Artemisitene triggers calcium-dependent ferroptosis by disrupting the LSH-EWSR1 interaction in colorectal cancer

Abstract

Colorectal cancer (CRC), propelled by extreme molecular heterogeneity and intractable drug resistance, is rapidly becoming a global health challenge. Ferroptosis offers a promising therapeutic strategy by exploiting the iron addiction and oxidative vulnerability of CRC cells. However, available methods to trigger ferroptosis are still limited, mostly focusing on antioxidant systems or iron metabolism. Here, we found that artemisitene (ATT), a bioactive natural sesquiterpene isolated from Artemisia annua, acted as a CRC therapeutic agent by promoting calcium-dependent ferroptosis. Integrative transcriptomics revealed that ATT repressed cytochrome P450 family 24 subfamily A member 1 (CYP24A1) expression, the pivotal mediator of this response. The ensuing calcium overload downregulated stearoyl-CoA desaturase (SCD) by CAMKK2/AMPK/SREBF1 axis, enriching oxidizable fatty acids and sensitizing CRC cells to lethal lipid peroxidation. Mechanistically, ATT was found to directly target lymphoid-specific helicase (LSH), covalently binding to the Cys205 residue of LSH and thereby disrupting its interaction with EWS RNA binding protein 1 (EWSR1). This disruption ultimately suppressed CYP24A1 transcription. Our findings revealed that pharmacological blockade of the LSH/CYP24A1/SCD axis triggers calcium-driven ferroptosis, positioning ATT as a potent, mechanism-based therapeutic for CRC.

Graphical abstract

Highlights

• •ATT induced calcium-driven fatty acid accumulation to trigger ferroptosis in CRC.
• •ATT directly bound to LSH, leading to inhibition of the LSH/CYP24A1/SCD axis.
• •Cys205 alkylation of LSH by ATT dismantled LSH-EWSR1 complex to silence CYP24A1.

1. Introduction

Colorectal cancer (CRC) is a major global health issue, making up nearly 10 % of all cancer diagnoses and deaths [1,2]. For CRC treatment, early-stage cases often undergo surgical resection, while advanced cases rely on chemotherapy and radiotherapy to slow progression and improve survival. However, recurrence and metastasis [3] remain major challenges in oncology. Amid these challenges, ferroptosis has emerged as a promising therapeutic strategy, with colorectal cancer (CRC), owing to its intrinsic iron reliance and redox fragility [4], appearing particularly susceptible to this form of cell death. Ferroptosis is a form of regulated cell death driven by lipid peroxidation [5], which bypasses apoptosis resistance and selectively exploits tumor-specific metabolic vulnerabilities [6]. Conventional ferroptosis inducers largely rely on targeting iron, lipids, and thiols [7,8]. However, metabolic disparities and pronounced heterogeneity in tumors often confer varying degrees of resistance to classical ferroptosis-inducing pathways [9], highlighting the imperative to develop alternative modalities for ferroptosis induction.

Calcium, a pivotal second messenger whose absorption largely occurs in the intestine, is often disrupted in CRC due to abnormal channel and pump function [10,11], thereby heightening cellular vulnerability to calcium perturbations. Accumulating evidence further implicates calcium as a multifaceted regulator of ferroptosis [12] by integrating ionic homeostasis with redox sensitivity [[13], [14], [15]]. It was reported that the enhanced calcium transfer between the endoplasmic reticulum and mitochondria is a prerequisite for ferroptosis induction [16,17], which strengthens intracellular oxidative stress and modulates the susceptibility of cancer cells to ferroptosis. Moreover, endoplasmic reticulum calcium depletion triggered the reprogramming of tumor cell membrane composition, characterized by an elevated proportion of saturated lipids that inherently suppresses ferroptosis [18], demonstrating that calcium fluxes can be bidirectionally tuned to either drive or restrain this cell-death pathway. This positions calcium signaling and the proteins that govern it as a versatile therapeutic lever. Within this network, CYP24A1, the cytochrome P450 enzyme that catalyzes the inactivation of active vitamin D (1, 25(OH)₂D₃), has emerged as a potential modulator of calcium signaling [19]. By reducing 1, 25(OH)₂D₃ levels, aberrant upregulation of CYP24A1 may impair calcium absorption and intracellular calcium availability, thereby disrupting calcium homeostasis [[20], [21], [22]]. Although calcium dysregulation has been implicated in CRC progression [[23], [24], [25]], the exact role of CYP24A1 in driving calcium-dependent ferroptosis remains undefined, underscoring its promise as a therapeutic node for ferroptosis-based therapy. Previous studies have proposed direct inhibition of CYP24A1 as a potential therapeutic avenue [26]. However, this approach is constrained by adverse effects such as calcium-phosphate disequilibrium [27] and compensatory metabolic feedback [28]. Recently, epigenetic modulation via chromatin remodeling and transcriptional control has emerged as a promising alternative [29], offering a means to overcome the inherent limitations of enzyme-centric interventions.

Lymphoid-specific helicase (LSH), an SNF2-family chromatin-remodeling ATPase, orchestrates epigenetic processes including nucleosome remodeling [30], DNA methylation, histone modification, and DNA repair [31,32]. In cancer, aberrantly overexpressed LSH drives tumor growth [33] and metastasis [34] across multiple malignancies. Accumulating evidence demonstrates that LSH exerts epigenetic control over ferroptosis-related genes by remodeling chromatin states at their promoter regions [35], such as CYP24A1. By regulating promoter accessibility and transcriptional dynamics, LSH contributes to the metabolic and redox pathways that determine ferroptotic susceptibility in cancer cells [30,36]. Beyond chromatin regulation, LSH activates p53 [37] and its downstream lipid metabolic programs [38] through post-translational modifications, linking it directly to ferroptosis [36,39] and underscoring its promise as a biomarker and therapeutic target.

Traditional Chinese medicine (TCM) offers a valuable reservoir of bioactive compounds with potential applications in cancer diseases. Artemisia annua, best known for artemisinin (qinghaosu), has demonstrated broad pharmacological effects in systemic disorders. In recent years, artemisinin derivatives have emerged as promising candidates in oncology [40], largely owing to their ability to perturb redox homeostasis and iron metabolism [41]. Generally, the endoperoxide bridge is recognized as the pharmacophoric group responsible for this activity [42], as it can react with heme to release ferrous ions and free radicals, thereby driving oxidative stress. However, this radical-driven reactivity often results in non-selective protein alkylation, and the precise contribution of the overall artemisinin scaffold to its antitumor activity, as well as its definitive biological targets, remains unclear. Artemisitene (ATT), a sesquiterpene lactone endoperoxide derived from Artemisia annua [43], is a natural artemisinin analogue reported to exhibit stronger anti-cancer [44,45] and anti-inflammation [[46], [47], [48]] activities than artemisinin and its derivatives. Although its pro-apoptotic activity in breast cancer has been linked to farnesyl-diphosphate farnesyltransferase 1 (FDFT1) [45], the underlying mechanism and the binding mode remain undefined. Therefore, clarifying ATT's direct targets and the downstream networks they command will be pivotal for engineering next-generation, artemisinin-inspired therapeutics with heightened potency.

