria-related studies have become a prominent area of focus in the field of cancer research [11]. Mitochondria may play pivotal roles in the development of tumors, orchestrating essential metabolic processes and signaling pathways that regulate cellular energy homeostasis, redox equilibrium, and apoptosis. Previous studies have suggested that dysregulation of mitochondrial metabolic enzymes can disrupt redox homeostasis, leading to a pronounced oxidative shift characterized by cumulative oxidative damage to mitochondrial constituents, compromised membrane integrity, initiation of apoptotic pathways, and eventual cell death [12]. Recent research in 2024 has demonstrated that the imbalance in mitochondrial dynamics critically contributes to lung cancer progression via mitochondrial apoptosis [13]. The pathogenesis of mitochondrial homeostasis is strongly associated with the excessive accumulation of mitochondrial reactive oxygen species (ROS), which overwhelms the mitochondria’s antioxidant defense ability [14]. ROS play a vital role in regulating mitochondrial functions. The accumulation of ROS can induce damage to the mitochondria, leading to alterations in mitochondrial membrane integrity, fragmentation, and dysfunction. Additionally, ROS are able to trigger the liberation of pro-apoptotic agents i.e. Cytochrome C and AIF, from the intermembrane space of mitochondria to the cytoplasm. These processes can lead to the activation of apoptosis-related signaling pathways, ultimately aggravating tissue damage [15]. Apoptosis can be divided into extrinsic and intrinsic types. The extrinsic apoptosis-related pathway is activated by the binding of extracellular ligands to the transmembrane death receptors; On the other hand, intrinsic apoptosis, also known as mitochondrial apoptosis, is initiated by irreversible mitochondrial outer membrane permeabilization (MOMP). This critical event is primarily regulated by the Bcl-2 protein family, with pro-apoptotic members Bax and Bak playing essential roles in mediating pore formation [16]. The above evidence suggests that mitochondrial dysfunction is closely related to tumor growth [17], and targeting the key enzymes involved in the mitochondrial metabolism may represent a novel approach for the management of cancers [18].

Serine hydroxymethyltransferase 2 (SHMT2) is a notable metabolic enzyme in the process of serine to glycine conversion that predominantly happened in mitochondria. Serine and glycine, as key one-carbon donors in one-carbon metabolism, support the synthesis of essential cellular components, including nucleic acids, proteins, and lipids. They also regulate the NADPH/NADP+ ratio and glutathione biosynthesis to sustain cellular redox balance [19, 20]. The isoenzyme of SHMT2, SHMT1 mainly plays a role in the cytoplasm, and the two isoforms of serine hydroxymethyltransferase family display an amino acid sequence identity of approximately 66% [21]. SHMT2, but not SHMT1, has been reported to be overexpressed in various tumors, associated with their progression, and potentially involved in processes of tumor cell migration, invasion, proliferation, and apoptosis [22, 23]. Moreover, SHMT2 has also been reported to be involved in the process of maintaining the ROS balance, genetic material synthesis, DNA repair, RNA translation, epigenetic changes, and redox defense [24, 25]. SHMT2 serves as a crucial oncogenic factor that promotes tumor progression, contributes to unfavour clinical outcomes, and enhances therapeutic resistance in various cancers [26–29]. Previous studies have demonstrated the oncogenic roles of SHMT2 in a variety of human cancers, including glioma [30], colorectal cancer [31], breast cancer [27], and pancreatic cancer [32]. SHMT2 may regulate mitochondrial function in different types of cancer. For example, hypoxia-inducible factor-1 can increase the expression of SHMT2 under hypoxic conditions, accelerate serine catabolism, and reduce intracellular production of ROS, thereby improving the survival of neuroblastoma and malignant glioma cells. Conversely, depletion of SHMT2 in cancer cells has been associated with mitochondrial dysfunction, causing an increase in intracellular ROS levels, and ultimately inducing tumor cell death [23]. These findings indicate that SHMT2 may play a significant role in regulating tumor cell metabolism and survival. In addition, the inhibition of SHMT2 has been shown to suppress the growth of various tumor cells, indicating that SHMT2 has the potential to become a molecular target for tumor therapy [20–33]. The roles of SHMT2 in lung cancer have also been reported [34, 35]. SHMT2 may serve as an independent prognostic marker for LUAD, and increased expression of SHMT2 is associated with poor patient prognosis [34–36]. These observations led us to investigate the potential role of SHMT2 in LUAD progression, with particular focus on its regulatory effects on mitochondrial function.

Based on the previous findings that SHMT2 is overexpressed in LUAD and linked to poor prognosis, we designed this study to investigate the functional role of SHMT2 in LUAD progression, with a specific focus on its regulatory effects on mitochondrial function. Our study systematically elucidated the molecular mechanisms underlying SHMT2 deficiency-induced ROS accumulation and mitochondrial apoptosis regulation in LUAD. These findings strongly suggest that SHMT2 represents a promising therapeutic target for LUAD intervention.

Methods

Clinical NSCLC patient specimens

NSCLC tumor tissues and paired paracancerous tissues were harvested from patients who were histopathologically diagnosed with NSCLC at the First Affiliated Hospital of Soochow University between 2015 and 2023. Before surgery, None of the patients had received any chemotherapy or radiotherapy. The tissue samples were immediately stored at -80℃ after surgery. All the clinical studies were conducted with relevant ethical regulations including the Declaration of Helsinki and were approved by the Ethics Committee of the First Affiliated Hospital of Soochow University (Ethics Code 2024662), and informed consent was obtained from each patient.

Bioinformatics

We downloaded the mRNA expression data and also the clinical information of LUAD (589 samples, 530 tumor samples and 59 normal samples) and LUSC (502 samples, 501 tumor samples and 51 normal samples) patients from The Cancer Genomes Atlas (TCGA) database through the GDC Data Portal (https://protal.gdc.cancer.gov). The GSE81089 datasets (containing clinical information of the patients) was obtained from the Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/) database. The Wilcoxon rank sum test or Student’s two-sided t test was used to compare the significance of the differences between two groups. The raw count data were converted to log2(TPM + 1) by using the “count2tpm” function from the R package “IOBR” (v0.99.8), and subsequently standardized. The mathematical principle of this standardization (Z-score) is as follows:

For a vector , the Z-score for each element is calculated as:

where: is the standardized value (Z-score) for the -th element. is the original value for the -th element. is the mean of all values in the vector. is the standard deviation of all values in the vector.

The overall survival (OS) data was grouped into categories of 7.5 years, with any data points exceeding this time frame being excluded from the analysis. The remaining data was then utilized for KM and Cox analysis. For Kaplan-Meier survival analysis, the “High” group comprised the top 25% of expression values, while the “Low” group comprised the bottom 25%. KM analysis and the log-rank test were performed to assess the statistical significance of differences between the high-SHMT2 and low-SHMT2 groups by function “survfit” from R packages “survival” (v3.7-0). For the Cox analysis, the optimal cutpoint for the standardized SHMT2 level was determined using the “surv_cutpoint” function from the R package “survminer” (v0.5.0). This cutpoint was then used to dichotomize patients into “High” and “Low” expression groups. Independent prognostic factors were determined by univariate and multivariate Cox proportional-hazards model analysis and were provided by function “coxph” from R packages “survival” (v3.7-0). All statistical analyses were performed using R software (version 4.4.1).

