Oroxin A suppresses colorectal tumor growth by regulating the TRIM24-mediated ferroptosis and TSPO pathway.

T.L., G.L., and W.L. conceived and supervised the study. C.H. and H.L. contributing to study design and data interpretation. C.H. and C.W. designed, conducted experiments, and drafted the manuscript. K. C. and Z.L. performed the pathology analyses and interpretation and contributed to manuscript drafting. All authors thoroughly reviewed and approved the final manuscript.

We validated the overexpression of TRIM24 mRNA in tumors compared to adjacent normal tissues across various cancers using the GEPIA database ( http://gepia2.cancer-pku.cn/ ) ( Figure 1 A). Notably, TRIM24 expression exhibited a significant upregulation in tumor tissues ( p < 0.01). Similarly, TRIM24 expression in colon adenocarcinoma (COAD) and rectum adenocarcinoma (READ) was markedly elevated when compared to normal tissues ( Figures 1 B and 1C). Immunohistochemistry (IHC) images from the The Human Protein Atlas database ( https://www.proteinatlas.org/ ) further confirmed that TRIM24 expression in colorectal cancer cells was significantly elevated compared to normal epithelial cells ( Figure 1 D). We further collected tumor and adjacent normal tissues from five colorectal cancer patients at The Affiliated Hospital of Hangzhou Normal University for western blot analysis. The results demonstrated a significantly higher expression of TRIM24 in tumor tissues compared to adjacent normal tissues ( Figure 1 E). Survival analysis was conducted for the canonical CMS subtype COAD, prevalent in the left colon, utilizing the Kaplan-Meier Plotter database ( https://kmplot.com/analysis/index.php?p=service&cancer=colon ). Elevated TRIM24 expression was found to be significantly associated with a poorer prognosis ( Figure 1 F). Moreover, analyses conducted using the Timer2.0 database ( http://timer.cistrome.org/ ) revealed a positive correlation between TRIM24 expression and the presence of CD8 + T cells and natural killer (NK) cells within the colorectal adenocarcinoma (COAD) tumor microenvironment. Conversely, a negative correlation was observed in the case of regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs) ( Figures 1 G–1J). In summary, these findings suggest that the heightened expression of TRIM24 in colorectal cancer is intricately linked to an unfavorable prognosis. Figure 1 TRIM24 is highly expressed in CRC and predicts poor prognosis (A) TRIM24 mRNA expression in GEPIA database. (B and C) TRIM24 expression in CRC tumor tissue versus normal tissue. (D) Representative IHC images of TRIM24 expression in normal colon and rectum tissues versus tumor tissues from the THPA database. (E) Western blot analysis of TRIM24 expression in tumor and adjacent non-tumor tissues from CRC patients at the affiliated hospital of Zhejiang normal university. (F) KM survival curves for CRC patients based on TRIM24 expression in Kaplan-Meier plotter database. (G and H) Timer2.0 database correlation between TRIM24 expression and CD8 + T cells or NK cells in COAD tumor microenvironments. (I and J) Timer2.0 database correlation between TRIM24 expression and Tregs or MDSCs in COAD tumor microenvironments. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. TRIM24 is highly expressed in CRC and predicts poor prognosis (A) TRIM24 mRNA expression in GEPIA database. (B and C) TRIM24 expression in CRC tumor tissue versus normal tissue. (D) Representative IHC images of TRIM24 expression in normal colon and rectum tissues versus tumor tissues from the THPA database. (E) Western blot analysis of TRIM24 expression in tumor and adjacent non-tumor tissues from CRC patients at the affiliated hospital of Zhejiang normal university. (F) KM survival curves for CRC patients based on TRIM24 expression in Kaplan-Meier plotter database. (G and H) Timer2.0 database correlation between TRIM24 expression and CD8 + T cells or NK cells in COAD tumor microenvironments. (I and J) Timer2.0 database correlation between TRIM24 expression and Tregs or MDSCs in COAD tumor microenvironments. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. Utilizing the crystal structure of TRIM24, we successfully identified a previously unidentified chemical scaffold with potential as a TRIM24 inhibitor—OA. Molecular docking analyses revealed that OA establishes five hydrogen bonds with residues GLU879, GLU864, GLU877, and ASP881, forming effective bonds with TRIM24 at distances ranging from 1.9 Å to 2.1 Å (depicted by yellow dashed lines, Figures 2 A–2C). Subsequent cellular thermal shift assays illustrated that, in the control group, TRIM24 protein underwent degradation with increasing temperature. In contrast, the interaction between OA and TRIM24 resulted in a deceleration of TRIM24 protein degradation ( Figures 2 E–2H). Drug affinity reactive target stability assay also confirmed that OA binds to TRIM24, rather than β-actin ( Figures 2 I and 2J). Furthermore, high concentrations of OA treatment significantly downregulated TRIM24 expression in CRC cells ( Figures 2 K–2L). Collectively, these findings indicate that OA exhibits the potential to selectively degrade TRIM24 in CRC cells. Figure 2 OA can regulate TRIM24 and reduce TRIM24 expression (A–C) Virtual docking of TRIM24 and OA. (D) The structure of OA. (E–H) Cellular thermal shift assay was performed with HCT-8 and SW-480 to evaluate the binding of OA and TRIM24. (I and J) Drug affinity reactive target stability assay was performed to exhibit the binding between TRIM24 and OA in HCT-8 and SW-480 cells. (K and L) Western blot assay was used for evaluating the inhibitory effect of OA and TRIM24 in HCT-8 and SW-480 CRC cells. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. OA can regulate TRIM24 and reduce TRIM24 expression (A–C) Virtual docking of TRIM24 and OA. (D) The structure of OA. (E–H) Cellular thermal shift assay was performed with HCT-8 and SW-480 to evaluate the binding of OA and TRIM24. (I and J) Drug affinity reactive target stability assay was performed to exhibit the binding between TRIM24 and OA in HCT-8 and SW-480 cells. (K and L) Western blot assay was used for evaluating the inhibitory effect of OA and TRIM24 in HCT-8 and SW-480 CRC cells. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. Various concentrations of OA were administered to four CRC cell lines and one normal colon epithelial cell line to elucidate the impact of OA on CRC cell viability. An evident inhibitory effect of OA on CRC cells was observed, notably contrasting with its impact on normal colon epithelial cells (IC50 = 4.028–6.296 μM vs. 20.65 μM) ( Figure 3 A). Colony formation assays further illustrated a pronounced inhibition rate for HCT-8 and SW-480 cells treated with OA ( Figures 3 B and 3C). Additionally, wound healing experiments revealed that OA effectively impeded the migration of HCT-8 and SW-480 cell lines ( Figures 3 D–3F). Similarly, in transwell assays, OA significantly inhibited the migration and invasion of HCT-8 and SW-480 cells ( Figures 3 G–3J). These findings collectively suggest that OA can effectively suppress the phenotypic characteristics of CRC cells without inducing harm to normal colon epithelial cells. Figure 3 OA inhibited the CRC cell proliferation, migration, and invasion in vitro (A) OA affected the viability of CRC cell lines via the CCK8 assay. Treatment with OA decreased HCT-8 and SW-480 colony-formation (B–C), wound-healing (D–F), migration and invasion (G–J) compared with their control group, respectively. Scale bar, 50 μm. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. OA inhibited the CRC cell proliferation, migration, and invasion in vitro (A) OA affected the viability of CRC cell lines via the CCK8 assay. Treatment with OA decreased HCT-8 and SW-480 colony-formation (B–C), wound-healing (D–F), migration and invasion (G–J) compared with their control group, respectively. Scale bar, 50 μm. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. To unravel the molecular mechanisms underlying the effects of OA, a quantitative proteomic analysis was conducted on HCT-8 CRC cells treated with OA. A total of 857 proteins were identified and annotated across all samples. Principal-component analysis (PCA) revealed significant alterations in PC1 and PC2 for the OA group compared to the control group ( Figure 4 A). Following meticulous screening, 78 differentially expressed proteins were co-identified ( Figure 4 B). The OA group exhibited an upregulation of 8 proteins and a downregulation of 70 proteins compared to the control group. Figure 4 Proteomic sequencing analysis revealed OA involved in various biological process in HCT-8 (A) PCA of proteomes represented in two dimensions with or without OA treatment. (B) The volcano map shows differentially expressed proteins with or without OA treatment in HCT-8. (C and D) The KEGG and GO analysis for the significantly differentially proteins after OA treatment in HCT-8. (E and F) Heat maps of differentially expressed proteins associated with ferroptosis and mitochondria. Proteomic sequencing analysis revealed OA involved in various biological process in HCT-8 (A) PCA of proteomes represented in two dimensions with or without OA treatment. (B) The volcano map shows differentially expressed proteins with or without OA treatment in HCT-8. (C and D) The KEGG and GO analysis for the significantly differentially proteins after OA treatment in HCT-8. (E and F) Heat maps of differentially expressed proteins associated with ferroptosis and mitochondria. Moreover, the Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis unveiled that differentially expressed proteins post-OA treatment were predominantly associated with ferroptosis, peroxisome, oxidative phosphorylation, and metabolism ( Figure 4 C). GO analysis indicated that these proteins are involved in processes such as the regulation of cellular metabolic processes, regulation of mitochondrion organization, and nucleosome-dependent ATPase activity ( Figure 4 D). To comprehensively understand the regulatory effects of OA, a heatmap of differentially expressed proteins relevant to the aforementioned processes were constructed ( Figure 4 E). The results suggested that OA could modulate ferroptosis and mitochondrial-related proteins through the regulation of iron ion transport, including TFRC, SLC3A2, ATP6V1C1, and NDUFAB1 ( Figures 4 E and 4F). Cancer cells rely on the Xc-cystine transporter system, composed of SLC7A11 and SLC3A2, to facilitate the uptake of extracellular cysteine. This process supports glutathione synthesis, contributing to antioxidant defense and the inhibition of ferroptosis. 21 Analysis using the GEPIA database revealed a significant positive correlation between the mRNA expression levels of TRIM24 and SLC3A2 in CRC tissues ( Figure 5 A). Western blot analysis demonstrated a dose-dependent reduction in SLC3A2 protein expression in HCT-8 and SW-480 cells following treatment with OA ( Figures 5 B and 5C). Additionally, OA treatment significantly decreased intracellular GSH levels and glutathione peroxidase (GPX) activity, as determined using GSH/GSSG and GPX activity assay kits, respectively ( Figure 5 D). Correspondingly, protein levels of SLC7A11 and GPX4 were markedly reduced after OA exposure, whereas the expression of ACSL4 was significantly upregulated ( Figure 5 E). These protein expression changes were corroborated by RT-qPCR results, which showed that OA treatment significantly downregulated the mRNA levels of SLC3A2, SLC7A11, and GPX4, while upregulating ACSL4 mRNA expression ( Figure 5 F). Moreover, OA significantly elevated intracellular reactive oxygen species (ROS) and lipid peroxidation