In this study, we integrated transcriptomic and lipidomic analyses and found that ATT triggers a calcium surge, disrupting lipid homeostasis and driving colorectal cancer cells into ferroptosis. At the mechanistic level, the electrophilic core of ATT covalently modified LSH, disrupting its interaction with EWSR1 and silencing the CYP24A1/SCD desaturation program. This calcium-ferroptosis link unveils a druggable vulnerability in CRC and points to artemisinin derivatives as templates for future therapeutic innovation.

2. Results

2.1. Anti-proliferative activity evaluation of ATT in CRC

Artemisinin derivatives exhibit broad-spectrum anticancer activity [49], and clinical studies have suggested potential therapeutic benefits of artesunate in colorectal cancer patients [50]. To map the targetable signaling landscape of this family in colorectal cancer, a more potent analogue must first be identified. We therefore carried out a systematic cytotoxicity screen of a panel of artemisinin-based analogues across multiple colorectal cancer-relevant tumor cell lines. The results showed that ATT exhibited the highest cytotoxicity, surpassing that of dihydroartemisinin and artesunate, whereas artemisinin and artemether were only weakly effective (Fig. S1). ATT induced cell death in a strictly time-dependent manner and displayed minimal toxicity toward normal NCM460 epithelial cells (Fig. 1B). These findings indicate that structural variations within the artemisinin scaffold underlie their divergent anticancer activities.

Fig. 1.

Proliferation inhibition of CRC cells by ATT. (A) Chemical structure of artemisitene (ATT). (B) HCT-116, HT-29, HCT-8 and NCM460 cells were treated with the indicated concentrations of ATT for 24 and 48 h. Cell viability was measured by the CCK-8 assay. (C) EdU staining and quantitative analysis of CRC cells after treatment with ATT (0, 5, 10, 20 μM) for 24 h. Scale bar: 50 μm. (D) Representative images and quantitative analysis of the colony formation in CRC cells treated with ATT (0, 5, 10, 20 μM) for 24 h and then cultured for 14 days. Data are expressed as mean ± SD (n = 3). Statistical significance was assessed using one-way ANOVA followed by Dunnett's post hoc test. P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

Indeed, structural modifications of the artemisinin scaffold confer distinct anticancer potencies. In both HCT-116 and HT-29 cells, ATT treatment markedly reduced EdU incorporation, indicating impaired DNA synthesis and proliferation (Fig. 1C). Consistently, colony formation was profoundly suppressed (Fig. 1D), underscoring its sustained cytostatic effect. Together, these data demonstrate that ATT selectively restrains CRC cell growth while sparing normal intestinal epithelial cells.

2.2. ATT downregulated CYP24A1 expression to elevate calcium levels and induce ferroptosis in CRC cells

Building on the pronounced antiproliferative efficacy of ATT, we next employed high-depth RNA-sequencing (RNA-seq) to dissect the underlying transcriptional circuitry driving its tumor-suppressive activity. The KEGG pathway enrichment analysis revealed that the signaling pathways related to the calcium signaling pathway and fatty acid metabolism underwent significant alterations (Fig. 2A). The heatmap depicted the relative expression profiles of these genes (Fig. 2B). Volcano plots illustrated both upregulated and downregulated differentially expressed genes (DEGs) in ATT-treated cells versus control cells (Fig. 2C).

Fig. 2.

Analysis of ATT-mediated calcium regulation via RNA-seq. (A) The analysis of the KEGG pathway for ATT-treated cells (10 μM). (B) Heatmap showing the relative expression of differentially expressed genes in HCT-116 cells with or without ATT treatment (10 μM). (C) Volcano plots of DEGs in ATT vs. control HCT-116 cells. |Fold change| > 2 and P < 0.05 were shown in red (upregulated) or blue (downregulated). (D, E) Cellular calcium level of CRC cells in HCT-116 and HT-29 cells following ATT treatment (0, 5, 10, 20 μM) for 24 h. CRC cells were treated with ATT (0, 5, 10, 20 μM) for 24 h. The influence of ATT on the RNA (F) and protein levels (G, H) of the differentially expressed protein CYP24A1 were assessed. (I) The immunofluorescence analysis of CYP24A1 expression in CRC cells with ATT treatment (10 μM) for 24 h. Scale bar: 50 μm. Data are expressed as mean ± SD (n = 3). Statistical significance was assessed using one-way ANOVA followed by Dunnett's post hoc test. P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

In light of transcriptomic findings showing that ATT significantly modulates calcium-signaling pathways, we quantitatively mapped intracellular Ca²⁺ dynamics with the Fluo-4 AM probe. This functional interrogation revealed a pronounced surge in cellular Ca²⁺ upon ATT treatment (Fig. 2D and E), corroborating the RNA-seq-predicted dysregulation of calcium homeostasis. Subsequent comparative transcriptomics was therefore directed toward screening for the differentially expressed genes selectively enriched for regulators of Ca²⁺ homeostasis that are causally linked to ATT-induced Ca²⁺ overload. Among the markedly down-regulated genes, CYP24A1 emerged as a key observation that disrupts calcium homeostasis to promote tumor cell proliferation [51], potentially playing a crucial role in the calcium overload triggered by ATT. Following RNA-seq results, the mRNA level of CYP24A1 was markedly decreased in both HCT-116 and HT-29 cells (Fig. 2F), and ATT dose-dependently inhibited CYP24A1 expression (Fig. 2G and H). Consistently, immunofluorescence staining further confirmed a pronounced reduction of CYP24A1 at the protein level upon ATT treatment (Fig. 2I). These data suggest that ATT induces calcium dysregulation by down-regulating CYP24A1. Calcium dysregulation has been reported to contribute to various forms of cell death, including apoptosis [52] and ferroptosis [16]. To define the mode of cell death triggered by ATT in CRC cells, we interrogated a panel of cell death inhibitors. Cell viability assays revealed that ATT-triggered death was almost completely prevented by NAC and DFOM, partially attenuated by Ferrostatin-1 and Z-VAD-FMK, and unaltered by Necrostatin-1 (Fig. 3A), indicating a predominant role for ferroptosis. Concordantly, ATT treatment led to a pronounced increase in intracellular oxidative stress and membrane lipid peroxidation, as detected by DCFH-DA and C11-BODIPY probe, respectively (Fig. 3B–E), and these increases were reversed by the ferroptosis inhibitors NAC and DFOM (Fig. S2). In contrast, ATT did not induce significant changes in intracellular calcium levels or oxidative stress in NCM460 cells, suggesting the tumor-specific selectivity of ATT (Fig. S3A-S3C).