Cell culture

The immortalized human bronchial epithelial cell line BEAS-2B, and the LUAD cell lines H838, A549, H1299, H358, and H1650 were purchased from Procell Life Science & Technology Co., Ltd. (Wuhan, China). The cells were grown in RPMI-1640 medium (HyClone, South Logan, UT, USA) supplemented with 1% penicillin-streptomycin (Beyotime, Shanghai, China) and 10% fetal bovine serum (FBS; Gibco, Carlsbad, CA, USA). Culturing was carried out at 37 °C in a humidified incubator with 5.0% CO₂. All these cell lines were passaged within 6 months.

Construction of SHMT2 knockout LUAD cells

SHMT2 knockout LUAD cell lines were generated using the CRISPR-Cas9 system. SgRNA sequences targeting human SHMT2 were designed by an online gRNA design tool (https://crispor.gi.ucsc.edu/). The guide sequences targeting exon 2 of the human SHMT2 gene (5’-CACCGTCTCAGGATCACTGTCCGAC-3’; 5’-CACCGGACAGGCAGTGTCGTGGCC-3’) were synthesized, subcloned into the LentiCRISPR v2 vector (Adddene #52961) and then sequenced by GENEWIZ Biotechnology (Suzhou, China). Then the constructed Sg-SHMT2 and the control plasmids were co-transfected in HEK293T cells with VSVG, REV, and MDL plasmids to package lentivirus. Lentivirus containing SHMT2-targeting constructs were transfected into A549 and H1299 cells, followed by the selection with 2.0 µg/mL puromycin (Beyotime Biotechnology, Shanghai, China) to establish the stable SHMT2-knockout cell lines.

Western blot

Protein samples from lung tissues and cells were extracted, and then sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was applied to separate the protein lysates. The samples were transferred to a nitrocellulose membrane (Merck Millipore Ltd., Tullagreen, Carrigtwohill, Co. Cork, Ireland). After 2 h of membrane transfer, immunoblotting was performed with specific primary antibodies at 4 °C overnight. On the next day, immunoblotting was performed with the corresponding secondary antibody at room temperature for 2 h. Then, Western Lightning ECL reagent (New Cell & Molecular Biotech Co., Ltd, Shanghai, China) was used for coloration, and the β-actin or GAPDH protein level was used as a control. The brands and concentrations of the antibodies used are shown in Supplementary Table S1.

CCK-8 (Cell Counting Kit-8) assay

Parental or SHMT2 knockout A549/H1299 cells were plated in 96-well plates with each well containing 3.0 × 10³ cells, and each experimental group contained three replicate wells. Following the treatment of parental cells with various concentrations of inhibitors (NAC, Fer-1, NEC, TTM and Z-VAD) for 24 h, 10 µL of CCK-8 (lot: K1018, APExBIO) was added to each well. After a 2 h incubation period, the absorbance at 450 nm of each well was assessed via a microplate reader (Thermo Scientific, Massachusetts, USA).

Colony formation assay

A total of 3000 cells were inoculated in 2 mL of RPMI 1640 medium containing 10% fetal bovine serum in 6 cm dishes. The cells were cultured for 7 days at 37 °C and 5% CO₂, fixed with methanol for 30 min, and then stained with 0.1% crystal violet (Beyotime, Shanghai, China) for 12 h. After being gently washed twice with PBS, the number of formed clones was photographed and counted.

EdU assay

The cells from the 6-well plates were harvested and then transferred to 96-well plates at a concentration of 3,000 cells per well. Following 24 h of incubation, the cells were exposed to medium containing EdU (20 µM/mL) for 2 h. Subsequently, the medium was aspirated, and the cells were rinsed twice with PBS. The cells were then fixed with prechilled cell fixative (4% paraformaldehyde) for 30 min at room temperature. After fixation, the cells underwent three washes with PBS containing 0.3% BSA to eliminate any residual fixative. The samples were then treated with penetrating reagent (PBS containing 0.9% Triton X-100) for 15 min and washed 3 times again with the washing solution, after which 50 µL of the CLICK reaction solution was added. Following 30 min of incubation at room temperature in the dark, the samples underwent another three washes with the cleaning solution. Subsequently, the nuclei were stained with Hoechst solution (lot: 3342, Beyotime, 10 µg/mL) and incubated in the dark at room temperature for 10 min. After three additional washes with a washing solution, cell proliferation was observed, and images were captured via an inverted fluorescence microscope.

Detection of the intracellular and mitochondrial ROS production

DCFH-DA (lot: D6470, Solarbio, 10 µM) and MitoSox Red (lot: 309704, MedChemExpress, 5 µM) were used to detect ROS production. Briefly, cells were seeded in special confocal dishes (BS-15-GJM, Biosharp Life Science), washed 3 times with preheated PBS, and then loaded with 5 µM MitoSox Red or 10 µM DCFH-DA at 37 °C for 30 min. After being washed with PBS for three times, cells were dyed with Hoechst (lot: 3342, Beyotime, 10 µg/mL) at room temperature for 10 min. The fluorescence intensity was observed via a confocal fluorescence microscope (Nikon Eclipse Ti, Tokyo, Japan) and quantified using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

NADPH/NADP+ determination

According to the manufacturer’s instructions, the ratio of NADPH /NADP + was assayed via a NADPH /NADP+ assay kit with WST-8 (S0180S, Beyotime Biotechnology, China). Briefly, 1 × 10⁶ cells were seeded and cultured in six-well plates for a set amount of time. Aspirate the culture medium and fill each well with 200 µL of NADPH/NADP+ extracting solution. Then, 200 µL of extraction solution was added, and the extracts were centrifuged at 12,000 × g and 4 °C for 10 min to collect the supernatant. To detect NADPH, NADP+ needs to be decomposed before the reaction. Aliquots containing 200 µL of extracted samples were placed into new microcentrifuge tubes and heated to 60 °C for 30 min to decompose NADP+. The supernatant (50 µL, either unheated or heated to decompose NADP⁺) was added to a 96-well plate. Thereafter, a reaction mixture containing 98 µL of NADPH cycling buffer and 2 µL of NADPH cycling enzyme mix was added, and 100 µL of the reaction mixture was added to each well containing a test sample. The plate was incubated at room temperature for 10 min, after which 10 µL of NADPH developer was added to each well. The plate was then incubated at room temperature for 1 h. The absorbance was measured via a microplate reader (Thermo Scientific, Massachusetts, USA) at 450 nm. Standard curves of gradient concentrations of NADPH and NADP+ were simultaneously determined, and the concentrations of the samples were calculated according to the standard curves.

GSH/GSSG determination

The intracellular concentrations of all glutathione and oxidized glutathione (GSSG) were measured via a GSH/GSSG detection kit (S0053, Beyotime Biotechnology, China) following the manufacturer’s instructions. In brief, the cells were resuspended in protein-removal reagent M, thoroughly vortexed, and lysed via 2–3 rapid freeze–thaw cycles alternating between liquid nitrogen and a 37 °C water bath. The supernatant served directly as the total GSH sample. For the GSSG samples, glutathione (GSH) was removed from the supernatant via the sequential addition of GSH clearing auxiliary liquid and GSH clearing reagent. Both sample types were incubated with GSH detection working solution (which contains GSH reductase and DTNB) at 25 °C for 5 min and then incubated with NADPH at 25 °C for 30 min. Finally, the absorbance at 412 nm was read via a multifunctional microplate reader (Thermo Scientific, Massachusetts, USA). The total GSH content was defined as GSH + GSSG × 2, so GSH = total GSH − GSSG × 2. The intracellular GSH status was reflected by the GSH/GSSG ratio.

