Transcriptome data from CRC and normal tissues were acquired from The Cancer Genome Atlas (TCGA) database via the GEPIA platform (http://gepia.cancer-pku.cn/). Differential mRNA expression analysis of the ASS1 gene was performed to characterize its expression profile in CRC.

Tumor and matched para-carcinoma samples were collected from patients with colorectal cancer patients undergoing surgical resection at the Central Hospital of Zibo. All samples were promptly snap-frozen in liquid nitrogen and maintained at –80 °C until analysis. The study protocol was approved by the Ethics Committee of the Central Hospital of Zibo (Approval No. IRB-ZBCH-2024-078A). Every participant provided written informed consent in accordance with the Declaration of Helsinki. For analysis, tissues were cut into 4 µm slices after being paraffin-embedded. For IHC, the following primary antibodies were used: anti-ASS1 (1:50, Cat. No. 16210-1-AP, Proteintech, Chicago, IL, USA), anti-CPT1A (1:500, Cat. No. 15184-1-AP, Proteintech, Chicago, IL, USA), and anti-S100A10 (1:100, Cat. No. 11250-1-AP, Proteintech, Chicago, IL, USA). Representative fields were captured using an Olympus BX-51 TF microscope (Olympus Corporation, Tokyo, Japan). IHC scoring was conducted according to the proportion of positive cells and staining intensity. Protein expression was assessed based on staining intensity (scored 0–3: negative, weak, moderate, strong) and the proportion of positive cells (scored 0–4: 0%, 1–25%, 26–50%, 51–75%, 76–100%). Multiplying the staining intensity by the proportion score of positive cells yielded the final IHC score, which ranged from 0 to 12. Two seasoned pathologists conducted the assessments separately and blindly with regard to clinical data.

CRC cell lines SW480 (Cat. No. CBP60019), HCT116 (Cat. No. CBP60028), HT29 (Cat. No. CBP60011), and CT26 (Cat. No. CBP30258L) were all bought from Nanjing Kebai Biotechnology Co., Ltd (Nanjing, Jiangsu, China). The NCM460 cell line, derived from normal human colonic epithelium (Catalog Number CD0408), was procured from Shanghai QiDa Biotechnology Co., Ltd (Shanghai, China). All cell lines were cultured at 37 °C in a humidified incubator. Specifically, SW480 cells were maintained in Leibovitz’s L-15 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin, and incubated without CO₂ supplementation. The other cell lines were cultured in a 5% CO₂ atmosphere: HCT116 and HT29 cell lines were propagated in McCoy’s 5A medium supplemented with 10% FBS and 1% penicillin-streptomycin. For the cultivation of CT26 and NCM460 cells, RPMI-1640 medium was employed, also enriched with 10% FBS and 1% penicillin-streptomycin. All reagents and materials for cell culture were sourced from Procell System (Wuhan, Hubei, China). Authentication of each cell line was performed using short tandem repeat (STR) profiling, and all tested negative for mycoplasma contamination.

Short hairpin RNAs (shRNAs) that target human ASS1 or a non-targeting control shRNA were cloned into the pLKO.1 vector (Cat. No. 8453, Addgene, Watertown, MA, USA) in order to reduce the expression of ASS1. The following were the ASS1 targeting sequences: shASS1-#1: 5^′-GAAAACAGATTCCAGACGCCG-3^′; shASS1-#2: 5^′-TTGATTTTGCGCACTTCCCG-3^′. For overexpression, full-length coding sequences of human ASS1, CPT1A, and S100A10 were individually ligated into the pLVX vector (Cat. No. 631253, Takara Bio, Kusatsu, Shiga, Japan). Lentiviruses were packaged and produced by GenePharma (Shanghai, China). To create stable cell lines, CRC cells were infected with the relevant lentiviruses and then selected using puromycin (Cat. No. P8833, Sigma-Aldrich, St. Louis, MO, USA) at a concentration of 2 µg/mL. Prior to conducting functional experiments, the efficiency of gene knockdown and overexpression was confirmed through quantitative reverse transcription polymerase chain reaction (qRT-PCR) and Western blot analysis.