Fig. 3.

ATT induced calcium-related ferroptosis. (A) HCT-116 and HT-29 cells were pretreated with various inhibitors (NAC (5 mM), DFOM (100 μM), Ferrostatin-1 (5 μM), Z-VAD-FMK (20 μM), or Necrostatin-1 (20 μM)) for 12 h, followed by ATT treatment for 24 h. Cell viability was then assessed. HCT-116 and HT-29 cells were exposed to ATT for 24 h. Flow cytometry analysis measured oxidation-dependent DCF fluorescence (1 μM) as an indicator of general cellular oxidative activity (B, C), and C11-BODIPY fluorescence (1 μM) as a marker of lipid peroxidation (D, E). HCT-116 and HT-29 cells were treated with ATT (10 μM) with or without BAPTA-AM (2 μM) or A23187 (5 μM) for 24 h. Intracellular oxidative activity (F) and lipid peroxidation (G, H) were measured by flow cytometry using DCFH-DA and C11-BODIPY probes, respectively. Data are expressed as mean ± SD (n = 3). Statistical significance was assessed using one-way ANOVA, followed by Dunnett's test for comparisons with the control group and Tukey's test for pairwise comparisons among all groups. P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

Having established that ATT engages ferroptosis, we next asked whether this process is calcium-dependent. We therefore co-treated cells with the membrane-permeable Ca²⁺ chelator BAPTA-AM or the Ca²⁺ ionophore A23187 alongside ATT. As predicted, BAPTA-AM blunted, whereas A23187 amplified, the ATT-elicited surge in cellular and lipid ROS (Fig. 3F–H). However, these modulatory effects were absent in NCM460 cells (Fig. S3D and S3E). Collectively, these results support a model in which ATT suppresses CYP24A1 and elevates intracellular calcium.

2.3. Calcium regulated fatty acid metabolism to induce ferroptosis

Leveraging our prior finding that ATT-induced Ca²⁺ surge is the proximal trigger of ferroptosis, we next dissected how this Ca²⁺ disturbance propagates its ferroptosis signal. Given that calcium homeostasis reconfigures lipid storage dynamics and membrane lipid composition, which are the primary biochemical determinants that calibrate the ferroptosis threshold [53,54], we reasoned that ATT-evoked calcium dysregulation would reprogram fatty acids. LC-MS profiling confirmed the hypothesis that ATT expanded the intracellular pools of both saturated and unsaturated free fatty acids (Fig. 4A–B), priming CRC cells for ferroptosis. Among unsaturated species, polyunsaturated fatty acids (PUFA) accumulated to a greater extent than monounsaturated fatty acids (MUFAs) (Fig. 4C), a shift known to heighten lipid peroxidation and ferroptosis sensitivity. More importantly, among the MUFAs, palmitoleic acid (C16-1) alone declined markedly after ATT exposure, diverging from the modest increases observed in other MUFAs, while its precursor palmitic acid (C16-0) accumulated. This pattern was consistently observed in both HCT-116 and HT-29 cells, but was absent in NCM460 cells (Fig. 4D and S3F). SCD has been reported to be suppressed upon cytosolic Ca²⁺ elevation [55]. These reciprocal changes pointed out that SCD was regulated by ATT. Consistently, ATT repressed SCD expression in a concentration-dependent manner in both HCT-116 and HT-29 cells, with reductions observed at both the protein and mRNA levels (Fig. 4E–G). To investigate how calcium regulates SCD expression, we examined the CaMKK2/AMPK/SREBF1 signaling axis previously implicated in lipogenic control [56]. ATT treatment markedly enhanced CaMKK2-dependent AMPK phosphorylation in HCT-116 and HT-29 cells, concomitant with reduced SREBF1 and consequent SCD down-regulation (Fig. 4H). These changes were undetectable in normal colon epithelial NCM460 cells, indicating that the CaMKK2/AMPK/SREBF1/SCD pathway is selectively engaged by ATT in colorectal cancer cells (Fig. S3G). Collectively, these data reveal that ATT-driven calcium overload disrupts fatty-acid homeostasis via SCD inhibition, thereby steering CRC cells toward ferroptosis.

Fig. 4.

The free fatty acid changes of CRC cells after ATT treatment. HCT-116 cells were treated with ATT (10 μM) for 24 h, and the overall changes in intracellular free saturated fatty acids (A) and unsaturated fatty acids (B) were observed. (C) The relative changes in monounsaturated fatty acids (MUFAs) and polyunsaturated fatty acids (PUFAs) within CRC cells. HCT-116 cells and HT-29 cells were treated with ATT (10 μM) for 24 h. (D)The content changes of palmitic acid and palmitoleic acid in cells after the treatment of ATT (10 μM). Statistical significance was assessed using unpaired two-tailed Student's t-test. The WB (E) and immunofluorescence (F) analysis of SCD expression in CRC cells with ATT treatment. Scale bar: 50 μm. (G)The RNA changes of SCD in CRC cells were tested. (H) The effects of ATT on the protein levels of CAMKK2, AMPK, p-AMPK, and SREBF1 were determined by Western blot analysis. Statistical significance was assessed using two-way ANOVA followed by Dunnett's post hoc test. Data are expressed as mean ± SD (n = 3). P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

2.4. ATT directly binds to LSH in CRC cells

Since CYP24A1 is transcriptionally controlled by regulators involved in fatty acid metabolism, we sought the molecular regulator by which ATT represses CYP24A1 expression. A focused screen singled out the lymphoid-specific helicase (LSH) as the most plausible mediator [38]. Biophysical analyses confirmed a direct interaction between ATT and LSH with micromolar affinity. Surface plasmon resonance (SPR) yielded a K_D of 6.18 μM (Fig. 5A), and both drug affinity responsive target stability (DARTS) and cellular thermal shift assays (CETSA) in HCT-116 (Fig. 5B–C) and HT-29 cells (Fig. S7) independently substantiated this binding.

Fig. 5.