Assessment of mitochondrial morphology

Mito-Tracker Deep Red FM (lot: C1032, Beyotime Biotechnology, 200 nM) and PK MITO Deep Red (lot: PKMDR-2, PKMITO®/Genvivo, 125 nM) were used for the observation of mitochondrial morphology. The cells were seeded in confocal dishes (BS-15-GJM, Biosharp Life Science; China), and the density of the cells during the detection period was maintained at 40–50% confluence. After treatment, the cells were washed with preheated PBS for three times and then incubated with the aforementioned two mitochondrial dyes at 37 °C for 30 min. Subsequently, the cells were rinsed three times with PBS to eliminate excess dye, followed by Hoechst staining (lot: 33342, Beyotime; 10 µg/mL) at room temperature for 10 min. The morphology of the mitochondria was visualized via a confocal fluorescence microscope (SP8, Leica Microsystems, Germany).

Transmission electron microscopy (TEM)

SHMT2-knockout H1299 cells and negative control cells were cultured and collected. The culture medium was discarded after centrifugation, and the precipitated cells were prefixed with TEM fixative (G1102, Servicebio, China) for 4 h at 4 °C. Then the harvested cells were washed three times with 0.1 M PBS (pH 7.4), preembedded in 1% agarose, postfixed with 0.1 M PBS containing 1% osmium tetroxide at room temperature for 2 h, and rinsed three times with 0.1 M PBS (pH 7.4) for 15 min each time. After dehydration in 30%, 50%, 70%, 80%, 95%, 100% and 100% alcohol, the cells were embedded in pure EMBed 812 (90529-77-4, SPI). The samples were then cut into 60–80 nm thick sections and stained with uranium-lead solution. Following drying overnight at room temperature, the mitochondrial morphology of the cells was observed via transmission electron microscopy (HITACHI) and photographed by Servicebio Technology Co., Ltd.

JC-1 staining

The mitochondrial membrane potential was measured with a JC-1 detection kit (C2006, Beyotime). The treated cells were incubated with JC-1 staining solution at 37 °C for 20 min, and then washed twice with the JC-1 buffer solution. Images of the JC-1 aggregates and monomers were captured via a confocal microscope (Nikon Eclipse Ti; Nikon, Japan). JC-1 aggregates in normal mitochondria emit red fluorescence, whereas JC-1 monomers in unhealthy mitochondria emit green fluorescence.

TUNEL staining

The cells were processed according to the manufacturer’s protocol for the colorimetric TUNEL apoptosis assay kit (C1090, Beyotime). Briefly, the cells were washed and incubated with 50 µL of recombinant cyanine-12-dUTP cocktail for 1 h at 37 °C in a fully-humidified chamber. Then the reaction was terminated, and the samples were washed three times in PBS and stained with Hoechst (lot: 3342, Beyotime, 10 µg/mL) at room temperature for 10 min. The degree of apoptosis of the cells was quantitatively evaluated by red fluorescence intensity via ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Flow cytometry assay

The SHMT2 stable cells and SHIN1-treated A549/H1299 cells were collected from 6-well plates, and then resuspended in 500 µL of binding buffer containing 5 µL of annexin V-FITC and 5 µL of propidium iodide (lot: K2003, APExBIO), and then incubated at room temperature in the dark for 30 min. A CytoFLEX™ flow cytometer (Beckman Coulter, Brea, CA) was used for flow cytometry detection. FlowJo software was used to analyze the results.

Histopathological analysis and immunohistochemistry

The xenograft tumors were isolated and fixed in 4% paraformaldehyde at room temperature for 7 days, and then embedded into paraffin. Initially, histological analysis was conducted, and the paraffin-embedded sections were stained with hematoxylin-eosin (HE). For immunohistochemistry, tissue sections were deparaffinized and hydrated with xylene, followed by the treatment with a series of ethanol solutions (100%, 95%, 80%, and 75%). The paraffin sections were then rinsed three times with PBS and subjected to antigen retrieval by heating at 90℃ for 30 min in sodium citrate buffer. Subsequently, after cooling the samples to room temperature, the sections were treated with endogenous peroxidase for 10 min at room temperature. The sections were blocked with immunohistochemical blocking solution for 1 h to prevent non-specific background staining. Then the blocking solution was removed, and the primary antibody working solution was added to the sections at 4 °C overnight. The following day, the sections were rinsed three times with TBST and then treated with Rabbit-on-Rodent HRP-Polymer/Mouse-on-Rodent HRP-Polymer at room temperature for 1 h. Subsequent to another round of washing with TBST, the sections were exposed to DAB for the specified duration (PCNA: 45 s, BAX: 3.5 min, Cleaved caspase 3: 9 min, cleaved caspase 9: 10 min).

Tissue microarrays (ZL-Lug1201) to detect SHMT2 protein expression were constructed by Shanghai Zhuoli Biotechnology Co., Ltd (Zhuoli Biotechnology Co, Shanghai, China). All the protocols were followed the same as before. The clinical characteristics of the NSCLC patients in tissue microarrays are shown in Supplementary Table S2. The brands and concentrations of the antibodies used are shown in Supplementary Table S3.

Fluorescence staining

The Mito-Tracker Deep Red FM (lot: C1032, Beyotime Biotechnology, 200 nM) was used to track mitochondria. The cells were incubated with BAX (E4U1V) rabbit mAb (lot: 41162, Cell Signaling Technology, 1:4,000) or Cyto C monoclonal antibody (lot: 66264-1-Ig, Proteintech, 1:500) with MitoTracker Deep Red FM at the same time to detect colocalization of BAX or Cyto C with mitochondria. Subsequently, the nucleus was re-stained with Hoechst (lot: 3342, Beyotime, 10 µg/mL). The cells were imaged using a confocal microscope (Nikon Eclipse Ti, Tokyo, Japan).

Propidium iodide staining

Cell death was quantified via propidium iodide (PI) (SIGMA, Cat# 537059) staining. For PI staining, cells were seeded at 20–30% confluence into 6-well plates overnight. On day 2, the cells were treated with a specified concentration of NAC or SHIN1 for 24 h. The prepared cells were incubated with PI solution (final concentration: 500 µg/mL) at 37 °C for 15 min, and then stained with Hoechst for 10 min. The PI-positive stained cells were subsequently analysed via the TRITC fluorescence channel with a fluorescence microscope.

TMRE staining

To assess the disturbance of the mitochondrial membrane potential (ΔΨm), the cells were stained with TMRE (lot: C2001S, Beyotime) and examined via confocal microscopy. ImageJ software (National Institutes of Health, Bethesda, MD) was used to measure the mean fluorescence intensity in 10 randomly selected fields to evaluate ΔΨm.