levels, indicating enhanced oxidative stress ( Figures 5 G and 5H). To further determine whether OA’s anti-tumor effects are mediated through ferroptosis, CRC cells were co-treated with OA and the ferroptosis inhibitor ferrostatin-1 (Fer-1). The inhibitory effect of OA on CRC cell proliferation was significantly reversed by Fer-1 ( Figure 5 I), supporting a role for ferroptosis in OA-induced cytotoxicity. Figure 5 OA treatment regulates ferroptosis related pathway and OA inhibited CRC cells growth by TRIM24 in vitro (A) The relationship between TRIM24 and SLC3A2 mRNA expression in the GEPIA database. (B and C) Western blot analysis of SLC3A2 expression in HCT-8 and SW-480 cells treated with OA for 24 h at concentrations of 0, 1, 2, and 5 μM. (D) GSH levels (left) and GPX activity (right) in HCT-8 and SW-480 cells following 24-h OA treatment. (E) Expression of key ferroptosis pathway proteins, SLC7A11, GPX4, and ACSL4, after treatment with different concentrations of OA (24 h at 0, 1, and 2 μM). (F) mRNA expression levels of SLC3A2, SLC7A11, GPX4, and ACSL4 after 24-h treatment with 2 μM OA, determined by RT-qPCR. (G) Representative lipid peroxide levels measured using a lipid peroxidation assay kit following 24-h treatment with 2 μM OA in HCT-8 and SW-480 cells. Scale bar, 25μm. (H) Representative results and statistical analysis of ROS levels following 2 μM OA treatment for 24 h in HCT-8 and SW-480 cells. (I) Cell viability measured by CCK-8 in HCT-8 and SW-480 cells treated with 2 μM OA, with or without the ferroptosis inhibitor Fer-1 (100 nM). (J) TRIM24 knockdown levels in HCT-8 and SW-480 cells treated with shRNA lentivirus. (K) Cell proliferation in HCT-8 and SW-480 cells after TRIM24 knockdown, measured by CCK-8 at 24 and 48 h. (L) Sensitivity of TRIM24 knockdown HCT-8 and SW-480 cells to ferroptosis inducer erastin, assessed by CCK-8. (M) Inhibition rate of OA on HCT-8 and SW-480 cells (control and TRIM24 knockdown) after 48 h of 2 μM OA treatment, measured by CCK-8. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. OA treatment regulates ferroptosis related pathway and OA inhibited CRC cells growth by TRIM24 in vitro (A) The relationship between TRIM24 and SLC3A2 mRNA expression in the GEPIA database. (B and C) Western blot analysis of SLC3A2 expression in HCT-8 and SW-480 cells treated with OA for 24 h at concentrations of 0, 1, 2, and 5 μM. (D) GSH levels (left) and GPX activity (right) in HCT-8 and SW-480 cells following 24-h OA treatment. (E) Expression of key ferroptosis pathway proteins, SLC7A11, GPX4, and ACSL4, after treatment with different concentrations of OA (24 h at 0, 1, and 2 μM). (F) mRNA expression levels of SLC3A2, SLC7A11, GPX4, and ACSL4 after 24-h treatment with 2 μM OA, determined by RT-qPCR. (G) Representative lipid peroxide levels measured using a lipid peroxidation assay kit following 24-h treatment with 2 μM OA in HCT-8 and SW-480 cells. Scale bar, 25μm. (H) Representative results and statistical analysis of ROS levels following 2 μM OA treatment for 24 h in HCT-8 and SW-480 cells. (I) Cell viability measured by CCK-8 in HCT-8 and SW-480 cells treated with 2 μM OA, with or without the ferroptosis inhibitor Fer-1 (100 nM). (J) TRIM24 knockdown levels in HCT-8 and SW-480 cells treated with shRNA lentivirus. (K) Cell proliferation in HCT-8 and SW-480 cells after TRIM24 knockdown, measured by CCK-8 at 24 and 48 h. (L) Sensitivity of TRIM24 knockdown HCT-8 and SW-480 cells to ferroptosis inducer erastin, assessed by CCK-8. (M) Inhibition rate of OA on HCT-8 and SW-480 cells (control and TRIM24 knockdown) after 48 h of 2 μM OA treatment, measured by CCK-8. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. To elucidate whether OA-induced ferroptosis is mediated by TRIM24, we knocked down TRIM24 expression using shRNA ( Figure 5 J). TRIM24 knockdown significantly suppressed CRC cell proliferation ( Figure 5 K) and enhanced cell sensitivity to ferroptosis inducers ( Figure 5 L). Notably, OA-mediated inhibition of CRC cell growth was significantly attenuated in TRIM24-deficient cells ( Figure 5 M), indicating that TRIM24 is a crucial mediator of OA’s anti-cancer effects. Collectively, these findings raise the possibility that OA induces ferroptosis and increases ferroptosis sensitivity in CRC cells, possibly mediated by TRIM24 degradation. The TRIM family represents a classical group of E3 ubiquitin ligases, 22 we formulated a hypothesis to explore whether OA could inhibit tumor growth by diminishing the presentation of the tumor suppressor protein TRIM24. Analyzing the proteomic data ( Figure 4 B), we observed a noteworthy increase in TSPO subsequent to OA intervention. Previous studies have indicated that TSPO plays a crucial role in regulating mitochondrial oxidative phosphorylation and glycolytic balance, and its deficiency has been associated with promoting glioma growth and angiogenesis. 23 The results from western blotting revealed that OA treatment not only inhibited TRIM24 but also significantly increased TSPO expression ( Figures 6 A and 6B). Additionally, TSPO expression was also elevated in TRIM24 knockdown colorectal cancer cells ( Figure 6 C), supporting the possibility that the increase in TSPO expression induced by OA treatment is regulated through TRIM24 levels. Furthermore, OA intervention suppressed the expression of vascular endothelial growth factor A (VEGFA), matrix metalloproteinase 2 (MMP2), and hypoxia-inducible factor HIF-1α, which may be downstream targets of TSPO ( Figures 6 D and 6E). To establish TSPO as a key protein in OA’s anti-cancer effect, siRNA was employed to rescue OA-treated cells. Results showed a partial restoration of VEGFA, MMP2, and HIF-1α expression upon TSPO siRNA intervention ( Figures 6 F and 6G). Importantly, TSPO siRNA partially restored OA’s inhibitory effect on CRC