After seeding in 96-well plates, CRC cells were cultured for specified durations. The plates were incubated at 37 °C for three hours after 10 µL of CCK-8 solution (Cat. No. C0037, Beyotime, Shanghai, China) was added to each well. A microplate reader was used to measure absorbance at 490 nm in order to determine viability. To evaluate colony-forming ability, a clonogenic assay was performed. Cells were seeded into 6-well plates, cultured for one week, then fixed and stained with crystal violet (Cat. No. C0775, Sigma-Aldrich, St. Louis, MO, USA). The resulting colonies were manually counted for analysis.

Cell migration and invasion capabilities of CRC cells were evaluated using Transwell chambers (8 µm pore size; Cat. No. PCEP06H48, MilliporeSigma, Burlington, MA, USA). For the invasion assay, the interior of the inserts was pre-coated with a 0.2% Matrigel layer (Cat. No. 356234, Corning, Corning, NY, USA). Subsequently, 2 $\times$ 10⁵ cells were plated in the upper compartment, while the lower chamber was filled with 600 µL of medium containing 10% fetal bovine serum (FBS) as a chemoattractant. Following a 24-hour incubation period, cells were fixed with 4% paraformaldehyde. Migrated or invaded cells on the membrane were stained with 0.1% crystal violet for 30 minutes. Finally, the number of cells was quantified by counting five randomly selected visual fields per membrane under a microscope.

RNA was extracted from both cellular and tissue samples with TRIzol reagent (Cat. No. 15596026CN, Invitrogen, Waltham, MA, USA). Subsequent quantitative reverse transcription PCR (qRT-PCR) was performed using SYBR Green master mix (Cat. No. 4309155, Applied Biosystems, Foster City, CA, USA). The relative expression levels of target genes were calculated via the 2^{-Δ⁢Δ⁢CT} method, with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) serving as the internal reference. All primer sequences are listed in Table 1.

Protein samples were obtained from tissue specimens or cultured cells using RIPA buffer (Cat. No. P0013B, Beyotime, Shanghai, China). Subsequently, the proteins were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then electrophoretically transferred onto PVDF membranes (Cat. No. IPVH00010, Millipore, Burlington, MA, USA). Following a 2-hour blocking step with 5% skimmed milk at room temperature, the membranes were incubated overnight at 4 °C with the specified primary antibodies. The specific primary antibodies utilized were: Anti-ASS1 (1:5000, 16210-1-AP, Proteintech, Chicago, IL, USA), anti-CPT1A (1:5000, 15184-1-AP, Proteintech, Chicago, IL, USA), anti-S100A10 (1:5000, 81017-1-RR, Proteintech, Chicago, IL, USA), anti-FLAG® M2 antibody (1:2000, F1804, Sigma-Aldrich, St. Louis, MO, USA), anti-HA antibody (1:1000, H6908, Sigma-Aldrich, St. Louis, MO, USA), anti-succinyl lysine (1:1000, PTM-401, PTM Bio, Chicago, IL, USA), and anti-GAPDH (1:20,000, 10494-1-AP, Proteintech, Chicago, IL, USA). After washing, the membranes were subjected to a 2-hour incubation at room temperature with the appropriate HRP-linked secondary antibodies. These included goat anti-rabbit immunoglobulin G (IgG) (H+L) conjugated with HRP (1:1000, A0208, Beyotime, Shanghai, China) and goat anti-mouse IgG (H+L) conjugated with HRP (1:1000, A0216, Beyotime, Shanghai, China). An improved chemiluminescence detection technique was used to observe protein bands.

CD8⁺ T cells were isolated using an immunomagnetic bead sorting kit (EasySep Human CD8+ T Cell Isolation Kit, Cat. No. 17953, STEMCELL Technologies, Vancouver, BC, Canada). Briefly, following the manufacturer’s protocol, biotinylated anti-human CD8 antibody was added to peripheral blood mononuclear cells (PBMCs, obtained from Shanghai Jinyuan Biotechnology Co., Ltd., Shanghai, China ) resuspended in sorting buffer, and the mixture was incubated at room temperature for 15 minutes. After centrifugation and supernatant removal, the cell pellet was resuspended in fresh sorting buffer. Streptavidin-conjugated magnetic beads were then introduced, followed by another 15-minute incubation at room temperature. Subsequently, the tube was placed on a magnetic stand for five minutes to allow the bead-bound CD8⁺ T cells to adhere to the tube wall. The unbound cells in the supernatant were carefully aspirated and discarded. Finally, to obtain the purified CD8⁺ T cells, 1 mL of complete medium was added to resuspend the isolated cells.