Mitochondrial extraction

To separate and extract mitochondrial proteins, mitochondrial extraction kits (C3601, Beyotime) were used according to the manufacturer’s protocols. In short, the cells were resuspended with 80 µL of mitochondria isolation reagent. After being incubated on ice for 15 min, the homogenates were centrifuged at 600 × g at 4 °C for 5 min. The supernatants were then carefully transferred to a new test tube and subjected to a second centrifugation step at 11,000 × g for 10 min to isolate the granular mitochondria. The mitochondrial pellets were collected, washed with PBS and resuspended in RIPA lysis buffer for Western blot analysis.

Glycine measurement

Cellular glycine levels were determined using a Glycine Assay Kit (Fluorometric, ab211100, Abcam) in accordance with the manufacturer’s protocol. Briefly, NSCLC cells were seeded in 10 mm culture dishes and treated with glycine for 24 h. After treatment, cells were washed with PBS and resuspended in 100 µL Assay Buffer 58 on ice, followed by homogenization via pipetting up and down and 10-minute incubation on ice. The cell homogenate was centrifuged at 10,000 × g for 5 min at 4 °C, and a 25 µL aliquot of the supernatant (adjusted to 50 µL with Assay Buffer 58) was mixed with 50 µL of prepared Reaction Mix in a 96-well black plate. The plate was incubated at 25 °C for 60 min protected from light, and fluorescence was detected with a microplate reader (ThermoFisher, Varioskan LUX, USA) at Ex/Em = 535/587 nm. Glycine concentrations were calculated using a standard curve generated with the kit’s glycine standard.

Subcutaneous xenograft model

BALB/c nude mice (5–6 weeks old; female) were purchased and randomly divided into three groups (n = 8 per group). A549 Sg-NC/Sg-SHMT2-1/Sg-SHMT2-2 cells were subcutaneously injected into the armpits of the mice. After 30 days, the xenograft tumors were isolated, weighed, and recorded. The tumor volume was calculated using the following formula: volume = length × width² ×  0.5. For the in vivo NAC rescue experiments, starting on day 15, the mice were intraperitoneally injected with NAC (S1623, Selleckchem) at a dose of 400 mg/kg body weight, dissolved in PBS (pH 7.4), every other day for two weeks. After 34 days, the mice were euthanized, and the tumors were dissected, weighed, and recorded. All the xenograft tumors were collected for hematoxylin and eosin (H&E) staining and immunohistochemistry assays later. The above animal experimental procedures obtained official approval from the Laboratory Animal Center of Soochow University (approval no. 202308A0815 and 202405A0636).

Statistical analysis

All statistical analyses were performed using GraphPad Prism 8.0 (San Diego, CA, USA). Bioinformatics analysis was performed with R software (version 4.4.1). Data are presented as mean ± standard deviation (SD). T-tests (two-sided) were employed for inter-group comparisons. For comparisons among multiple groups, one-way ANOVA test was utilized, with statistical significance set at P < 0.05. All experiments were conducted in triplicate.

Results

SHMT2 is highly expressed in human NSCLC tissue and is associated with poor prognosis in patients with LUAD

First, we analyzed the expression of SHMT2 in NSCLC tissue by using The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus Series (GSE81809) databases. We observed significantly elevated expression of SHMT2 in LUAD, LUSC and NSCLC tissues compared with normal lung tissues, respectively (Fig. 1A&B). To validate the bioinformatic findings, we performed qRT-PCR analysis to assess SHMT2 mRNA expression in tumor tissue and paired paracancerous tissue obtained from NSCLC patients (N = 60). Consistently, SHMT2 mRNA expression was markedly increased in tumor tissue compared with that in paired paracancerous tissues (Fig. 1C). Moreover, the results of the Western blot analysis also revealed the increased protein expression of SHMT2 in the tumor tissue of randomly selected NSCLC patients in comparison with the paired paracancerous tissue (N = 14) (Fig. 1D). Furthermore, immunohistochemical staining results from 60 patients further indicated the increased expression of SHMT2 protein in NSCLC tissue compared to the paired normal lung epithelial tissue (Fig. 1E).

Fig. 1.

SHMT2 is highly expressed in human NSCLC tissues and is associated with poor prognosis of patients with LUAD. (A-B) The expression levels of SHMT2 in NSCLC, LUAD, or LUSC tissues and normal lung tissue were assessed utilizing the TCGA and GSE81809 databases. (C) The mRNA expression of SHMT2 in tumor tissue (“T”) and paired non-cancerous tissue (“N”) of 60 patients with NSCLC was detected by qRT-PCR method. (D) The protein expressions and quantification of SHMT2 in 14 pairs of tumor tissue and paired non-cancerous tissue in NSCLC patients. (E) Immunohistochemical staining images depicting levels of SHMT2 in tumor tissue and matched normal tissue from 60 typical NSCLC patients. (F-G) Overall survival of the patients with high or low expression of SHMT2 in NSCLC, LUAD or LUSC from TCGA (F) and GSE81089 (G) databases. (H-I) Univariate and multivariate Cox analyses involving pathologic TNM stages, age, gender, smoker and SHMT2 expression for LUAD patients. Quantitative data were presented as mean ± SD. *P < 0.05, **P < 0.01

Moreover, the results of survival analysis revealed that patients with high SHMT2 expression from TCGA and GSE81809 database had shorter survival durations than those with low SHMT2 expression in LUAD, but not in LUSC or NSCLC (Fig. 1F&G). Besides, the expression of SHMT2 was related to the histological stage of the LUAD patients (Fig. S1). Next, to validate the correlation between SHMT2 expression and LUAD prognosis, we conducted univariate and multivariate Cox regression analyses involving pathologic TNM stages, age, gender, smoker and SHMT2 expression. Univariate Cox regression analysis demonstrated that clinical stage, TNM stage, and SHMT2 expression levels were significantly associated with lung cancer risk (Fig. 1H). Moreover, multivariate analysis further revealed that SHMT2 could act as an independent prognostic risk factor for NSCLC patients (HR = 1.391, 95% CI: 1.015–1.906, P = 0.04) (Fig. 1I).

Knockout of SHMT2 inhibits the growth of human LUAD cells in vitro and in vivo

To investigate the potential effects of SHMT2 on LUAD cells, we first compared the protein expression of SHMT2 in a human normal lung epithelial cell line (BEAS-2B) and LUAD cell lines (A549, H1299, H838, H358 and H1650). The results revealed that SHMT2 was highly expressed in LUAD cells compared with BEAS-2B cells, especially in A549 and H1299 cells (Fig. 2A). Furthermore, the CRISPR-Cas9 genome editing system was employed to establish SHMT2 knockout stable cells of A549 and H1299 cells with SHMT2 knockout sequences (Sg-SHMT2) or control negative sequences (Sg-NC). The results suggested that the targeting sequences only knocked out SHMT2 protein expression in stable knockout cell lines, and had no effect on the protein expression of the cytoplasmic isoform SHMT1 (Fig. 2B).

Fig. 2.