cells ( Figures 6 H–6J). To further demonstrate TRIM24’s crucial role as an E3 ubiquitin ligase, co-immunoprecipitation experiments revealed a verified interaction between TRIM24 and TSPO ( Figure 6 K). Additionally, the co-immunoprecipitation data indicated that OA treatment reduced the binding of TSPO to ubiquitin ( Figure 6 L), suggesting that OA decreases the ubiquitination level of TSPO. In conclusion, we propose that OA may inhibit CRC growth by suppressing TRIM24, consequently elevating TSPO levels. Figure 6 OA regulates the TSPO in CRC cells (A and B) Western blotting assay of TPSO with OA treatment in HCT-8 and SW-480. (C) TSPO expression levels in HCT-8 and SW-480 cells after TRIM24 knockdown. (D and E) Western blotting assay of MMP2, HIT-α, VEGFA with OA treatment in HCT-8 and SW-480. (F and G) Western blotting assay of MMP2, HIT-α, and VEGFA after OA treatment combined with TPSO knockdown in HCT-8 and SW-480. (H–J) Colony-formation (H), wound-healing (I), migration and invasion assays (J) after OA treatment combined with TPSO knockdown in HCT-8 and SW-480. Scale bar, 50 μm. (K) Lysates from HCT-8 cells were immunoprecipitated with control IgG or an anti-TRIM24 antibody. (L) After 24-h OA treatment, HCT-8 cells were immunoprecipitated with TSPO antibody, and ubiquitination of TSPO was detected using an anti-ubiquitin antibody. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. OA regulates the TSPO in CRC cells (A and B) Western blotting assay of TPSO with OA treatment in HCT-8 and SW-480. (C) TSPO expression levels in HCT-8 and SW-480 cells after TRIM24 knockdown. (D and E) Western blotting assay of MMP2, HIT-α, VEGFA with OA treatment in HCT-8 and SW-480. (F and G) Western blotting assay of MMP2, HIT-α, and VEGFA after OA treatment combined with TPSO knockdown in HCT-8 and SW-480. (H–J) Colony-formation (H), wound-healing (I), migration and invasion assays (J) after OA treatment combined with TPSO knockdown in HCT-8 and SW-480. Scale bar, 50 μm. (K) Lysates from HCT-8 cells were immunoprecipitated with control IgG or an anti-TRIM24 antibody. (L) After 24-h OA treatment, HCT-8 cells were immunoprecipitated with TSPO antibody, and ubiquitination of TSPO was detected using an anti-ubiquitin antibody. Data shown is mean (±SEM, n = 3). “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. The efficacy of OA treatment was evaluated in a subcutaneous CRC mouse model. The impact of OA on CRC growth was assessed, as depicted in Figures 7 A–7C and Tables S1 and S2 . OA demonstrated inhibitory effects on HCT-8 subcutaneous tumor growth by 36.03% (5.0 mg/kg/d) and 61.83% (10.0 mg/kg/d), respectively, whereas the inhibitory effect of 5-fluorouracil (5-FU) was 57.57% (5 mg/kg/d). Furthermore, treatment with OA at doses of 5.0 and 10.0 mg/kg/d had no significant impact on average body weight, alanine transaminase (ALT), aspartate transaminase (AST), total bilirubin (TBIL) levels in peripheral blood, and morphology of multiple organs ( Figures 7 D–7F). Conversely, treatment with 5-FU significantly reduced mouse body weight ( Figure 7 D). These results suggest that, within an effective anti-cancer range, OA has a better capacity to reduce drug toxicity to the host compared to traditional chemotherapy drugs. Moreover, OA treatment significantly reduced the expression of TRIM24 and SLC3A2 while elevating TSPO expression in subcutaneous tumor tissues as observed through immunohistochemistry ( Figure 7 G). Figure 7 OA treatment inhibits HCT-8 subcutaneous tumor growth without causing significant harm to the host (A) The subcutaneous tumor volumes were measured once a week. (B and C) At day 35, the tumor picture and weight of mice were recorded and measured. (D–F) The mice were observed for alterations in body weight, blood biochemistry (the levels of ALT, AST, and TBIL), and morphology of multiple organs (H&E staining) as an indicator of potential harm. (G) Representative immunohistochemical images of TRIM24, SLC3A2, and TSPO expression after OA treatment. (H) Tumor volumes in mice of CRC PDX model. (I and J) Tumors from the CRC PDX model were photographed and weighed. (K) Western blot analysis of TRIM24 expression in tumor tissues from the control and OA-treated groups in the CRC PDX model. Data shown is mean (±SEM, n = 6) “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. OA treatment inhibits HCT-8 subcutaneous tumor growth without causing significant harm to the host (A) The subcutaneous tumor volumes were measured once a week. (B and C) At day 35, the tumor picture and weight of mice were recorded and measured. (D–F) The mice were observed for alterations in body weight, blood biochemistry (the levels of ALT, AST, and TBIL), and morphology of multiple organs (H&E staining) as an indicator of potential harm. (G) Representative immunohistochemical images of TRIM24, SLC3A2, and TSPO expression after OA treatment. (H) Tumor volumes in mice of CRC PDX model. (I and J) Tumors from the CRC PDX model were photographed and weighed. (K) Western blot analysis of TRIM24 expression in tumor tissues from the control and OA-treated groups in the CRC PDX model. Data shown is mean (±SEM, n = 6) “ns” stands for “no significant difference”, ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. To further validate the antitumor efficacy of OA in colorectal cancer, a patient-derived xenograft (PDX) model was employed. The results demonstrated that OA treatment significantly inhibited tumor growth in the CRC PDX model ( Figure 7 H; Tables S3 and S4 ), with a tumor mass inhibition rate of 43.40 ± 10.60% ( Figures 7 I and 7J). Subsequent western blot analysis further confirmed that OA markedly downregulated the expression of TRIM24 ( Figure 7 K). These findings collectively indicate that OA can effectively inhibit colon tumor growth through pathways involving TRIM24.