A CD8⁺ T Cell Isolation Kit was used to separate human CD8⁺ T cells from PBMCs. Before the test, the cells were kept in RPMI-1640 medium supplemented with 10% FBS and 100 U/mL IL-2 (Cat. No. 200-02, PeproTech, Cranbury, NJ, USA) for 72 hours after being stimulated with anti-CD3/CD28 beads (Cat. No. 11131D, Thermo Fisher Scientific, Waltham, MA, USA). During the logarithmic growth phase, HCT116 or HT-29 CRC cells were collected. Then, 100 µL of CRC cells (target cells, T) were co-cultured with 100 µL of activated CD8⁺ T cells (effector cells, E) at an E:T ratio of 5:1 in a 96-well round-bottom cell culture plate. Wells containing target cells alone served as a negative control. Maximum release wells were treated with 100 µL of 1% NP-40. The following formula was used to determine the percentage of LDH release following a 48-hour co-culture at 37 °C with 5% CO₂: LDH release = [(OD experimental group – OD natural release) / (OD maximum release group – OD natural release)] $\times$ 100%.

Commercial sandwich ELISA kits were used to measure the amounts of interferon (IFN)-$\gamma$ (Cat. No. CB10293-Hu, COIBO BIO, Shanghai, China), tumor necrosis factor (TNF)-$\alpha$ (Cat. No. CB11762-Hu, COIBO BIO, Shanghai, China), granzyme-B (Cat. No. ab235635, Abcam, Cambridge, UK), and perforin (Cat. No. ab46068, Abcam, Cambridge, UK) in the supernatants of co-culture systems (centrifuged at 1000 $\times$g for 10 minutes). The assays were carried out in accordance with the manufacturer’s instructions. A BioTek Synergy H1 microplate reader was used to detect absorbance at 450 nm with a correction at 570 nm.

Following co-culture, cells were extracted and stained for flow cytometry to assess the CD8⁺ T cell that expressed Ki-67 and PD-1. An anti-CD8 antibody (Human: 1:50, Cat. No. 557834; Mouse: 1:50, Cat. No. 553035; BD Biosciences, San Jose, CA, USA) was used for surface staining following viability staining to rule out dead cells. A fluorescently conjugated antibody was used to detect PD-1 (Human: 1:100, Cat. No. 561272; Mouse: 1:100, Cat. No. 570654; BD Biosciences, San Jose, CA, USA) or TIM-3 (Human: 1:100, Cat. No. 571403; Mouse: 1:100, Cat. No. 569935; BD Biosciences, San Jose, CA, USA). Following fixation and permeabilization, the cells were intracellularly labeled using an anti-Ki-67 antibody (1:50, Cat. No. 556026, BD Biosciences, San Jose, CA, USA). Appropriate fluorescence-minus-one (FMO) controls and isotype controls were included in each experiment to establish positive gates and confirm staining specificity. The percentages of Ki-67⁺ and PD-1⁺ subsets were calculated by gating CD8⁺ T cells using data collected on a flow cytometer and analyzed using FlowJo software (v10, BD Biosciences, San Jose, CA, USA).

Protease inhibitor-supplemented RIPA buffer was used to extract the total protein, which was then incubated for 30 minutes and centrifuged for 15 min at 12,000 $\times$g. After adding 50 µL of Protein A/G agarose beads and 2 µg of normal IgG to the resultant lysate, it was gently shaken at 4 °C for one hour. The anti-CPT1A antibody was added to the supernatant and incubated at 4 °C for the entire night after centrifugation was completed to eliminate non-specific binding. After that, 50 µL of Protein A/G agarose beads (Cat. No. 20422, Thermo Fisher Scientific, Waltham, MA, USA) were added, and they were given time to bind. SDS loading buffer was added to the eluates following elution. Western blotting was utilized to evaluate the immunoprecipitated products, and S100A10 was detected using an anti-S100A10 antibody. To guarantee result specificity, the IgG group acted as a negative control and the input sample as a positive control. Total protein extracts from the transfected HEK293T, HT29, or HCT116 cells were prepared using the method described above. Protein was used for immunoprecipitation with anti-S100A10 antibody, following the same procedure as above. The total succinylation level and the succinylation level at the K47 site in the eluted products were detected by Western blotting. Protein expression was verified via the input sample.