Knockout of SHMT2 inhibits growth of human LUAD cells in vitro and in vivo. (A) Western blotting analysis showed SHMT2 expression in a panel of human NSCLC cells and immortalized normal bronchial epithelial cells (BEAS-2B). (B) Western blotting analysis showed SHMT2 and SHMT1 protein expressions in control and SHMT2-knockout A549 and H1299 cells. (C-F) A549 and H1299 cells were cultured for the indicated times and then assessed by CCK-8 assay (C), colony formation (D) and EdU staining assay. The Edu staining results were quantified and analyzed. Scale bar = 100 μm (E-F). (G-H) The expression of PCNA protein was detected by Western blotting assay in SHMT2 knockout A549 and H1299 cells. GAPDH was used as an loading control. (I-K) The removed xenograft tumors were photographed on day 30 (I), and the xenograft tumors’ volume (J) and weight (K) were calculated and statistically analyzed. (L-M) H&E staining and immunohistochemical staining was performed on the xenograft tumors of the three groups of nude mice to detect the expression of PCNA in the tumor tissues, with brown representing positive staining, scale bar = 200 μm, n = 8 per group. (N) The expression of PCNA protein in xenografts was detected by Western blotting, n = 4 per group. Quantitative data are presented as mean ± SD. ***P < 0.001

Then we elucidated the effects of SHMT2 knockout on A549 and H1299 cell growth with stable cell lines and the inhibitor SHIN1. As a specific inhibitor of SHMT1/2, SHIN1 exerts a highly specific inhibitory effect on LUAD cells, and this effect mainly depends on the overexpression of SHMT2 [35]. The results of the CCK-8, colony formation and EdU staining assays demonstrated that CRISPR/Cas9-mediated SHMT2 knockout as well as SHIN1 treatment significantly reduced A549 and H1299 cell viability (Fig. 2C, S2A) as well as cell proliferation (Fig. 2D, S2B, Fig. 2E&F, S2C), respectively. Furthermore, the results of Western blot assay showed that the expression of proliferation-related protein PCNA was markedly decreased in SHMT2 knockout as well as SHIN1-treated A549 and H1299 cells (Fig. 2G&H, S2D). To explore the effects of SHMT2 on the growth of LUAD cells in vivo, we injected SHMT2 knockout (Sg-SHMT2) or blank control (Sg-NC) A549 stable cells into the right axilla of nude mice to establish subcutaneous xenograft tumor models. The results showed that SHMT2 knockout markedly reduced the growth of xenograft tumor in vivo (Fig. 2I-K), and that the protein expression of PCNA was downregulated (Fig. 2L-N). In summary, SHMT2 has positive effects on the growth of LUAD cells and may contribute to its carcinogenic effects in vitro and in vivo.

Inhibition of SHMT2 promoted ROS production and disrupted mitochondrial homeostasis in LUAD cells

To elucidate the potential role of SHMT2 in regulating the growth of LUAD cells, we performed GO enrichment analysis on LUAD cells after SHMT2 knockout, and the results suggested that SHMT2 might be related to oxidative stress (Fig. 3A). Oxidative stress is characterized by the generation and accmulation of ROS. As SHMT2 is a catalytic enzyme within the mitochondria, its function is intricately linked with ROS generation. Therefore, we further examined whether SHMT2 knockout affects ROS levels in LUAD cells. The results of the subsequent analysis revealed a notable increase in the levels of intracellular and mitochondrial ROS in SHMT2-deficient cells compared to the controls (Fig. 3B&C). Meanwhile, SHIN1 treated cells showed higher ROS production in A549 and H1299 cells (Fig. S3A&B). Subsequently, compared to control cells, SHMT2 knockout significantly reduced the NADPH/NADP+ ratio and the GSH/GSSG ratio, suggesting that SHMT2 deficiency impairs the cellular antioxidant buffering capacity (Fig. 3D). Moreover, the TEM results showed that knockout of SHMT2 induced significant mitochondrial damage, which was characterized by morphological contraction swelling, crista loss, and fragmentation (Fig. 3E). As Fig. 3F & Fig. S3C showed that most mitochondria showed red fluorescence in the control group, while green was predominant in fluorescence in SHMT2 knockout cells and SHIN1-treated cells. This shift from red to green fluorescence indicates that targeted inhibition of SHMT2 caused a decrease in mitochondrial membrane potential. Furthermore, MitoTracker and PK mito deep red staining results verified the changes of mitochondrial morphology and distribution in SHMT2 knockout cells and SHIN1-treated cells (Fig. S3D-F). Finally, TMRE staining results confirmed the decrease in the mitochondrial membrane potential (Fig. S3G). These results indicate that SHMT2 knockout can affect oxidative stress and mitochondrial homeostasis in LUAD cells.

Fig. 3.

Inhibition of SHMT2 disrupted mitochondrial homeostasis by promoting ROS production in LUAD cells. (A) The results of gene ontology (GO) analysis of differential expressed genes (DEGs) in biological process (BP) between normal and SHMT2- knockout cells. (B) Detection of ROS production using immunofluorescence in SHMT2 knockout LUAD cells. Green: ROS, Blue: Hoechst, Scale bar = 100 μm. (C) MitoSOX™ Red was used to detect mtROS using immunofluorescence in SHMT2 knockout LUAD cells. Red: mtROS, Blue: Hoechst, Scale bar = 50 μm. (D) The effect of SHMT2 deficiency on NADPH/NADP+ ratio and GSH/GSSG ratio. (E) Representative TEM images of mitochondrial morphology in SHMT2 knockout LUAD cells. Scale bar = 20 μm. (F) JC-1 staining results showed mitochondrial membrane potential of SHMT2 knockout LUAD cells, scale bar = 50 μm. The relative fluorescence intensity was quantified using ImageJ. n = 3 per group. Quantitative data were presented as mean ± SD. **P < 0.01, ***P < 0.001

Inhibition of LUAD cell growth induced by SHMT2 deficiency is mediated primarily through the accumulation of ROS

These results indicate that the function of SHMT2 is be closely related to ROS production. To further explore whether the inhibitory effect of SHMT2 knockout on cell growth was associated with the up-regulation of ROS, we applied ROS scavenger N-acetylcysteine (NAC) to treat LUAD cells. The results of CCK-8 and colony formation assays revealed that NAC treatment reversed the inhibition of cell growth induced by SHMT2 deficiency (Fig. 4A-C). We also found that NAC had a stable recovery effect at 2.5 mM in A549/H1299 SHMT2 knockout and SHIN1-treated cells, thus 2.5 mM NAC was used in the subsequent experiments. Next, to determine the in vivo functional contribution of SHMT2 to tumor progression, we established a LUAD xenograft mouse model and treated it with NAC. Knockout of SHMT2 significantly reduced tumor growth, and NAC treatment (intraperitoneal (i.p.)) injection, 400 mg/kg/day) significantly reversed the effects of SHMT2 knockout (Fig. 4D-G). Then intracellular and mitochondrial ROS levels were determined via the fluorescence analyses, and the results revealed that NAC inhibited ROS production in SHMT2 knockout and SHIN1-treated cells (Fig. S4A-D). Moreover, NAC treatment effectively rescued the SHMT2 deficiency-induced mitochondrial dysfunction, as evidenced by the restoration of ΔΨm (Fig. 4H-I) and the normalization of mitochondrial morphology and spatial distribution (Fig. S4E-F). In summary, genetic ablation of SHMT2 triggered ROS accumulation and the subsequent oxidative stress, ultimately suppressed LUAD cell proliferation via disruption of mitochondrial homeostasis.