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Guodong Li ( [email protected] ). This study did not generate new unique reagents. • Data: this paper does not report new dataset. • Code: this paper does not report original code. • All other items: any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request. Data: this paper does not report new dataset. Code: this paper does not report original code. All other items: any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

CRC, characterized by a predominant malignancy and poor prognosis, necessitates the identification of new therapeutic targets. The ubiquitin-proteasome system plays a crucial role in the occurrence and development of cancer cells. 24 , 25 Within this intricate system, E3 ubiquitin ligases emerge as a vital component, exerting key regulatory functions. 26 Mdm2 (Mouse double minute 2), as a prototypical E3 ubiquitin ligase, exerts its inhibitory effect on p53 function through ubiquitination, ultimately leading to p53 degradation. 27 Similarly, Skp2 (S-phase kinase-associated protein 2), acting as another E3 ubiquitin ligase, contributes to the uncontrolled proliferation of tumor cells by promoting the degradation of the cell cycle inhibitor CDK inhibitor p27. 28 Specific key E3 ubiquitin ligases may, therefore, constitute a critical strategy for CRC therapy. TRIM24, functioning as an E3 ubiquitin ligase, exhibits a significant correlation with CRC cell proliferation. 29 Our study showed that TRIM24 expression was found to be significantly upregulated in tumor tissues compared to adjacent normal tissues, based on data from public databases and CRC clinical samples from The Affiliated Hospital of Hangzhou Normal University. Elevated TRIM24 expression was found to be associated with a poorer prognosis, and a negative correlation was identified between TRIM24 and the infiltration of tumor-killing cells in tumor tissues, while a positive correlation was established between TRIM24 and immune inhibitory cells. Consequently, the expression of TRIM24 is intricately linked to the development and prognosis of CRC. Recently, four effective and specific TRIM24 inhibitors have been developed with significant tumor inhibition, including compound 34, 30 IACS-6558, 31 IACS-9571, 32 and Y08624 . 33 Studies have shown that Y08624 has the ability to inhibit the growth of prostate cancer. 33 Han et al. demonstrated that TRIM24 inhibitors (IACS-6558, IACS-9571) are potent in preventing cell growth and invasion in various patient-derived glioblastoma stem cells. 34 Building upon these findings, we conducted a target screening of the TRIM24 protein. Virtual docking and cellular thermal shift assays (CETSA) revealed a potential binding interaction between OA and TRIM24. Furthermore, OA markedly reduced TRIM24 protein expression and exhibited potent anti-cancer effects in colorectal cancer (CRC) cells. These results suggest that OA may serve as a promising therapeutic candidate for CRC. Previous studies have provided evidence that TRIM24 can target and degrade p53, and it also interacts with NLRP3, contributing to the progression of endometriosis. 35 , 36 To gain deeper insights into the potential mechanisms underlying the inhibitory effects of OA on CRC, our drug intervention revealed its involvement in the ferroptosis pathway—a process crucial for maintaining cellular balance, differentiation, and the regulation of growth and proliferation. 37 This pathway plays diverse roles at different stages of tumor development and is jointly regulated by various mechanisms, including genetic and transcriptional control. Our findings revealed that OA treatment significantly altered the expression of ferroptosis-associated proteins and genes, including SLC3A2, SLC7A11, GPX4, and ACSL4. Additionally, OA decreased the GSH/GSSG ratio and increased ROS levels, thereby promoting ferroptosis in CRC cells. The ferroptosis inhibitor ferrostatin-1 was able to partially rescue the OA-induced suppression of CRC cell proliferation, further supporting ferroptosis involvement. Moreover, TRIM24 knockdown attenuated the sensitivity of CRC cells to both ferroptosis inducers and OA treatment, indicating that TRIM24 plays a crucial role in mediating OA-induced ferroptosis. Collectively, these results suggest that OA promotes ferroptosis in CRC cells by targeting key regulators of the ferroptosis pathway, with TRIM24 acting as a significant mediator. Additionally, we observed an increase in TSPO expression upon OA treatment. TSPO exhibits contrasting roles in different tumors, and recent research indicates that TSPO deficiency inhibits mitochondrial oxidative phosphorylation, promoting glucose absorption, lactate conversion, and facilitating the growth and angiogenesis of gliomas. 23 In our study, the OA-mediated degradation of TRIM24 resulted in elevated TSPO expression, subsequently suppressing VEGF-A, MMP2, and HIF-1α, thereby inhibiting the malignant phenotypes of CRC cells, including proliferation, migration, and invasion. Therefore, targeting TRIM24 expression with OA represents a promising therapeutic strategy for the treatment of colorectal cancer. In this study, we elucidated the characteristics of TRIM24 in CRC and identified a previously unidentified inhibitor, OA. OA demonstrated significant antitumor efficacy by regulating TRIM24, without inducing significant host toxicity. This effect can be attributed to the induction of ferroptosis and the modulation of TSPO expression. In summary, our findings underscore TRIM24 as a potential biomarker for CRC, and we propose OA as a promising therapeutic option for CRC. Despite the promising findings, this study has several limitations that warrant consideration. First, the number of clinical CRC tissue samples used for validation was limited ( n = 5), which may restrict the generalizability and statistical robustness of the conclusions drawn regarding TRIM24 expression in patient tissues. Future studies should involve a larger, more diverse patient cohort to confirm these findings and strengthen clinical relevance. Second, although we demonstrated that OA can induce ferroptosis and inhibit tumor growth through the downregulation of TRIM24, the precise molecular mechanisms by which TRIM24 regulates ferroptosis-related proteins (e.g., SLC3A2, SLC7A11, GPX4, and ACSL4) remain unclear. It is not yet fully established whether TRIM24 directly interacts with or transcriptionally regulates these ferroptosis-associated targets. Third, while TSPO was identified as a potential downstream effector of TRIM24, the regulatory axis connecting TRIM24, TSPO, and ferroptosis requires more comprehensive mechanistic validation, including chromatin immunoprecipitation (ChIP), transcriptional activity assays, and identification of intermediate signaling components. Finally, the specific mechanism by which OA suppresses colorectal tumor growth through the regulation of TRIM24 remains unclear. It is yet to be determined whether OA directly binds to TRIM24, as suggested by molecular docking simulations and CETSA experiments, as well as the exact binding domains or sites involved. Moreover, the manner in which OA regulates the degradation or downregulation of TRIM24 expression is still unknown. Further investigations will be carried out based on the findings of this study.