ATT directly binds to LSH. (A) The dissociation constant between ATT (0.625 μM–10 μM) and LSH was determined using SPR. (B) The cell lysates of HCT-116 were pre-incubated with or without ATT (10, 50 μM), followed by treatment with different concentrations of pronase to detect protein changes. Statistical significance was assessed using two-way ANOVA followed by Dunnett's post hoc test. (C) HCT-116 cells were treated with DMSO or ATT (20 μM) for 2 h at 37 °C, followed by gradient heating from 37 °C to 77 °C. Soluble LSH was detected by Western blotting. Data are expressed as mean ± SD (n = 3). Statistical significance was assessed using two-way ANOVA followed by Šídák's multiple comparisons test. P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

Following the identification of LSH as the target of ATT, we hypothesized that the electrophilic α, β-unsaturated ketone in ATT might drive covalent adduction through Michael addition. LC-MS/MS of ATT-incubated recombinant LSH revealed a site-specific modification at Cys205 (Fig. 6A and S4). Replacing this cysteine with alanine (C205A) (Fig. 6B) markedly attenuated ATT binding in CETSA, DARTS, and SPR assays (Fig. 6C–E), and critically abolished ATT-induced ferroptosis in HCT-116 cells (Fig. 6F). Collectively, these data establish that ATT alkylates Cys205 of LSH, thereby licensing its downstream suppression of CYP24A1 and consequent sensitization to ferroptosis.

Fig. 6.

ATT covalently modified the Cys205 residue of LSH. (A) Recombinant LSH protein (50 μg) was incubated with ATT (50 μM). MS/MS analysis was performed and mass spectra of the peptide containing the target site were obtained. (B) Schematic diagram of site mutation. (C) CETSA and (D) DARTS experiments were conducted to investigate the binding affinity of ATT to Flag-tagged wild-type LSH protein and C205A mutant protein. (E) Determination of dissociation constants between mutant recombinant proteins and ATT (1.25–80 μM) in vitro. (F) Effect of the mutant on ATT-induced increase in lipid reactive oxygen species. Data are expressed as mean ± SD (n = 3). Statistical significance was assessed using one-way ANOVA followed by Tukey's post hoc test. P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

Given these observations, the functional relevance of LSH in ATT-induced ferroptosis was evaluated through loss- and gain-of-function approaches. Cell viability assays revealed that LSH knockout attenuated the cytotoxic effect of ATT (Fig. 7A), whereas LSH overexpression rendered cells more sensitive to ATT (Fig. 7B). Consistently, LSH knockdown attenuated the ATT-induced rise in oxidative stress and intracellular Ca²⁺ levels (Fig. 7C–E). In contrast, LSH overexpression further promoted ferroptosis susceptibility, together with stronger oxidation-dependent fluorescence and Ca²⁺ accumulation compared with ATT alone (Fig. 7F–H). With respect to CYP24A1 regulation, LSH silencing diminished ATT-mediated repression of CYP24A1, approaching levels observed in the knockdown group (Fig. 7I). Conversely, the upregulation of CYP24A1 driven by LSH overexpression was efficiently suppressed by ATT (Fig. 7J). Similarly, LSH knockdown in HT-29 cells produced comparable effects, confirming the generality of this regulatory mechanism across colorectal cancer cell lines (Fig. S8). These results underscore the functional significance of LSH in ATT-induced ferroptosis. Together, these results reveal that ATT directly targets LSH to mediate its ferroptosis-inducing activity in CRC cells.

Fig. 7.

The crucial role of LSH in ATT-induced ferroptosis in HCT-116. HCT-116 cells were pre-treated with siLSH (A) or Flag-LSH (B) for 24 h and then treated with or without ATT (10 μM) for 24 h. Cytotoxicity of ATT in cells was tested. Flow cytometry was performed to evaluate oxidation-dependent DCF fluorescence (C), C11-BODIPY-detected lipid peroxidation (D), and intracellular Ca²⁺ levels (E) in CRC cells with LSH knockdown after ATT treatment. HCT-116 cells were pre-treated with Flag-LSH for 24 h and then treated with or without ATT (10 μM) for 24 h. Flow cytometry was used to analyze the effects of ATT on intracellular oxidative activity (DCFH-DA) (F), lipid peroxidation (C11-BODIPY) (G), and calcium levels (H) in CRC cells with LSH overexpression. Effects of ATT on target protein LSH and downstream protein CYP24A1 upon LSH knockdown (I) and LSH overexpression (J). Data are expressed as mean ± SD (n = 3). Statistical significance was assessed using one-way ANOVA followed by Tukey's post hoc test. P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

2.5. ATT alkylates LSH to dismantle the LSH-EWSR1 complex and silence CYP24A1 transcription

As an SNF2-family chromatin remodeler, LSH normally drives CYP24A1 transcription by evicting promoter-proximal nucleosomes [36]. Quantitative PCR and immunoblot analyses revealed that ATT did not significantly alter LSH mRNA or protein abundance in either HCT-116 or HT-29 cells (Fig. S5). We therefore propose that ATT allosterically disrupts the recruitment of transcriptional co-factors to the LSH complex, thereby repressing CYP24A1 expression without affecting LSH itself.

We next sought to understand how ATT influences CYP24A1 expression via LSH by mapping the LSH interactome through co-immunoprecipitation (co-IP). Silver staining revealed a discrete ∼100 kDa band whose constituents were identified by LC-MS/MS. Guided by peptide counts and abundance, we prioritized eleven candidates (Fig. 8A) and confirmed EWSR1 (EWS RNA binding protein 1) as the principal LSH partner. Reciprocal co-IP analysis demonstrated robust LSH-EWSR1 association in HCT-116 cells, an interaction that was markedly attenuated by ATT (Fig. 8B). Confocal microscopy further showed nearly complete nuclear co-localization of LSH and EWSR1, a pattern disrupted upon ATT treatment (Fig. 8C). Finally, ectopic expression of Flag-tagged LSH-C205A bearing the covalent ATT-binding cysteine mutated to alanine displayed significantly impaired EWSR1 binding relative to wild-type LSH (Fig. 8D). Collectively, these results establish that ATT alkylates Cys205 of LSH, thereby dissociating the LSH-EWSR1 complex.

Fig. 8.