Fig. 4.

Inhibition of LUAD cell growth induced by SHMT2 deficiency is primarily mediated through the accumulation of ROS. (A) Cell viability was assessed by CCK-8 assay to detect the effect of SHMT2 deficiency or SHIN1 treatment on growth of LUAD cells, and the effects of NAC (0, 1, 2.5, 5 or 10 mM, 24 h). (B-C) Colony formation assays were performed to detect the effect of SHMT2 deficiency or SHIN1 treatment on growth of LUAD cells, and the effects of NAC (0, 1, 2.5 or 5 mM, 24 h). (D-G) Sg-SHMT2 or Sg-NC A549 cells (2.5 × 10⁶ cells/per mouse) were injected into nude mice. Mice were treated with or without NAC, individually (D). Tumor images (E), tumor volumes (F) and tumor weight (G) are shown. The data are represented as the averages, n = 7 per group. (H-I) TMRE staining was used to assess the mitochondrial membrane potential in SHMT2-knockout cells or SHIN1-treated cells, as well as the recovery effects of NAC (2.5 mM, 24 h). The mitochondrial membrane potential was quantified by TMRE fluorescence intensity, scale bar = 100 μm. Quantitative data are presented as mean ± SD. ns: not significant, *P < 0.05, **P < 0.01, ***P < 0.001

SHMT2 deficiency induces the apoptosis of LUAD cells

To explore the potential mechanism by which SHMT2 deficiency suppresses LUAD cell growth, different types of inhibitors were used to detect the cell death caused by SHMT2 deficiency at the indicated different concentrations. The results of the CCK-8 assays revealed that Z-VAD, an apoptosis inhibitor, reduced SHMT2 knockout-induced cell death in A549 stable cells at a concentration of 30 µM, whereas ferrostatin-1 (FER-1, a ferroptosis inhibitor), necrostatin (NEC, a necroptosis inhibitor), and tetrathiomolybdate (TTM, a copper death inhibitor), had no obvious effect on the cell death of SHMT2-knockout cells (Fig. 5A). These results suggested that apoptosis was the major form of SHMT2-knockout induced cell death. Furthermore, the ability of 30 µM Z-VAD to improve the viability of SHMT2 knockout A549 and H1299 cells, was repeatedly verified (Fig. 5B). Moreover, PI staining demonstrated that the proportion of apoptotic cells was significantly increased in SHMT-knockout cells (Fig. 5C) and SHIN1-treated cells (Fig. 5D). The results of flow cytometry assay verified this conclusion (Fig. 5E&F). In summary, genetic depletion of SHMT2 significantly promotes apoptotic cell death in LUAD cells, thereby substantially compromising their viability.

Fig. 5.

SHMT2 deficiency induces apoptosis of LUAD cells. (A) Viability of A549 Sg-NC and Sg-SHMT2 cells cultured with or without Fer-1, NEC, TTM and Z- VAD after 24 h. (B) Viability of A549 and H1299 Sg-NC and Sg-SHMT2 cells cultured with or without Z-VAD (30 µM) after 24 h. (C-D) Detection of apoptosis in SHMT2 knockout and SHIN1-treated cells via PI staining. (E-F) Detection of apoptosis in SHMT2-knockout and SHIN1-treated cells via flow cytometry methods. Scale bar = 200 μm. Quantitative data are presented as mean ± SD. ns: not significant, *P < 0.05, **P < 0.01, ***P < 0.001

SHMT2 deficiency triggers the mitochondrial-mediated apoptotic signaling pathway

The above results indicated that SHMT2 knockout can induce serious damage to the mitochondrial function and promote cell apoptosis. Therefore, we speculate whether SHMT2 knockout promotes cell apoptosis via a mitochondrial-related pathway in LUAD cells. The mitochondrial apoptosis pathway is known to involve multiple signaling pathways. The conduction of apoptotic signals recruits BAX to the outer membrane of the mitochondria, forming the mitochondrial permeability transition pore (mPTP), which mediates the release of Cyto C and mitochondrial DNA (mtDNA) from the mitochondria. mtDNA released into the cytoplasm activates the cGAS-STING pathway or the NF-κB pathway to induce caspase-independent cell death, while Cyto C induces caspase-dependent mitochondrial apoptosis (Fig. 6A).

Fig. 6.

SHMT2 deficiency triggers mitochondrial-mediated apoptotic signaling pathway. (A) Schematic diagram of multiple pathways of mitochondrial apoptosis. (B) Western blot was used to determine the expression of BAX apoptosis pathway-related proteins in SHMT2 knockout and SHIN1-treated cells. (C) The protein expression of Bax and Cyto C in mitochondria or cytoplasm of SHMT2 knockout cells. (D-E) The co-localization of BAX and mitochondria in SHMT2 knockout and SHIN1-treated cells. Red: Mito-tracker; Green: BAX; Blue: Hoechst; scale bar = 20 μm. The quantitative statistics are shown in the right image. (F-G) The co-localization of Cyto C and mitochondria in SHMT2-knockout and SHIN1-treated cells. Red: Mito-tracker; Green: Cyto C; Blue: Hoechst; scale bar = 20 μm. The quantitative statistics are shown in the right image

In SHMT2-knockout cells and SHIN1-treated cells, the levels of intrinsic apoptotic pathway proteins i.e. BAX, cleaved caspase 3, and cleaved caspase 9, as determined by Western blotting, showed a significant increase, while the expression of the anti-apoptotic protein bcl-2 decreased (Fig. 6B). Furthermore, the levels of cGAS-STING pathway or NF-κB pathway-related proteins induced by mtDNA release remain unchanged (Fig. S5A&B). Moreover, the increased expression of apoptosis related proteins can also be observed in mice xenografts of SHMT2 knockout cells by IHC (Fig. S5C&D), suggesting that SHMT2 knockout may promote the activation of BAX-dependent mitochondrial apoptosis. Next, total mitochondrial proteins and cytoplasmic proteins were extracted separately from SHMT2 knockout cells, and Western blot results showed that SHMT2 knockout promoted the mitochondrial translocation of Bax and the release of Cyto C (Fig. 6C). The co-localization of BAX or Cyto C with mitochondria was further detected by Immunofluorescence assay. The results revealed that the colocalization of BAX with mitochondria was increased (Fig. 6D&E), while the co-localization of Cyto C with mitochondria was decreased (Fig. 6F&G) in SHMT2 knockout cells and SHIN1-treated cells.

These results suggested that blocking SHMT2 could increase the recruitment of BAX to mitochondria, reducing the ΔΨm, which in turn increases the permeability of the membrane. This process results in the release of Cytochrome C from mitochondria, triggering the activation of the caspase cascade and initiating intrinsic mitochondrial apoptosis.