The incidence and mortality rates of colorectal cancer (CRC) stand among the most elevated globally. 1 The escalating prevalence of CRC can be ascribed to alterations in dietary patterns, heightened consumption of red meat, and diminished levels of physical activity concomitant with the amelioration in global economic status. 2 However, deficiencies in early diagnosis and efficacious treatment profoundly influence the mortality rates associated with CRC. In the pursuit of heightened precision and personalized therapeutic approaches for CRC, the introduction of the colorectal cancer consensus molecular subtypes (CMS) classification has become imperative. 3 This classification system delineates CRC into four discrete subtypes: CMS1 (MSI immune), CMS2 (canonical), CMS3 (metabolic), and CMS4 (mesenchymal). Nevertheless, the exact therapeutic implications predicated on CMS classification remain uncertain. 4 Hence, the identification of new targets and pharmaceutical agents is imperative to address the prevailing challenges in enhancing the efficacy of CRC treatment. The tripartite motif (TRIM) family has been delineated as a category of E3 ubiquitin ligases, characterized predominantly by the presence of a RING domain capable of binding to the ubiquitin-loaded E2 enzyme. This enables the facilitated transfer of ubiquitin to target proteins. 5 Members within the TRIM family play pivotal roles in the initiation, progression, and therapeutic resistance across diverse cancer types. They manifest both oncogenic and tumor-suppressive functions, exhibiting context-dependent actions in various human cancer types. 6 As a constituent of the TRIM family, TRIM24 assumes a consequential role in the regulatory mechanisms governing tumorigenesis. 7 , 8 Numerous studies substantiate that TRIM24 possesses the capability to ubiquitinate and facilitate the proteasome-mediated degradation of p53 and TERX1. This cascade ultimately contributes to DNA damage repair mechanisms, thereby fostering tumor growth. 9 , 10 Furthermore, TRIM24 exhibits heightened expression in CRC tissues, thereby contributing to the remodeling of the tumor stroma through the stimulation of angiogenesis and recruitment of tumor-associated macrophages (TAMs). 11 Nevertheless, the viability of targeting TRIM24 as a therapeutic approach for CRC remains uncertain and warrants further investigation. In recent years, there has been a notable surge in the exploration of small molecule inhibitors and natural products within contemporary medical research, positioning them as prospective therapeutic agents. 12 , 13 Small molecule inhibitors and natural products demonstrate the capacity to selectively target specific proteins or biomolecules, thereby modulating physiological and pathological processes implicated in diseases, such as cancer, inflammation, and metabolic disorders. 14 , 15 , 16 The flavonoid compound Oroxin A(OA), derived from the medicinal herb Oroxylum indicum (L.) Kurz, has garnered prominence as a noteworthy example. Traditionally employed for the treatment of conditions such as cough, pharyngitis, pertussis, bronchitis, and others, oroxylum indicum (L.) Kurz holds therapeutic significance. 17 Specifically, the chemical constituents, encompassing flavonoid glycosides, cyclohexanol, and OA, have been acknowledged for their discernible anti-inflammatory and anti-cancer properties. 18 , 19 , 20 Nevertheless, the precise therapeutic potential of OA in the context of CRC, along with its underlying mechanistic intricacies, remains yet to be fully elucidated and warrants further investigation. In this study, we meticulously scrutinized TRIM24 expression and prognostic implications in CRC and identified OA as a small molecule inhibitor of TRIM24. We systematically elucidated the anti-CRC activity and underlying mechanisms of OA, both in vitro and in vivo . Our findings robustly suggest that OA exhibits promising potential as a therapeutic agent against CRC.

The authors declare no competing interests.

REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies TRIM24 proteintech RRID: AB_2619207 SLC3A2 proteintech RRID: AB_2254909 SLC7A11 Thermo Fisher Scientific RRID: AB_3093725 GPX4 Cell Signaling Technology RRID: AB_2924984 ACSL4 Thermo Fisher Scientific RRID: AB_3247784 TSPO Abcam RRID: AB_10897701 HIF-1α proteintech RRID: AB_10732601 VEGF-A proteintech RRID: AB_2212657 MMP2 proteintech RRID: AB_2250823 ubiquitin proteintech RRID: AB_671515 β-actin proteintech RRID: AB_2923704 Biological samples Human CRC tissues The Affiliated Hospital of Hangzhou Normal University This paper Chemicals, peptides, and recombinant proteins Erastin TargetMol 571203-78-6 OA TargetMol 31567-75-6 Fer-1 MedChemExpress HY-100579 10X TNC buffer ECOTOP ED-9270 M-PER buffer Thermo Fisher Scientific 78501 1640 Gibco 12633020 DMEM Gibco C11965500BT PBS Gibco 70011044 Triton X-100 Sigma-aldrich Cat# SLBT3016 TWEEN J.T.Baker X251-07 CAS 9005-64-5 BSA Sigma-Aldrich A7906 CAS 9048-46-8 Critical commercial assays Cell Counting Kit-8 HY-K0301 GSH and GSSG Assay Kit Beyotime S0053 Reactive Oxygen Species Assay Kit Beyotime S0033 Glutathione Peroxidase (GPX) activity Assay Kit Beyotime S0056 BODIPY 581/591 C11 Beyotime S0043 BCA assay Thermo Fisher Scientific A53225 RNA Extraction Kit Yishan Biotechnology RN001-50Rxns SSuper SYBR Green qPCR Master Mix Yishan Biotechnology ES-QP002 Co-IP kit Epizyme YJ201 Experimental models: Cell lines CCD-841 Chinese Academy of Sciences N/A SW-480 Chinese Academy of Sciences N/A SW-620 Chinese Academy of Sciences N/A DLD-1 Chinese Academy of Sciences N/A HCT-8 Chinese Academy of Sciences N/A Experimental models: Organisms/strains Nude mice Shanghai Slack Laboratory Animal Co. Ltd N/A Oligonucleotides Slc3a2 (forward: 5’-TGAATGAGTTAGAGCCCGAGA-3’, reverse: 5’-GTCTTCCGCCACCTTGATCTT-3’), This paper N/A Slc7a11 (forward: 5’-TTACTACTTCTGGATTGGCTA-3’, reverse: 5’-CTTGTATTTAAGCGCCTGCC-3’), This paper N/A GPX4 (forward: 5’-ATGAGCCTCGGCCGCCTTTG-3’, reverse: 5’-CCCACAAGGTAGCCAGGGGT-3’), This paper N/A ACSL4 (forward: 5’-GCAGAGTACCCTGAAGGATTTG-3’, reverse: 5’-CGTTGGTCTACTTGGAGGAATG-3’). This paper N/A Software and algorithms GraphPad Prism version 10.4.0 N/A GraphPad Software, CA, USA FlowJo version 10.8.1 N/A FlowJo, LLC SPSS version 25 N/A IBM,USA AutoDockTools 1.5.7 N/A USA PyMOL 2.3.3 N/A USA MaxQuant 1.6.17.0 N/A Germany The study involved 5 male human participants, aged between 18 and 80 years. All procedures performed in studies involving human participants were in accordance with the ethical standards of the Affiliated Hospital of Hangzhou Normal University (Approval No.2025(E2)-HS-029. The present study was approved by the Laboratory animal management and ethics committee of Zhejiang Chinese Medical University (Approval No. IACUC-2022011-02). All the procedures for the care of the rats were in accordance with the institutional guidelines for animal use in research. 