Succinylation mutants of S100A10, in which Lysine 47 was substituted with arginine (K47R), were generated by Cenepharma (Shanghai, China). All mutant expression plasmids were generated by Sangon Biotech (Shanghai, China). For the overexpression of Flag-CPT1A, HA-S100A10, and its mutant HA-S100A10-K47R, the respective constructs were transfected into HEK293T cells using Lipofectamine 3000 (Cat. No. L3000015, Invitrogen, Waltham, MA, USA), following the manufacturer’s protocol. Anti-Flag M2 or anti-HA antibodies coupled to magnetic beads were used to immunoprecipitate proteins, which were then eluted with the appropriate Flag or HA peptides. The immunoprecipitated samples (Flag-CPT1A, HA-S100A10, and HA-S100A10-K47R) were incubated in reaction buffer at 37 °C for 1 h, followed by protein immunoblotting analysis.

In accordance with the experimental groups, HCT116 or HT29 cells were plated in 6-well plates and transfected. Fresh media containing 50 µg/mL cycloheximide (CHX, Cat. No. C7698, Sigma-Aldrich, St. Louis, MO, USA) was added after 48 hours of culture to allow target gene expression. At 2, 4, 6, and 8 hours after CHX treatment, cells were taken. Then, protein lysates were prepared, quantified, and subjected to Western blotting to evaluate S100A10 degradation kinetics.

C57BL/6 mice (6–8 weeks, 20–22 g) were acquired from Changzhou Cavins Laboratory Animal Co., Ltd (Changzhou, Jiangsu, China). This research received ethical approval from the Ethics Committee of the Central Hospital of Zibo (Protocol No. IACUC-ZBCH-2024-110B). Twelve C57BL/6 mice were randomly divided into 2 groups, with 6 mice per group. Then, 5 $\times$ 10⁵ ASS1-knockdown CT26 cells were subcutaneously injected to establish transplanted tumor models, designated as sh-NC and sh-ASS1 groups, respectively. The mice were monitored daily after inoculation. Tumor dimensions were measured every third day using a digital caliper, recording the longest (L) and shortest (W) diameters. Tumor volume (mm³) was then calculated using the formula: Volume = L $\times$ W² / 2. At the conclusion of the experiment, all mice were euthanized by exposure to 40% CO₂ followed by cervical dislocation. Tumor volume measurements and endpoint analyses were performed by investigators blinded to group allocation. No animals were excluded from this analysis.

Collagenase IV (1 mg/mL, Cat. No. 17104019, Gibco, Waltham, MA, USA) and DNase I (0.1 mg/mL, Cat. No. 18047019, Gibco, Waltham, MA, USA) were used to mechanically and enzymatically break down tumor tissue in RPMI-1640 media for 30 minutes at 37 °C in order to produce single-cell suspensions. After passing through a 70 µm strainer, the cells were cleaned. Antibodies against CD3 and CD8 were used to stain the cells in order to analyze CD8 T cell invasion. The fraction of CD8⁺ T/CD3⁺ cells in the live leukocytes was calculated after fixation. Using a commercial isolation kit, isolated tumor-infiltrating leukocytes were selected for CD8⁺ T cells for PD-1 expression. Antibodies against PD-1 and CD8 were used to stain the cells. There were also isotype controls. PD-1 expression in gated CD8⁺ T lymphocytes was measured, and samples were examined using flow cytometry.

All statistical analyses were conducted with GraphPad Prism (v9.5, GraphPad Software, Inc., San Diego, CA, USA). Results are expressed as mean $\pm$ standard deviation. For comparisons between two groups, unpaired Student’s t-tests were applied, whereas comparisons across multiple groups were performed using one-way analysis of variance. A statistically significant p-value was defined as ˂0.05.

To assess ASS1 expression in CRC, we first performed a bioinformatics analysis of TCGA data, which exhibited significant ASS1 mRNA upregulation in tumor tissues compared to normal samples (Fig. 1A). To validate this finding, we collected 40 pairs of tumor and matched para-carcinoma tissues from patients with CRC. qRT-PCR results confirmed that ASS1 mRNA expression was significantly elevated in tumor tissues relative to matched paracancerous tissues (Fig. 1B). Similarly, Western blotting of five representative paired samples revealed increased ASS1 protein expression in tumor tissues (Fig. 1C,D). IHC staining further supported these results, exhibiting weak ASS1 staining in para-carcinoma tissues but strong staining in CRC specimens (Fig. 1E,F). Moreover, in cellular models, CRC cell lines (HCT116, SW480, HT29, and CT26) exhibited considerably higher levels of ASS1 mRNA and protein expression than NCM460 (Fig. 1G–I). Collectively, these data demonstrate elevated expression of ASS1 in both CRC tissues and cellular models, suggesting its potential role as an oncogenic driver in colorectal carcinogenesis.