Identification of LSH-interacting proteins perturbed by ATT. HCT-116 cells were treated with ATT (10 μM) for 24 h. (A) LC-MS/MS analysis of LSH-immunoprecipitated proteins in HCT-116 cells (left). Representative peptide counts and molecular weights of indicated proteins are shown (right). (B) Co-immunoprecipitation (Co-IP) analysis of the interaction between LSH and EWSR1 in CRC cells. (C) Immunofluorescence co-localization analysis of EWSR1 and LSH in CRC cells. Scale bar: 5 μm. (D) HCT-116 cells were pre-treated with Flag-LSH for 24 h and then treated with or without ATT (10 μM) for 24 h. Co-IP analysis of interactions between Flag-tagged wild-type LSH, mutant, and EWSR1. (E) HCT-116 cells were treated with ATT (0, 5, 10, 20 μM) for 24 h. Protein levels of EWSR1 in CRC cells were determined. (F) HCT-116 cells were treated with siEWSR1 for 48 h. Effect of EWSR1 knockdown on HCT-116 cell viability. Influence of EWSR1 knockdown on lipid peroxidation (C11-BODIPY) (G) and intracellular calcium levels (H). Statistical significance was assessed using one-way ANOVA followed by Dunnett's post hoc test. Impact of EWSR1 knockdown on (I) RNA and (J) protein expression of LSH and CYP24A1. Statistical significance was assessed using two-way ANOVA followed by Dunnett's post hoc test. (K) HCT-116 cells were transfected with siEWSR1 for 48 h, and a dual-luciferase assay was performed to examine its effect on the promoter activity of CYP24A1 mediated by exogenous LSH. Statistical significance was assessed using one-way ANOVA followed by Tukey's post hoc test. Data are expressed as mean ± SD (n = 3). P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

Remarkably, ATT left both the mRNA and protein levels of EWSR1 unperturbed (Fig. 8E and S5C). To dissect the role of EWSR1 in ferroptosis, we depleted it in HCT-116 cells. EWSR1 silencing triggered marked cell death (Fig. 8F), together with elevated membrane lipid peroxidation (Fig. 8G) and increased intracellular Ca²⁺ levels (Fig. 8H). Concurrently, the transcript and protein abundance of CYP24A1 and LSH were sharply reduced (Fig. 8I–J). The dual-luciferase reporter assay confirmed that the LSH-induced enhancement of CYP24A1 promoter activity was attenuated upon EWSR1 knockdown (Fig. 8K). We conclude that EWSR1 cooperates with LSH to drive CYP24A1 transcription, whereas ATT, by covalently engaging LSH, disrupts this complex and extinguishes CYP24A1 expression.

2.6. ATT blocked CRC progression in vivo

An HCT-116 xenograft model was employed to evaluate ATT's antitumor activity in vivo. Compared with vehicle control, ATT administration produced a pronounced reduction in both tumor volume and tumor weight (Fig. 9A–C). Hematoxylin and eosin staining revealed sparse tumor cellularity, extensive nuclear pyknosis, and widespread cellular disintegration in ATT-treated lesions. Concordantly, Ki-67 immunostaining confirmed a marked decline in proliferative indices (Fig. 9D). Immunohistochemical quantitation and Western blot jointly showed that ATT selectively down-regulated CYP24A1 and SCD while leaving LSH abundance unchanged (Fig. 9E), mirroring the in vitro observations. Co-IP assay additionally verified that ATT inhibited the LSH-EWSR1 interaction within the tumors (Fig. 9F). Importantly, ATT showed an excellent safety margin, as body-weight curves overlapped between groups and histologic examination of major organs revealed no treatment-related pathology (Fig. S9A and S9C). Similarly, in non-tumor-bearing mice treated with ATT alone, no significant toxicity was observed based on body-weight curves and H&E staining of major organs (Fig. S9B and S9D). Collectively, these data demonstrate that ATT potently restrains CRC progression in vivo while exhibiting negligible systemic toxicity.

Fig. 9.

The effect of ATT on CRC progression in vivo. (A) Representative images of tumor samples from each group (n = 6). (B) Measurement of tumor volume and (C) weight (n = 6). (D) HE staining of tumor tissues (Scale bar: 100 μm) and IHC detection of Ki-67, CYP24A1, SCD, and LSH expression in tumor tissues (Scale bar: 50 μm). (E) Protein expression levels of CYP24A1, SCD, and LSH in tumor tissues were determined. (F) Impact of ATT on the interaction of LSH and EWSR1 in vivo was assayed by Co-IP assay. Statistical significance was assessed using one-way ANOVA followed by Dunnett's post hoc test. P-values are labeled directly in figure. P < 0.05 was considered statistically significant, and P ≥ 0.05 was considered non-significant.

3. Discussion

Ferroptosis, a programmed cell death driven by lipid peroxidation, GSH depletion, and iron overload, has emerged as a compelling therapeutic target in oncology [5]. Traditionally, ferroptosis has been characterized as an iron-dependent form of cell death driven predominantly by excessive lipid peroxidation. Most known ferroptosis inducers act through canonical pathways centered on GPX4 [57], GSH [58], and iron metabolism [59]. Although these downstream effectors serve as critical executors of ferroptosis, targeting upstream regulatory proteins may offer broader and more flexible strategies to trigger ferroptosis. Increasing evidence indicates that calcium signaling plays a pivotal role in modulating ferroptosis sensitivity. In colorectal cancer, intracellular calcium levels are frequently elevated, supporting tumor proliferation at moderate concentrations [60] but triggering cell death when excessively high [[61], [62], [63]]. The intrinsic calcium burden of CRC cells renders them particularly susceptible to calcium-dependent ferroptosis, with ER-mitochondria calcium transfer contributing to this vulnerability [16]. CYP24A1 is a mitochondrial enzyme that attenuates the antitumor effects of calcitriol (1,25(OH)₂D₃) by catalyzing its inactivation. It has been reported that CYP24A1 activated the MAPK pathway to promote tumor survival by regulating calcium levels [51,64]. Our findings revealed that ATT induces a calcium-dependent form of ferroptosis, representing a mechanistically distinct route from the canonical GPX4-or iron-centered pathways. Specifically, ATT treatment triggered calcium-dependent ferroptosis in colorectal cancer cells by suppressing CYP24A1 expression, thereby disrupting calcium homeostasis and promoting oxidative stress-mediated lipid peroxidation.

Integral to this program is the ensuing disruption of fatty-acid homeostasis. Given that fatty acids are key mediators of ferroptosis, tightly controlling their metabolic pathways is essential, and perturbation of this network constitutes a potent therapeutic approach. Previous studies have demonstrated that blocking FASN (fatty acid synthase) in mutant KRAS lung cancer promotes ferroptosis [65]. Calcium signaling is increasingly recognized as a pivotal regulator of cancer metabolism, influencing glycolysis and lipid synthesis [66] and tumor progression through modulation of calcium channels and downstream kinases such as CaMKs and PKC [67]. Previous evidence reported that CaMKK2 acts as an upstream kinase of AMPK in response to calcium flux [56], and that AMPK activation represses SREBP1-dependent transcription of lipogenic genes [68], including SCD. Dysregulation of the CaMKK2/AMPK/SREBF1 axis has been implicated in aberrant metabolic reprogramming and malignant growth, underscoring its relevance in tumor lipid metabolism. Given that lipid homeostasis critically determines ferroptosis sensitivity, perturbation of calcium signaling offers an alternative means to modulate lipid-driven cell death. In our study, metabolomic profiling revealed that ATT triggers a global remodeling of free fatty acids with SCD, the USP7 substrate that suppresses ferroptosis, showing the sharpest decline [69]. In line with this, the tight concordance between SCD loss and ATT-induced fatty acid signatures confirms that the ATT links calcium dysregulation to lipid metabolic collapse, driving ferroptosis. Moreover, activation of the CaMKK2/AMPK/SREBF1 signaling cascade was observed, which ultimately led to the suppression of SCD. The resulting alterations in lipid composition rendered cells more vulnerable to ferroptosis damage. Collectively, our findings not only identify calcium overload as a novel upstream event in ferroptosis regulation, but also expand the current understanding of how cellular metabolic and signaling networks converge to determine ferroptosis susceptibility.