4-6-week male nude mice were purchased from Shanghai Slack Laboratory Animal Co. Ltd. The human CRC cell linesHCT-8, SW-480, SW-620, DLD-1 and normal cells CCD-841 were obtained from the Chinese Academy of Sciences. These cell lines were cultured in RPMI-1640 medium or Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 g/ml streptomycin at 37°C. The cell lines used in this study were authenticated by Short Tandem Repeat (STR) profiling to confirm their identity and genetic consistency. Additionally, all cell lines were tested for mycoplasma contamination using a PCR-based detection method, which ensures the purity of the cultures and the reliability of the experimental results. The docking of TRIM24 was systematically executed utilizing the AutodockTools software in accordance with established protocols. 38 The structural information of OA (PubChem CID: 5320313) was retrieved from the PubChem Compound Database, whereas the crystal structure of TRIM24 (PDB code: 4YAB ) was acquired from the Research Collaboratory for Structural Bioinformatics Protein Data Bank. Subsequent to data retrieval, a visual examination of the optimal complex was conducted utilizing AutodockTools 1.5.7 and PyMOL 2.3.3 for comprehensive analysis. Tumor tissues and matched adjacent normal tissues were obtained from five colorectal cancer patients who underwent surgical resection at The Affiliated Hospital of Hangzhou Normal University (approval number 2025(E2)-HS-029). All patients provided informed consent, and the study was approved by the institutional ethics committee. Tissues were immediately snap-frozen in liquid nitrogen and stored at -80°C until use. CRC cells seeded in a 6 cm dish were treated with 2.0 μM OA for 24 hours, after which the cells were collected. Knockdown cells were collected during the exponential growth phase. Total protein extraction from cells subjected to various treatments was carried out. The proteins were subsequently separated through sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred onto polyvinylidene fluoride (PVDF) membranes, and then blocked with 5% skim milk. Following this, membranes were probed overnight with primary antibodies, succeeded by secondary antibodies (ProteinTech). Chemiluminescence signals were detected utilizing enhanced chemiluminescence detection reagents (ECL; Biosharp). Quantification and normalization of the data were performed utilizing β-actin as a loading control. For cellular thermal shift assay, HCT-8 and SW-480 CRC cells were subjected to a 1-hour treatment with 20 μM OA. Subsequently, the cells were harvested in pre-cooled phosphate-buffered saline (PBS) supplemented with 1% protease inhibitor in 1.5 mL tubes. These tubes were then sequentially incubated at specified temperatures, including room temperature and -80°C, for 3 minutes each. The ultimate results were discerned through western blotting analysis. For the drug affinity reactive target stability assay, HCT-8 and SW-480 CRC cells were lysed in M-PER buffer (Thermo Fisher Scientific, cat#78501) with protease and phosphatase inhibitors. The lysates were mixed with 10X TNC buffer (ECOTOP, cat#ED-9270), and protein concentration was determined using the BCA assay. The lysates were incubated with Oroxin A or DMSO for 1 hour at room temperature, followed by pronase digestion (Roche, cat#10165921001) for 30 minutes at room temperature. A cell seeding density of 2000-3000 cells per well was employed in a 96-well plate, followed by co-cultivation with varying concentrations of OA (TargetMol, cat# 31567-75-6) (1.0, 2.0, 4.0, 6.0, 8.0, 10.0, 15.0, 20.0 μM) or dimethyl sulfoxide (DMSO) for a duration of 48 hours. Subsequently, 10% CCK-8 solution was introduced into each well and incubated for 1 hour at 37 °C. The optical density (OD) value was quantified at 450 nm. For the TRIM24 knockdown experiment, 3000 CRC cells were seeded in a 96-well plate. CCK-8 solution (10%) was added at 0h, 24h, and 48h, and the OD value at 450 nm was measured 2 hours post-incubation. In the experiment assessing TRIM24's effect on ferroptosis induction by erastin (TargetMol, cat#571203-78-6), 3000 CRC cells were seeded into a 96-well plate. The cells were co-cultivated with varying concentrations of erastin (1.0, 2.0, 4.0, 8.0, 10.0, 20.0 μM) or DMSO for 48 hours, followed by the addition of CCK-8 solution and OD measurement at 450 nm. To evaluate whether the ferroptosis inhibitor Fer-1 could rescue OA-induced cell growth inhibition, 2,000–3,000 cells per well were treated with 2 μM OA alone or in combination with 100 nM Fer-1 for 48 hours. After treatment, CCK-8 was added and incubated for 2 hours prior to OD450 measurement. To assess OA's inhibitory effect in TRIM24 knockdown cells, both control and TRIM24-silenced CRC cells (2,000–3,000 cells per well) were treated with 2 μM OA for 48 hours. Cell viability was evaluated as above, and the inhibition rate was calculated using the formula: Inhibition rate (%) = [1 − (OD value of knockdown group / average OD value of control group)] × 100%. A total of 1000 CRC cells from the HCT-8 and SW-480 cell lines were seeded overnight in a 3.5cm dish and subsequently treated with OA at concentrations of 2.0 or 5.0 μM, or dimethyl sulfoxide (DMSO) for a duration of 24 hours. Following this treatment, the cells were cultured for an additional 14 days in normal medium. Additionally, these cells were seeded into the upper transwell chambers (Corning) containing 2.0 μM OA or DMSO in the absence of fetal bovine serum (FBS), while the lower chamber received medium supplemented with 10% FBS. After 48 hours, the cells remaining in the upper chamber were removed using sterile cotton swabs. The dishes and transwell chambers were subsequently