Fig. 1.

ASS1 is highly expressed in colorectal cancer (CRC) tissues and cell lines. (A) ASS1 mRNA expression levels in CRC and normal tissues from the The Cancer Genome Atlas (TCGA) database (n = 349). (B) qRT-PCR analysis of ASS1 mRNA expression in 40 pairs of CRC tumors and matched paracancerous tissues (n = 40). (C,D) Western blotting of ASS1 protein expression in five representative pairs of CRC and adjacent non-tumor tissues. GAPDH was used as a loading control (n = 6). (E,F) Representative immunohistochemistry (IHC) staining of ASS1 in CRC and paracancerous tissues (n = 40, normal, 200 µm, enlarged, 100 µm). **p 0.01, compared with para-carcinoma. (G–I) ASS1 mRNA (n = 6) and protein (n = 3) expression levels in CRC cell lines (HCT116, SW480, HT29, and CT26) and normal colonic epithelial cell line (NCM460). **p 0.01, compared with NCM460.

To study the biological functions of ASS1 in CRC, we generated stable overexpression (ASS1-OE) and knockdown (sh-ASS1 #1 and sh-ASS1 #2) models in SW480, HCT116, and HT29 cell lines. We utilized Western blotting and qRT-PCR in parallel to verify ASS1 expression (Fig. 2A,B). ASS1 knockdown reduced cell proliferation, colony formation, migration, and invasion in comparison to the control group, according to functional assays. Conversely, ASS1 overexpression notably enhanced cell proliferation and clonogenicity (Fig. 2C–F). Together, these results show that ASS1 deletion reduces important malignant behaviors in CRC cells, such as invasion, migration, colony formation, and proliferation, indicating that ASS1 functions as an oncogenic driver supporting CRC progression.

Fig. 2.

ASS1 promotes the malignant phenotypes of CRC cells in vitro. (A,B) Efficiency of ASS1 overexpression (ASS1-OE) and knockdown (sh-ASS1 #1 and #2) in SW480, HCT116, and HT29 cells was confirmed by qRT-PCR (n = 6) and Western blotting (n = 3). (C–F) Functional assays revealed that ASS1 knockdown significantly inhibited cell proliferation (C, n = 6), colony formation (D, n = 6, scale bar: 10 mm), migration (E, n = 5, scale bar: 100 µm), and invasion (F, n = 5, scale bar: 100 µm), whereas ASS1 overexpression enhanced proliferative and colony-forming capabilities compared to the controls. **p 0.01, compared with control.

CD8⁺ T cells were purified to a purity exceeding 95%, as confirmed by flow cytometric analysis (Fig. 3A and Supplementary Fig. 1A). In comparison to those co-cultured with sh-NC control cells, CD8⁺ T cells stimulated with ASS1-knockdown HCT116 cells (sh-ASS1) showed noticeably higher release of cytotoxic mediators, such as IFN-$\gamma$, TNF-$\alpha$, perforin, and granzyme B (Fig. 3B–E). We used flow cytometry to further assess PD-1 expression, which is higher in fatigued T cells [19]. The results revealed enhanced proliferation (indicated by Ki67⁺ frequency) and effector function, along with reduced PD-1 and TIM-3 expression, in CD8⁺ T cells cultured with ASS1-depleted cells (Fig. 3F,G and Supplementary Fig. 1B–D). Functional assessment revealed that ASS1 knockdown significantly sensitized CRC cells to CD8⁺ T cell-mediated killing, as evidenced by the increased LDH release (Fig. 3H). Conversely, ASS1 overexpression conferred resistance to CD8⁺ T cell cytotoxicity (Fig. 3B–H). All of these findings show that ASS1 depletion improves CD8⁺ T cell anti-tumor activity and reduces T cell fatigue.

Fig. 3.

ASS1 expression modulates CD8⁺ T cell function against CRC cells. (A) Purity analysis of isolated human CD8⁺ T cells. (B–E) Secretion of cytotoxic mediators by CD8⁺ T cells after co-culture with sh-ASS1 or sh-NC CRC cells (n = 6). (F,G) Proliferation (Ki67) and exhaustion markers (PD-1, TIM-3) on CD8⁺ T cells were analyzed using flow cytometry (n = 5). (H) Functional assessment of CD8⁺ T cell-mediated killing via LDH release assays (n = 6). **p 0.01, compared with control or sh-NC.