Recent investigations have revealed a second, rapidly expanding therapeutic dimension of artemisinin and its analogues: potent anticancer activity that is intimately coupled to the iron-dependent form of regulated cell death known as ferroptosis [70]. The underlying mechanisms involve downregulation of SLC7A11 and GPX4 expression [71], inhibition of PRDX1/2 enzymatic activity [72], and modulation of cellular iron metabolism [73]. Among these mechanisms, the regulation of iron homeostasis has been particularly well characterized. Artesunate, for example, docks directly onto the transferrin receptor (TFRC) [74], stabilizing its plasma membrane residency and promoting iron import, whereas dihydroartemisinin chelates ferritin through its endoperoxide bridge, facilitating lysosomal degradation and intracellular iron release. However, current research remains largely focused on the reactivity of the endoperoxide bridge, which appears indispensable for the biological activity of artemisinin compounds, particularly in iron metabolism-related mechanisms. Whether other structural motifs within this chemical family can serve as alternative active centers remains unclear. Moving beyond traditional solubilizing modifications, the deliberate installation of alternative electrophilic or metal-binding motifs may open an orthogonal path to next-generation artemisinins with amplified and mechanistically diversified bioactivity. Earlier computational efforts nominated FDFT1 as a putative mediator of the apoptotic effects of ATT [45]. However, neither the essential pharmacophoric elements nor the precise binding mode of ATT has been experimentally characterized. In this study, we found that the introduction of a double bond within the molecular scaffold markedly enhanced both the antitumor potency of ATT compared with artemisinin derivatives lacking this feature. In normal colon epithelial NCM460 cells, ATT treatment did not induce detectable cell death or ferroptosis hallmarks, and in vivo administration further confirmed its favorable safety profile. Through comprehensive transcriptomic profiling and integrative functional analyses, we identified LSH as a direct and druggable target that mediates ATT-induced ferroptosis. As an epigenetic regulator, LSH orchestrates the transcriptional regulation of ferroptosis-related genes through chromatin remodeling. Specifically, LSH functions as a pivotal upstream modulator of CYP24A1 expression, thereby establishing a mechanistic connection between ATT-induced LSH targeting, calcium dysregulation, and ferroptosis execution. SPR analysis confirmed the binding between ATT and protein LSH, while mass spectrometry pinpointed a critical Cys205 residue as the covalent binding site. Mutation of this residue markedly reduced the binding and reversed the ferroptosis phenotype induced by ATT. The unique double-bond structure of ATT functions as a novel electrophilic center that enables direct interaction with the chromatin remodeler LSH. In addition, artemisinin, devoid of the electrophilic α, β-unsaturated ketone, exhibited a markedly reduced affinity for LSH (K_D = 32.22 μM) (Fig. S6), underscoring this moiety as an indispensable determinant of target engagement. The superior LSH-binding capacity of ATT is thus a primary driver of its substantially enhanced cytotoxicity relative to the parent compound. Collectively, we unveil a previously unrecognized paradigm in which an artemisinin derivative ignites ferroptosis through direct alkylation of LSH. These findings emphasize that, beyond the canonical endoperoxide pharmacophore, peripheral electrophilic substituents on the artemisinin scaffold dictate molecular specificity, thereby providing a rational blueprint for the next generation of artemisinin-based antineoplastics.

LSH is an SNF2-family chromatin remodeler whose intrinsic nucleosome-binding affinity is weak, usually relying on auxiliary partners to achieve stable genomic engagement [75]. Using ATT-treated CRC cells as a functional filter, we performed LSH immunoprecipitation-MS and identified the RNA-binding protein EWSR1 as an ATT-sensitive LSH interactor. This finding constitutes the central mechanistic advance of the present study. EWSR1 is an RNA-binding protein (RBP) that orchestrates circular RNA (circRNA) biogenesis to drive tumorigenesis by interfering with the function of transcription initiation complexes [76]. The aberrant transcriptional function of EWSR1 arises from either chromosomal translocations that generate chimeric proteins (e.g., EWSR1-FLI1 in Ewing sarcoma) or direct interactions with transcription factors that reprogram downstream oncogenic pathways [77,78]. Recent work has further shown that EWSR1 recruits the chromatin-remodeling SWI/SNF complex to trigger aberrant oncogene transcription and translation, thereby accelerating tumor progression [79]. Our data now position EWSR1 as an obligate partner of LSH. siRNA-mediated knockdown of EWSR1 extinguished CYP24A1 transcription, phenocopying the transcriptional shutdown observed upon ATT treatment. Moreover, ATT selectively disrupted the LSH-EWSR1 interface without altering the abundance of either protein, thereby preventing LSH from engaging the chromatin-remodeling machinery required for transcriptional activation. This ligand-induced disassembly of the LSH-EWSR1 complex represents the key molecular event that converts ATT binding into ferroptosis sensitivity and defines the principal discovery of our work.

Collectively, we uncover a previously unrecognized pathway in which ATT alkylates LSH, disrupts its partnership with EWSR1, and silences the LSH/CYP24A1/SCD axis. The ensuing Ca²⁺ overload and fatty acid remodeling converge on lethal lipid peroxidation, driving ferroptosis in colorectal cancer cells. By precisely targeting the EWSR1-LSH interface, ATT not only offers a potent therapeutic adjuvant for CRC but also establishes a paradigm that may extend to other malignancies reliant on this epigenetic circuitry. Future studies employing unbiased chemical proteomics and complementary multi-omics approaches will be instrumental in mapping the broader target landscape of ATT and refining its translational potential.

4. Conclusion

In summary, ATT reprograms lipid metabolism in CRC by coupling calcium overload to ferroptosis. Covalent inactivation of LSH disrupts the LSH-EWSR1 complex, collapses the LSH/CYP24A1/SCD axis, thereby halting tumor growth in vitro and in vivo. These results establish LSH as a central regulator of calcium-dependent ferroptosis and highlight ATT, as well as artemisinin derivatives in general, as targeted agents for malignancies reliant on redox-sensitive lipid pathways.