subjected to staining with crystal violet (Solarbio). A total of 8 ∗ 10ˆ4 CRC cells from the HCT-8 and SW-480 cell lines were seeded into culture inserts (Ibidi GmbH). Subsequent to the removal of the Culture-Inserts, cells underwent treatment with 2.0 μM OA or dimethyl sulfoxide (DMSO). Wound healing images were captured at 0, 24, and 48 hours using an optical microscope. Cells were exposed to 2.0 μM OA or an equivalent volume of solvent for 24 hours, followed by collection for the construction of a proteome library and subsequent sequencing at GeneChem Biotechnology Company. The procedures encompassed protein extraction, quantification, SDS-PAGE, enzymatic hydrolysis of proteins, and mass spectrometry analysis, facilitated by MaxQuant software (version 1.6.17.0). To ensure stringent criteria for peptide and protein identification, a false discovery rate cut-off of 0.01 was applied. Differentially expressed proteins were defined by a fold change > 1.2 and a p-value (Student's t-test) < 0.05. Each experimental group was replicated three times to facilitate a comprehensive assessment of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment. CRC cells, specifically HCT-8 and SW-480, were subjected to a 24-hour treatment with 2 μM OA or dimethyl sulfoxide (DMSO). Subsequently, the cells underwent processing utilizing the Reactive Oxygen Species Assay Kit and Lipid Peroxidation Assay Kit. Briefly, cells were incubated with 5 μM DCFH-DA or 2 μM BODIPY 581/591 C11 for 20 minutes at 37°C. The analysis was then conducted and recorded using flow cytometry (Agilent) or fluorescence microscopy, and the data were quantified using FlowJo (v10.8.1). CRC cells, specifically HCT-8 and SW-480, were treated with 2 μM OA or dimethyl sulfoxide (DMSO) for a duration of 24 hours. Subsequently, the assessment of Glutathione Peroxidase (GPX) activity and the determination of the ratio of oxidized glutathione (GSSG) to reduced glutathione (GSH) were conducted using respective kits. The final optical densities at 412 nm and 340 nm were measured employing a Thermo Varioskan Flash microplate reader. Total RNA was extracted from the samples using an RNA Extraction Kit according to the manufacturer's instructions. Reverse transcription was performed using 1 μg of total RNA to synthesize cDNA. Quantitative PCR was carried out with SSuper SYBR Green qPCR Master Mix following the manufacturer's protocol. All reactions were conducted in triplicate, and relative gene expression levels were calculated using the 2ˆ−ΔΔCt method with GAPDH (or β-actin) as the internal reference gene. The TSPO small interfering RNA (siRNA; siTSPO) was procured from Jiman Biotechnology. Subsequently, TSPO siRNA at a concentration of 100 nM was transfected into HCT-8 and SW-480 CRC cells following a 24-hour incubation period. The transfection was facilitated using Lipofectamine 2000 (Invitrogen, 11668–019), and the cells were incubated for an additional 44 hours in fresh medium. Lentiviruses harboring an shRNA against TRIM24, a TRIM24 overexpression plasmid, or a scramble shRNA from Shanghai GeneChem Co Ltd, China. ShRNAs against TRIM24 were CCGATCCCAAGCTCATCATTT and GCGCCTCCTTAAAGTTGCCAT. Infections were carried out as directed by the manufacturer for 72 h, and transfection efficacy was determined by western blot. Co-immunoprecipitation (Co-IP) assays were executed utilizing a Co-IP kit (Epizyme, Cat No. YJ201). HCT-8 cells were treated with 2 μM OA or not for 24 hours. Cells were harvested and lysed, and in accordance with the manufacturer's instructions, proteins underwent successive incubation with beads and antibodies (anti-TRIM24 or anti-TSPO antibody) overnight at 4°C. Subsequently, the resultant protein-antibody-bead complexes were subjected to boiling with loading buffer and subsequently analyzed through western blotting. Animals were sourced from the experimental animal department of Zhejiang Chinese Medical University, specifically BALB/c nude and NOD/SCID mice (male, 4 weeks old). In strict adherence to the regulations and guidelines of Zhejiang Chinese Medical University institutional animal care (Approval No. IACUC-2022011-02), all animal experiments, encompassing euthanasia, were conducted in accordance with AAALAC and IACUC standards. Subcutaneous xenograft tumor models were established in nude mice by subcutaneously injecting 5×10ˆ6 HCT-8 CRC cells. Fresh colorectal tumor tissues obtained from surgical resection of patients were implanted subcutaneously into NOD/SCID mice to establish primary PDX (patient-derived xenograft) models. Once tumors reached an appropriate size, tumor fragments were harvested and passaged subcutaneously into BALB/c nude mice for further expansion and treatment studies. Tumor dimensions were measured weekly using a Vernier caliper, and tumor volumes were calculated employing the formula 1/2a×bˆ2, where ‘a' and ‘b' represent the larger and smaller dimensions of the tumor, respectively. Tumor-bearing mice were randomly assigned to vehicle and treatment groups once the average tumor volume reached 100 mmˆ3. OA at doses of 5.0 mg/kg or 10.0 mg/kg, or 5-Fluorouracil (5-FU) at a dose of 5.0 mg/kg, was intraperitoneally injected once daily for 3 weeks. The control group received an equivalent volume of the vehicle. Tumor volumes and weights were measured twice a week. Tumor tissues were collected and weighed, peripheral blood was obtained for enzyme-linked immunosorbent assay (ELISA) to assess blood biochemical markers, and tumor and various organs underwent histological examinations, including hematoxylin and eosin (H&E) and immunohistochemical (IHC) staining. Data were analyzed using SPSS version 25. Data were analyzed using the Student’s t-test for comparisons between two groups. For comparisons involving three or more groups, one-way analysis of variance (ANOVA) was performed. A p-value less than 0.05 was considered statistically significant. Results are expressed as follows: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ns denotes no significant difference. All other items: any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

The authors thank Jianliang Wu and Liwei Sun for assisting in the preparation of this manuscript. This work was supported by Hangzhou Agricultural and Social Development Research Guidance Project ( 20241029Y026 ).