Analysis of 40 clinical CRC samples revealed significantly elevated mRNA and protein levels of CPT1A in tumor tissues compared to matched para-carcinoma tissues (Fig. 4A,B). IHC analysis further supported these observations, exhibiting faint CPT1A immunoreactivity in paracancerous regions; however, strong positive staining was observed in CRC tissues (Fig. 4C). Correlation analysis revealed a significant positive association between ASS1 and CPT1A protein expression levels (Fig. 4D). We transfected ASS1-targeting shRNA into HCT116 cells to examine the regulatory mechanisms underlying this relationship. Downregulation of CPT1A expression was observed following ASS1 knockdown, with a notable decline at both the protein and mRNA levels. Conversely, ASS1 overexpression in SW480 cells substantially increased CPT1A protein levels (Fig. 4E,F). To investigate how ASS1 regulates CPT1A, we focused on its role in arginine synthesis. First, we determined that ASS1 knockdown significantly reduced intracellular arginine levels (Fig. 4G). We next asked whether this arginine reduction mediated the downregulation of CPT1A. Exogenous arginine supplementation in ASS1-deficient cells completely rescued the downregulation of CPT1A mRNA and protein levels (Fig. 4H,I). Conversely, supplementation with a control amino acid (glycine) was ineffective. These results demonstrate that ASS1 controls CPT1A transcription by maintaining intracellular arginine levels through its enzymatic activity.

Fig. 4.

ASS1 positively regulates CPT1A expression in CRC. (A,B) mRNA (n = 40) and protein (n = 5) levels of CPT1A were significantly upregulated in tumor tissues compared to matched para-carcinoma tissues from 40 patients with CRC. (C) Representative IHC images exhibited weak CPT1A staining in para-carcinoma tissues and strong staining in CRC tissues (n = 40, normal, 200 µm; enlarged, 100 µm). **p 0.01, compared with para-carcinoma. (D) A positive correlation between ASS1 and CPT1A protein expressions in clinical CRC samples. (E,F) Western blotting (n = 3) and qRT-PCR (n = 6) analyses confirmed that ASS1 knockdown downregulated CPT1A protein and mRNA levels in HCT116 cells, whereas ASS1 overexpression upregulated CPT1A expression in SW480 cells. (G) Intracellular arginine levels in CRC cell lines (HCT116 and HT29) were determined using a colorimetric assay. (H,I) CPT1A protein (n = 3) and mRNA (n = 6) expression levels in CRC cell lines (HCT116 and HT29). **p 0.01, compared with sh-NC or NC-OE, ^#⁢#p 0.01, compared with sh-ASS1; ns, not significant.

To evaluate functional rescue, we established CPT1A-overexpressing HCT116 and HT29 cell lines (Fig. 5A). ASS1 knockdown dramatically reduced cell proliferation, clonogenicity, migration, and invasion in both cell lines, according to functional experiments. These inhibitory effects were effectively reversed by CPT1A overexpression (Fig. 5B–E). In co-culture assays, ASS1 knockdown potentiated the anti-tumor function of CD8⁺ T cells in co-culture, as demonstrated by two key findings: elevated secretion of effector molecules (IFN-$\gamma$, TNF-$\alpha$, perforin, and granzyme B) (Fig. 5F) and a decreased proportion of PD-1⁺ (or TIM-3⁺) exhausted T cells (Fig. 5G and Supplementary Fig. 2A,B). In line with this enhanced cytotoxicity, ASS1-deficient tumor cells showed significantly increased LDH release upon co-culture (Fig. 5H). CPT1A overexpression attenuated these immunostimulatory effects (Fig. 5F–H). All of these findings point to CPT1A as a crucial downstream effector that ASS1 uses to encourage immune evasion in CRC.

Fig. 5.

CPT1A overexpression rescues the tumor-suppressive and immunostimulatory effects of ASS1 knockdown in CRC cells. (A) Establishment of CPT1A-overexpressing HCT116 and HT29 cell lines (n = 3). (B–E) Functional assays exhibited that ASS1 knockdown inhibited proliferation (B, n = 6), colony formation (C, n = 5, Scale bar: 10 mm), migration (D, n = 5, Scale bar: 100 µm), and invasion (E, n = 5, Scale bar: 100 µm) in CRC cells, which were reversed by CPT1A overexpression. (F) Increased secretion of CD8⁺ T cell effector molecules (interferon [IFN]-$\gamma$, tumor necrosis factor [TNF]-$\alpha$, perforin, and granzyme B) upon co-culture with ASS1-knockdown tumor cells, attenuated by CPT1A overexpression (n = 6). (G) Reduced frequency of exhausted PD-1⁺/CD8⁺ T cells (or TIM-3⁺/CD8⁺ T cells) after co-culture with ASS1-deficient cells, rescued by CPT1A expression (n = 5). (H) Enhanced CD8⁺ T cell-mediated killing of ASS1-knockdown tumor cells, measured by LDH release, and counteracted by CPT1A overexpression (n = 6). **p 0.01, compared with sh-NC. ^#⁢#p 0.01, compared with sh-ASS1.

We further studied S100A10 expression in CRC clinical samples. Elevated S100A10 mRNA levels in tumors versus paracancerous tissues were detected by qRT-PCR (Fig. 6A). Consistent with the mRNA data, Western blotting confirmed that S100A10 protein was also markedly upregulated in tumor samples (Fig. 6B, and Supplementary Fig. 3A). IHC analysis corroborated these results, demonstrating weak S100A10 staining in paracancerous areas but intense positive staining in CRC tissues (Fig. 6C). Strong significant correlations between S100A10 expression and CPT1A and ASS1 expression were found by correlation analysis (Fig. 6D). Importantly, elevated expression of each gene (ASS1, CPT1A, or S100A10) correlated with advanced tumor stage (Stage III/IV) (Supplementary Fig. 4A–F). Furthermore, Kaplan-Meier analysis revealed that high expression of these genes was individually associated with significantly shorter overall survival in CRC patients (Supplementary Fig. 4G–I). In HT29 and HCT116 cells, co-IP studies verified a direct connection between CPT1A and S100A10 (Fig. 6E). Immunoprecipitation using an anti-CPT1A antibody efficiently enriched the S100A10. When Flag-CPT1A and HA-S100A10 were co-overexpressed in HEK293T cells, strong S100A10 succinylation was observed. Conversely, the K47R mutant exhibited a significantly reduced succinylation (Fig. 6F). Similarly, CPT1A overexpression in CRC cell lines HT29 and HCT116 markedly enhanced S100A10 succinylation, while the K47R mutant revealed substantially less succinylation (Fig. 6G). CHX-chase assays indicated that CPT1A overexpression delayed the S100A10 protein degradation, suggesting that CPT1A stabilizes S100A10 via promoting its succinylation (Fig. 6H,I, and Supplementary Fig. 3B,C). Even under CPT1A overexpression, the K47R mutant was rapidly degraded, implying that succinylation at lysine 47 is essential for maintaining S100A10 protein stability. Additional experiments suggested that ASS1 regulates S100A10 expression through CPT1A. Specifically, ASS1 knockdown significantly reduced S100A10 protein levels, and this reduction was reversed by CPT1A overexpression (Fig. 6J,K, and Supplementary Fig. 3D,E). Together, these findings demonstrate that CPT1A promotes succinylation of S100A10 at K47, thereby stabilizing the protein and upregulating its expression in CRC cells.

Fig. 6.

CPT1A-mediated succinylation stabilizes S100A10 and promotes its expression in CRC cells. (A) qRT-PCR analysis of S100A10 mRNA expression in CRC tumors and matched paracancerous tissues (n = 40). (B) Western blotting of S100A10 protein levels in clinical CRC samples and paracancerous tissues (n = 5). (C) Representative IHC images exhibited S100A10 expression in CRC and paracancerous tissues (n = 40, normal, 200 µm; enlarged, 100 µm). (D) Correlation analysis between S100A10 expression and CPT1A or ASS1 levels in CRC samples. (E) Co-immunoprecipitation (Co-IP) assays confirmed the direct interaction between CPT1A and S100A10 in HT29 and HCT116 cells. (F) Succinylation of S100A10 in HEK293T cells co-expressing Flag-CPT1A and HA-S100A10 wild-type (WT) or K47R mutant. (G) Enhanced succinylation of S100A10 WT but not K47R mutant after CPT1A overexpression in HT29 and HCT116 cells. (H,I) Cycloheximide (CHX) chase assays exhibited the S100A10 protein degradation with or without CPT1A overexpression (n = 3). (J,K) Western blotting revealed that ASS1 knockdown reduced S100A10 protein levels, which could be rescued by CPT1A overexpression (n = 3). **p 0.01, compared with para-carcinoma.