The mitochondrial-dependent apoptosis induced by SHMT2 deficiency is mediated through ROS accumulation

Finally, to ascertain whether the mitochondrial apoptosis triggered by SHMT2 deletion stems from increased intracellular ROS, we investigated the levels of mitochondrial apoptosis following the treatment of SHMT2-knockout cells with the mitochondrial ROS scavenger NAC for 24 h. The results of flow cytometry (Fig. 7A&B, Fig. S6A), PI staining (Fig. 7C, Fig. S6B) and TUNEL assays (Fig. 7D) showed that NAC treatment significantly reduced the apoptosis in both SHMT2-knockout, and SHIN-treated cells. We subsequently investigated the expression levels of PCNA and mitochondrial apoptosis-related proteins via IHC in LUAD xenograft mice models with NAC treatment. Furthermore, PCNA was increased and the cleavage products of caspase 9, caspase 3 and BAX in SHMT2 knockout mouse xenografts showed significant reduction after NAC treatment (Fig. 7E). These findings provide compelling evidence that genetic ablation of SHMT2 induces mitochondria-mediated apoptosis via ROS accumulation both in vitro and in vivo.

Fig. 7.

Mitochondrial-dependent apoptosis induced by SHMT2 deficiency is mediated through ROS accumulation. (A-D) Flow cytometry (A & B), PI staining (C), and TUNEL staining (D) were used to analyze the apoptosis ratio of SHMT2-deficient cells, as well as the effects of NAC recovery (2.5 mM, 24 h), and all results were quantified, scale bar = 200 μm. (E) SHMT2 knockout and control tumors with or without NAC treatment were stained to show representative IHC staining images of PCNA, BAX, Cleaved-caspase 3, Cleaved-caspase 9, as well as representative images of H & E staining. Scale bar = 200 μm, Average Optical Density (AOD) displayed on the right side. Quantitative data are presented as mean ± SD. ns: not significant, *P < 0.05, **P < 0.01, ***P < 0.001

SHMT2 deficiency-induced growth Inhibition and mitochondrial-dependent apoptosis are mediated by loss of SHMT2 enzymatic activity

As the key mediator in one-carbon metabolism, SHMT2 facilitates the transfer of a carbon moiety from tetrahydrofolate to 5,10-methylene tetrahydrofolate, while also mediating the conversion of serine to glycine [37]. Previous studies have shown that SHMT2 depletion significantly downregulates the level of glycine, which is the direct product of the reaction catalyzed by SHMT2 [38, 39]. To check how SHMT2 deficiency affects intracellular glycine levels and whether adding external glycine helps, we first measured glycine content in Sg-SHMT2 stable cells. Results showed that Sg-SHMT2 stably cells had much lower intracellular glycine levels than control cells. However, when we added external glycine to these SHMT2-deficient cells, their intracellular glycine levels were upregulated (Fig. 8A). These findings confirm that SHMT2 deficiency reduces the cell’s ability to make glycine, and adding external glycine can reverse this reduction. Based on these findings, we further examined whether glycine supplementation has an impact on ROS accumulation. Indeed, glycine supplementation reversed the ROS accumulation induced by SHMT depletion (Fig. 8B). Similarly, in line with these results, glycine supplementation can restore the proliferative capacity of LUAD cells upon SHMT2 deficiency (Fig. 8C&D). Meanwhile, glycine supplementation also reversed the elevated apoptotic levels induced by SHMT2 depletion in LUAD cells (Fig. 8E). In summary, we found that SHMT2 deficiency suppresses the proliferation of LUAD cells through the loss of its enzymatic function.

Fig. 8.

SHMT2 deficiency-induced growth inhibition and mitochondrial-dependent apoptosis in LUAD cells are mediated by the loss of SHMT2 enzymatic activity. (A) A Glycine Assay Kit was used to detect cellular glycine levels by immunofluorescence in SHMT2-knockout cells, as well as the recovery effect of glycine. (B) CM-H2DCFDA was used to detect ROS production by immunofluorescence in SHMT2-knockout cells, as well as the recovery effect of glycine (0.1 mM, 24 h). Scale bar = 50 μm. (C) To detect the rescue effect of glycine (0.1 mM, 24 h) treatment on the proliferation of SHMT2-knockout lung cancer cells via the CCK8 assay. (D) To detect the effect of glycine (0.1 mM, 24 h) treatment on PCNA protein expression in SHMT2-knockout cells. (E) Flow cytometry was used to analyze the apoptosis ratio of SHMT2-deficient cells, as well as the effects of glycine recovery (0.1 mM, 24 h), and all results were quantified. Quantitative data are presented as mean ± SD. ns: not significant, *P < 0.05, **P < 0.01, ***P < 0.001

This study contributes to our understanding of SHMT2’s roles in LUAD by demonstrating its involvement in regulating tumor cell apoptosis and prognosis. Knocking out of SHMT2 significantly increases ROS accumulation, leading to mitochondrial-dependent apoptosis. Mechanistically, SHMT2 inhibition increases the expression of the proapoptotic protein BAX and triggers Cyto C release from mitochondria to the cytoplasm in an ROS-dependent manner. These changes can activate the caspase cascade, initiating the ROS-dependent mitochondrial apoptosis pathway, to suppress the progression of lung adenocarcinoma (Fig. 9).

Fig. 9.

Graphic illustration of eliminating SHMT2-mediated mitochondrial apoptosis suppresses the progression of lung adenocarcinoma dependent on ROS production

Discussion

SHMT2 exerts oncogenic metabolic effects in LUAD, and it has been demonstrated to promote the growth of various tumors [40, 41]. The roles of SHMT2 in LUAD have also been reported; however, the precise molecular mechanisms underlying SHMT2 inactivation-mediated regulation of LUAD cell survival and proliferation remain to be fully elucidated. In this study, we provide evidence revealing the critical roles of SHMT2 in LUAD cell growth and elucidate the molecular mechanisms underlying the ROS-mediated mitochondrial apoptosis triggered by SHMT2 inhibition through the intrinsic apoptotic pathway. From a mechanistic perspective, the inhibition of SHMT2 resulted in the accumulation of ROS, which in turn led to oxidative stress and the impairment of mitochondrial homeostasis. Apoptotic signals can promote the mitochondrial translocation of the downstream molecule BAX, reduce the ΔΨm and increase membrane permeability. The above effects may lead to the release of Cyto C, and induce the activation of the caspase cascade, finally leading to intrinsic mitochondrial apoptosis. Notably, inhibition of ROS generation by the specific scavenger NAC significantly attenuated SHMT2 deficiency-induced apoptotic cell death through the intrinsic mitochondrial pathway. Furthermore, we demonstrated that SHIN1, a SHMT2 inhibitor, exerts the proapoptotic effects observed in SHMT2-deficient LUAD cells, triggering mitochondrial-mediated apoptotic cell death. In summary, the knockout or inhibition of SHMT2 has been shown to inhibit LUAD cell growth by promoting mitochondrial apoptosis, suggesting that SHMT2 may serve as a novel therapeutic target for LUAD.

Mitochondria are the central hub of mammalian cells, and the homeostasis as well as function of mitochondria play crucial roles in maintaining cellular homeostasis and the survival of cells. Mitochondria serve as central executioners of programmed cell death, mediating apoptosis through the release of proapoptotic factors from the intermembrane space and generating cytotoxic reactive oxygen species [42]. In this study, we constructed SHMT2 knockout cell lines, and subsequently employed probe staining to detect levels of ROS in LUAD cells and mitochondria. Furthermore, MitoTracker, JC-1 and TMRE staining assays were employed to confirm alterations in mitochondrial structure and membrane potential. Our data suggested that ROS levels were increased in SHMT2 knockout or SHIN1-treated LUAD cells. Consistent with the results of previous findings [43], oxidative stress caused by elevated levels of ROS may further lead to decreased mitochondrial membrane potential and also morphological changes. On the other hand, NAC treatment, scavenging excessive ROS, can rescue the decreased cell growth ability induced by SHMT2 knockout. These results indicated that the loss of SHMT2, a pivotal mitochondrial metabolism enzyme, leads to mitochondrial homeostasis disorders, consequently impairing growth of LUAD cells.