5. Materials and methods

5.1. Cell lines, plasmid, and agents

HCT-116, HT-29, HCT-8 and NCM460 cell lines were purchased from the Cell Bank of Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences (Shanghai, People's Republic of China). HCT-116 and HT-29 cells were cultured in McCoy's 5A medium (KeyGEN Biotechnology, China) supplemented with 10 % fetal bovine serum (Lonsera, USA). HCT-8 cells were maintained in RPMI-1640 medium (GIBCO, USA) containing 10 % FBS. NCM460 cells were maintained in Dulbecco's Modified Eagle Medium (Hyclone, USA) supplemented with 10 % FBS (Lonsera, USA). All cells were incubated at 37 °C in a humidified atmosphere containing 5 % CO₂. Reagents used in cell culture were as follows: artemisitene (Baoji Herbest Biotechnology, HA155873, China), artemisinin (Macklin, A800831, China), artesunate (Macklin, A800614, China), dihydroartemisinin (Macklin, D831931, China), artemether (Macklin, A800587, China), DFOM (Aladdin, D409090, China), Ferrostatin-1 (Aladdin, F129882, China), NAC (Aladdin, A105422, China), Z-VAD-FMK (Aladdin, Z408507, China), Necrostatin-1(Aladdin, N125522, China), A23187 (Aladdin, C137844, China), BAPTA-AM (Aladdin, B115502, China).

5.2. Cell viability assay

According to the manufacturer's protocol, the Cell Counting Kit-8 (Apexbio, K1018, USA) assay was performed to measure the IC₅₀ value of artemisitene. Briefly, cells were seeded into 96-well plates at 5 × 10³ per well, treated with medium containing different concentrations of compounds, and incubated at 37 °C for 24 and 48 h. The absorbance was read at 450 nm in a SpectraMax Plus 384 plate-reader (Molecular Devices, Sunnyvale, CA, USA). IC₅₀ values were calculated on GraphPad Prism 9.

5.3. Transfection

Plasmids and siRNAs were purchased from Saisofi Biotechnology (China) and Tsingke Biotechnology (China). Lipofectamine 3000 Transfection Reagent (Invitrogen, L3000015) was used for transfection. In simple terms, 30 % of the fused cells were transfected according to the instructions. After 6 h, the medium was replaced, and the cells were harvested after 48 h. The relevant sequences are provided in the Supplementary Table S2.

5.4. DNA synthesis assay

Cells were cultured in 96-well culture plates and treated with artemisitene (0, 5, 10, 20 μM) for 24 h. EdU (Apexbio, K1076, China) was added to the incubator for an additional 2 h. Then, cells were fixed with 4 % paraformaldehyde for 30 min and were permeabilized with 0.5 % Triton X-100. Afterward, the cells were individually stained with the fluor Apollo 567 and Hoechst 33342 for 30 min. HCT-116 cells were observed with an imageXpress® confocal microscope (Molecular Devices, CA, USA).

5.5. Colony formation assay

Cells (2000 cells/well) were seeded in 6-well plates and cultured overnight. Then the cells were treated with serial concentrations of artemisitene for 24 h. The medium was discarded and replaced with a new one. The procedure was repeated until the colonies formed after 14 days. Then the cells were fixed with 4 % paraformaldehyde for 30 min and stained with a crystal violet solution. Colonies were photographed and counted.

5.6. Measurement of intracellular oxidative stress and lipid peroxidation

Cells were seeded in 6-well plates and cultured overnight. After artemisitene treatment for 24 h, the cell medium was replaced with a medium containing the DCFH-DA probe (1 μM) (Beyotime Biotechnology, S0033S, China) or C11-BODIPY^581/591 dye (1 μM) (Invitrogen, D3861, USA). After 30 min, cells were collected and analyzed by flow cytometry. The results were analyzed by FlowJo 10.8.1 software.

5.7. Intracellular Ca²⁺ detection

Cells were collected, washed with PBS, and incubated with Fluo-4 AM (5 μM) (Beyotime Biotechnology, S1060, China) for 30 min. Cells were resuspended in PBS and analyzed by flow cytometry. The results were analyzed by FlowJo 10.8.1 software.

5.8. Cellular thermal shift assay (CETSA)

CETSA was performed as previously described. Briefly, cells were treated with DMSO and artemisitene (20 μM) for 2 h. Then the cells were collected and resuspended with PBS containing phenylmethylsulfonyl fluoride (PMSF) (Beyotime Biotechnology, ST505, China). Each group was equally divided into nine groups and heated at different temperatures for 3 min, followed by cooling for 3 min on ice. Finally, suspensions were freeze-thawed three times with liquid nitrogen and centrifuged at 12,000 rpm for 10 min at 4 °C to obtain the soluble fractions. The supernatants were then subjected to SDS-PAGE and Western blot.

5.9. Drug affinity responsive target stability (DARTS)

Untreated cell lysates were incubated with DMSO and artemisitene (10 μM and 50 μM) at 37 °C. After 1 h, the pronase (1 μg/ml) (Roche, 10165921001, Germany) was added to each sample and incubated at 37 °C for 30 min. The reaction was terminated by PMSF (1 mM), followed by SDS-PAGE and Western blot analysis.

5.10. Determination of affinity constant

SPR measurements were performed on a Biacore S200 instrument (GE Healthcare, Uppsala, Sweden). LSH protein was immobilized on a CM5 sensor chip using an amine coupling procedure. Gradient concentrations of artemisitene were injected into the LSH chip for 60 s, followed by dissociation for 120 s. The dissociation constant (K_D) was calculated by Biacore S200 evaluation software.

5.11. Purification of LSH

The full length of LSH was inserted into the pET28a vector. The plasmid was transformed into E. coli BL21 (DE3) cells, LSH were expressed in E. coli BL21(DE3) cells overnight at 37 °C. Overexpression was induced at 20 °C with isopropyl-β-D-thiogalactoside (0.2 mM) for 14 h. The cells were collected and lysed ultrasonically in buffer A (10 mM imidazole, 300 mM KCl, 50 mM KH₂PO₄, PH 8.0). Then the supernatant was separated from cell debris by centrifugation at 18,000×g for 30 min at 4 °C and loaded onto a Ni-NTA column (Smart-Lifesciences, China) that equilibrated with buffer A. LSH was gradient eluted with buffer A supplemented with 20 mM–250 mM imidazole. Subsequently, the target protein LSH was determined by Coomassie blue stain and concentrated by Ultrafiltration centrifugation (Meck, USA).