ASS1 knockdown dramatically reduced invasion, migration, colony formation, and cell proliferation. These inhibitory effects were markedly reversed by S100A10 overexpression (Fig. 7A–D). In a co-culture system with CD8⁺ T cells, ASS1 knockdown significantly enhanced CD8⁺ T cell cytotoxicity against HCT116 cells. This enhanced cytotoxicity was reversed by S100A10 overexpression, as indicated by the restoration of cytotoxic factor secretion to baseline levels and a significant increase in the PD-1⁺/CD8⁺ and TIM-3⁺/CD8⁺ T cell proportion (Fig. 7E,F). Consistent with these findings, ASS1-knockdown tumor cells revealed a significantly higher LDH release rate when co-cultured with CD8⁺ T cells, which was attenuated after S100A10 overexpression (Fig. 7G).

Fig. 7.

S100A10 overexpression rescues the tumor-promotive and immune-evasive functions impaired by ASS1 knockdown. (A) Western blotting of S100A10 in HCT116 cells (n = 3). (B) CCK-8 assay (n = 6). (C) Colony formation assay (n = 6, scale bar: 10 mm). (D) Transwell migration/invasion assay (n = 5). Scale bar: 100 µm. (E) Enzyme-linked immunosorbent assay (ELISA) analysis of cytotoxic factors (IFN-$\gamma$, TNF-$\alpha$, perforin, and granzyme B in a co-culture system (n = 6). (F) PD-1⁺ (or TIM-3⁺) exhausted CD8⁺ T cells rate (n = 5). (G) LDH release rate (n = 6). **p 0.01, compared with sh-NC. ^#⁢#p 0.01, compared with sh-ASS1.

According to experimental results, the sh-ASS1 group showed considerably lower tumor volume and weight than the sh-NC group in a CT26 cell transplantation tumor model created in C57BL/6 mice (Fig. 8A–C). IHC and Western blotting revealed markedly decreased ASS1, CPT1A, and S100A10 expression levels in tumor tissues from the sh-ASS1 group (Fig. 8D,E). Flow cytometry results indicated a significant increase in the CD8⁺ T cell proportion and a notable decrease in the PD-1⁺ rate and TIM-3⁺ rate among tumor-infiltrating lymphocytes in the sh-ASS1 group compared to the sh-NC group (Fig. 8F,G and Supplementary Fig. 5A–C). ELISA revealed that sh-ASS1 CT26 cell-derived tumors contained significantly elevated levels of the cytotoxic effector molecules IFN-$\gamma$ and granzyme B compared to sh-NC controls (Fig. 8H). Consistent with these findings, the LDH release rate in the sh-ASS1 group was significantly increased (Fig. 8I). These findings suggest that ASS1 knockdown suppresses colorectal tumor growth in vivo and promotes a more anti-tumor immune microenvironment.

Fig. 8.

ASS1 knockdown inhibits tumor growth and modulates the immune microenvironment in a CT26 murine CRC model. (A) Tumor appearance in each group (n=6). (B,C) Tumor volume curve (B) and tumor weight (C) were significantly reduced in mice injected with CT26 cells stably expressing sh-ASS1 compared to the sh-NC group (n = 6). (D) Representative IHC staining and quantitative analysis exhibited decreased protein expression of ASS1, CPT1A, and S100A10 in the sh-ASS1 group tumor tissues (n = 6). Scale bar: 100 µm. (E) Western blotting of ASS1, CPT1A, and S100A10 in tumor tissues (n = 3). (F,G) Flow cytometry analysis revealed an increased CD8⁺ T cell proportion (F) and a decreased PD-1⁺ rate (or TIM-3⁺ rate) among tumor-infiltrating lymphocytes (G) in the sh-ASS1 group compared to sh-NC (n = 6). (H) Enzyme-linked immunosorbent assay (ELISA) analysis of cytotoxic factors (IFN-$\gamma$, TNF-$\alpha$, perforin, and granzyme B in tumor tissues (n = 6). (I) LDH release rate (n = 6). **p 0.01, compared with sh-NC.