Apoptosis, a genetically programmed cell death mechanism, plays a crucial role in maintaining tissue homeostasis and internal environmental stability through two distinct molecular pathways: the extrinsic (death receptor-mediated) pathway and the intrinsic (mitochondria-dependent) pathway. Previous studies have established that the intrinsic apoptotic pathway, also known as the mitochondria-dependent pathway, represents one of the fundamental signaling cascades that regulate cell apoptosis. ROS act as proapoptotic factors [44, 45]. The accumulation of ROS also reduces ΔΨm, which is a precursor event to mitochondrial apoptosis [46]. This process facilitates MOMP, thereby enabling the translocation of Cyto C from the mitochondrial intermembrane space to the cytosol. Subsequently, Cyto C forms an apoptosome with caspase-9 and APAF-1, leading to the activation of the caspase-9 signaling pathway and the initiation of apoptosis [47]. In this study, the levels of mitochondrial apoptosis-related proteins, i.e. BAX, cleaved-caspase 9 and cleaved-caspase 3, were assessed via immunofluorescence, immunohistochemistry and Western blot assays. Our experimental data demonstrated a significant elevation in mitochondrial-mediated apoptosis following SHMT2 knockout. Nevertheless, the precise molecular mechanisms underlying the role of ROS-mediated oxidative stress in the recruitment and activation of BAX at the mitochondrial outer membrane remain to be fully elucidated.

SHMT2 is highly expressed in lung cancer [48]; however, the effects of SHMT2, a metabolic regulator, on mitochondrial function have not yet been reported. While SHMT2’s oncogenic role in LUAD has been reported, our study extends these findings by systematically demonstrating that targeting SHMT2 disrupts mitochondrial homeostasis through promoting mitochondrial ROS production, which in turn drives mitochondrial apoptosis in LUAD cells. Given that SHMT2 depletion inhibits LUAD cell growth and induces apoptosis in our preclinical models, the development of stable, clinically applicable SHMT2-targeted inhibitors could represent a potential therapeutic direction for LUAD, which merits further preclinical exploration. Inhibitors of SHMT2 have substantial potential for the treatment of lung adenocarcinoma [35], bladder cancer [24], pancreatic cancer [37] and other diseases. SHIN1, a SHMT2 inhibitor, has considerable anti-tumour effects on colon cancer cell line HCT116. SHIN1 treatment significantly downregulated the expression of SQSTM1, the autophagy adaptor protein in HCT116 cells, leading to disruption of mitochondrial homeostasis and consequent inhibition of cellular growth [37]. In this study, SHIN1 treatment was observed to significantly reduce the viability of LUAD cells. However, inhibitors targeting SHMT2 only showed anti-tumor effects at the cellular level and thus cannot be used in animal studies due to the poor stability. Currently, inhibitors that target SHMT2 have not yet been approved for clinical application [49]. In summary, our findings demonstrate the therapeutic potential of SHMT2-targeted pharmacological interventions in LUAD treatment, providing a strong preclinical foundation for future clinical translation and drug development.

Conclusions

Our findings suggest that SHMT2 is vital in regulating LUAD apoptosis and that the knockout of SHMT2 enhances the sensitivity of tumor cells to apoptosis through the ROS-BAX-Cyto C-Caspase signaling. Targeting SHMT2 may offer a potential approach for LUAD treatment, providing new preclinical insights into metabolic targeted therapy. Future investigations should focus on the development of novel SHMT2 inhibitors and systematically evaluate their potential synergistic effects with existing therapeutic modalities, particularly in combination with targeted therapies and immunotherapies.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

NSCLC

Non-small cell lung cancer

SCLC

Small cell lung cancer

LUAD

Lung adenocarcinoma

LUSC

Lung squamous cell carcinoma

LLC

Large cell carcinoma

SHMT2

Serine hydroxymethyltransferase

ROS

Reactive oxygen species

NAC

N-Acetyl cysteine

PCNA

Proliferating cell nuclear antigen

Z-VAD

z-VAD-Ala-Asp-fluoromethyl ketone

FER-1

Ferrostatin-1

NEC

Necrostatin

TTM

Tetrathiomolybdate

Cyto C

Cytochrome C

ΔΨm

Mitochondrial membrane potential

MOMP

Mitochondrial outer membrane permeabilization

QRT-PCR

Quantitative real-time polymerase chain reaction

IHC

Immunohistochemistry

H&E

Hematoxylin and eosin

IF

Immunofluorescence

TUNEL

Terminal deoxynucleotidyl transferase dUTP nick end labeling

TCGA

The Cancer Genome Atlas

GSE

Gene Expression Omnibus Series

PBS

Phosphate Buffer Saline

PI

Propidine iodide

MPT

Mitochondrial permeability transition

AOD

Average optical density

MPTP

Mitochondrial permeability transition pore

NADPH

Reduced nicotinamide adenine dinucleotide phosphate

NADP⁺

Oxidized nicotinamide adenine dinucleotide phosphate

GSH

Glutathione

GSSG

Glutathione disulfide

Author contributions

Yili Chen: Data curation, Methodology, Visualization, Writing-original draft, Writing-review & editing. Xinyu Zhang: Formal analysis, Investigation, Methodology, Writing-original draft. Xiaodong Pang: Data curation, Investigation, Formal analysis. Yang Yang: Data curation, Methodology, Visualization. Di Lu: Data curation, Investigation. Jian Zhao: Data curation, Formal analysis. Jianjie Zhu: Data curation, Formal analysis, Funding acquisition. Wenwen Du: Data curation, Formal analysis. Chen Gu: Data curation, Formal analysis. Jianjun Li: Data curation, Formal analysis. Lei Gu: Data curation, Formal analysis. Jian-an Huang: Formal analysis, Visualization, Funding acquisition. Zeyi Liu: Conceptualization, Formal analysis, Visualization, Writing-review & editing. Yuanyuan Zeng: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Visualization, Writing-original draft, Writing-review & editing.

Funding

This work was supported by grants from the National Natural Science Foundation of China (No. 82272648, No. 82202886), the Science and Technology Plan Project of Suzhou (SKY2022133, SKY2023160), the Suzhou Gusu Medical Talent (GSWS2021001) and Jiangsu Provincial Medical Key Discipline (ZDXK202201).

Data availability

Data support the results of this study will be provided by the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

All the clinical studies were conducted with relevant ethical regulations, including the Declaration of Helsinki and were approved by the Ethics Committee of the First Affiliated Hospital of Soochow University (Ethics code 2024662), and the informed consent was acquired from each patient. All animal experimental procedures were obtained official approval from the Laboratory Animal Center of Soochow University (approval no. 202308A0815 and 202405A0636).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.