5.12. Immunofluorescence

The cells treated with artemisitene were washed with PBS, fixed with 4 % paraformaldehyde for 30 min, and then permeabilized with 0.5 % Triton X-100 for 10 min. Then, 5 % BSA was added and incubated for 2 h. The primary antibody and fluorescent secondary antibody were incubated overnight at 4 °C and 2 h at room temperature, respectively. Finally, DAPI (Beyotime Biotechnology, C1002, China) was stained for 10 min. Fluorescent images were observed with a confocal laser scanning microscope (CLSM, LSM 800, Zeiss, Germany) and processed using the ZEN imaging software. The antibodies used are as follows: LSH (Santa Cruz, sc-46665, USA), EWSR1 (ABclonal, A9640, China), CYP24A1 (Origene, TA368684S, USA), and SCD (ABclonal, A26246, China).

5.13. RNA extraction and RT-qPCR

RNA was extracted by RNA-Quick Purification Kit (Yishan Biotechnology, ES-RN001, China) and converted to cDNA using Reverse Transcription Kit (Vazyme, R223-01, Nanjing, China). Real-time quantitative PCR (qPCR) was performed on a LightCycler 480 system (Roche, Switzerland) using SYBR Green qPCR Mix (Vazyme, Q311-02, China). Primer sequences are listed in Supplementary Table S1. RNA sequencing (RNA-seq) was conducted by Tsingke Biotechnology (China).

5.14. Protein mass spectrometry identification

The samples from immunoprecipitation were separated by SDS-PAGE and stained with Coomassie blue. Then, the target gel bands were excised and sent for mass spectrometry analysis (Personal Biotechnology, Shanghai, China). Briefly, the gel bands were subjected to protease digestion and desalting, followed by liquid chromatography-mass spectrometry (LC-MS) detection. Finally, the results were identified through mass spectrometry database searching.

Studies on small-molecule binding sites. In vitro recombinant protein LSH was incubated with artemisitene at 30 °C for 2 h, separated by SDS-PAGE electrophoresis, followed by trypsin digestion, peptide extraction, and subsequent LC-MS/MS analysis.

5.15. Free fatty acid determination

Cells were treated with artemisitene (10 μM) for 24 h, the cell pellets were collected and analyzed by gas chromatography-mass spectrometry (GC-MS) (Personal Biotechnology, China). The content changes of palmitic acid (Blbio, ml306954H, China) and palmitoleic acid were measured using ELISA kits (Blbio, ml602364H, China).

5.16. Protein detection by western blotting and immunoprecipitation assay

Cells were collected, washed with PBS, and lysed on ice in RIPA buffer (Beyotime Biotechnology, P0013B, China) for 30 min. Cell lysates were centrifuged at 12,000g for 10 min at 4 °C, and the supernatant was collected. Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE, BioRad Laboratories, Hercules, CA), transferred to polyvinylidene fluoride (PVDF) (Merck, USA) membranes. The membranes were blocked in 5 % skim milk for 2 h and incubated with primary antibodies at 4 °C overnight. After washing three times with TBST, membranes were incubated with HRP-conjugated secondary antibodies (Proteintech, SA00001-2, USA). The antibodies used are as follows: GAPDH (ABclonal, A19056, China), LSH (ABclonal, A5831, China), CYP24A1 (Origene, TA368684S, USA), SCD (ABclonal, A26246, China), EWSR1 (ABclonal, A9640, China), FALG (Proteintech, 20543-1-AP, China), CAMKK2 (ABclonal, A9899, China), AMPK (ABclonal, A17290, China), P-AMPK (ABclonal, A1441, China), SREBF1 (ABclonal, A15586, China) and MYC (ABclonal, AE070, China). The results were visualized using Super ECL with the ChemiDOC™ XRS⁺ system (BioRad Laboratories, Hercules, CA).

For the Co-IP assay, cells were treated with artemisitene for 24 h and lysed with NP-40 buffer (Beyotime Biotechnology, P0013F, China). 1 mg of total protein was diluted to 300 μL. Then the primary antibody was added and shaken slowly at 4 °C. After 4 h, 20 μL Protein A/G magnetic beads (MedChemexpress, HY-K0202, USA) were added and incubated overnight. The bead-bound proteins were identified by mass spectrometry or western blotting.

5.17. Tumor xenograft

Experiments were conducted in accordance with the principles of the NIH Guide for the Care and Use of Laboratory Animals and approved protocols of the Institutional Animal Care and Use Committee (IACUC) of China Pharmaceutical University Experimental Animal Center. Female BALB/c nude mice were purchased from GemPharmatech (Nanjing, China).

HCT-116 cells (5 × 10⁷ in 0.15 mL PBS) were injected subcutaneously into the right flank of nude mice. Each experimental group included six mice, all of which were five weeks old at the start of the study. ATT administration was initiated once the tumor volume reached approximately 50 mm³, with the compound delivered intraperitoneally every two days for a total of eight doses. Body weight and tumor volume were recorded at two-day intervals throughout the treatment period. Mice were sacrificed two weeks after treatment initiation. Following euthanasia, tumors and major visceral organs were collected and fixed in 4 % paraformaldehyde for subsequent histological examination.

5.18. Statistical analysis

Statistical significance was assessed using unpaired two-tailed Student's t-test and one-way analysis of ANOVA followed by Dunnett's test for comparisons with the control group and Tukey's test for pairwise comparisons among all groups and P values were calculated by GraphPad Prism 9.0 (GraphPad Software Inc., San Diego, CA, USA). Quantifications were performed from at least three independent experiments, and statistical analyses were conducted using appropriate tests as indicated in the figure legends. Results were considered statistically significant when P < 0.05, whereas P > 0.05 was regarded as not significant.

CRediT authorship contribution statement

Ling Zhu: Data curation, Formal analysis, Investigation, Writing – original draft. Qimei Tan: Investigation. Yuxia Wang: Investigation. Lihong Hong: Investigation. Chen Chen: Supervision. Lingyi Kong: Supervision. Jianguang Luo: Supervision, Writing – review & editing.

Declaration of competing interest

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

Abbreviations

AMPK

AMP-activated protein kinase

ATT

Artemisitene

CaMKK2

Calcium/calmodulin-dependent protein kinase kinase 2

CETSA

Cellular thermal shift assay

CRC

Colorectal cancer

CYP24A1

Cytochrome P450 Family 24 Subfamily A Member 1

DARTS

Drug affinity responsive target stability

EWSR1

EWS RNA binding protein 1

HE

Hematoxylin and eosin

IHC

Immunohistochemistry

LSH

Lymphoid-specific helicase

NAC

N-Acetylcysteine

SCD

Stearoyl-CoA desaturase

SREBF1

Sterol regulatory element-binding transcription factor 1

SPR

Surface plasmon resonance

Appendix A. Supplementary data

The following is the Supplementary data to this article: