Multi-omics unravels multiple pharmacological actions in a murine model of infantile hemangioma receiving oxymatrine therapy

Abstract

Background

Approximately 10% of the patients with infantile hemangioma (IH) may exhibit resistance to propranolol (PRN) therapy, and thus alternative strategies are required. Our previous studies reported that oxymatrine (OMT) could inhibit the growth of hemangiomas, however the underlying pharmacological actions have not been fully addressed.

Methods

In this study, a murine IH model was constructed by implantation of EOMA cells into nude mice. OMT was administrated (50 mg/kg; i.p) for 21 days. Metabolic changes were examined by proteomics and metabolomics, followed by in vitro experimental validation using EOMA cells.

Results

OMT significantly suppressed the growth of hemangioma in vivo without significant adverse effects. A total of 869 differentially expressed proteins and 38 metabolites were identified. In addition to canonical apoptosis regulation, OMT also caused significant metabolic disturbances, particularly in purine and pyrimidine metabolism. Furthermore, ferroptosis may be involved in the therapeutic effect of OMT. In the validation experiments in vitro, we found that OMT dose-dependently reduced the viability of EOMA cells, concomitant with increased production of lipid reactive oxygen species (ROS) and Fe2 + accumulation.

Conclusions

In conclusion, these findings suggested that treatment with OMT could suppress the growth of hemangiomas through metabolic disturbances and inducing ferroptosis, which may provide new insights to the management of IH.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-026-04536-w.

Introduction

Infantile hemangioma (IH) is the most common benign vascular tumor of infancy, with a prevalence estimate of 5–10% in full-term newborns [1, 2]. The natural cycle of IH includes the fast proliferation phase, the plateau phase, and the slow involution phase [3]. However, currently there are no known biological markers capable of distinguishing between invasive and non-invasive vascular tumors. Consequently, it is imperative to decelerate the progression to preserve their benign characteristics, especially in cases where the surgical risks are considered high. Overall, most of the drug therapies for IH were designed either to block the vascular formation or to accelerate the involution process [4]. Propranolol (PRN) therapy has emerged as the first-line treatment for IH since 2008, which is considered safer and more effective than previously used medications such as corticosteroids, interferon, or vincristine [5]. However, approximately 10% of IH patients may exhibit resistance to PRN therapy, and 19% of IH patients may experience recurrence within the first year of life. Moreover, a growing body of research indicated that PRN treatment in infants may raise a variety of adverse events, including hypotension, hypoglycemia, and seizures [6, 7]. Therefore, alternative treatments for IH are required for those patients who are intolerable to PRN therapy or at a higher risk of complications.

Oxymatrine (OMT, C₁₅H₂₄N₂O₂, molecular weight: 264.360), is a quinolizidine alkaloid extracted from Sophora medicinal plants. Its structure is remarkably similar to matrine, which contains more than one oxygen atom at carbon C-1. Recently, OMT has been shown to have a wide range of beneficial pharmacological values, including anti-cancer, anti-diabetic, anti-virus, and anti-inflammation activities. Moreover, the anti-cancer mechanisms of OMT include the repression of proliferation and metastasis, induction of cell differentiation and cell-cycle arrest, promotion of apoptosis and ferroptosis, and inhibition of tumor angiogenesis. A variety of cancer cells have been shown sensitive to OMT treatment, including colorectal cancer, gall bladder carcinoma, and leukemia [8]. Nevertheless, the effect of OMT on IH has not been fully elucidated. Our previous studies demonstrated that OMT can inhibit the growth of hemangiomas [9, 10], but the underlying pharmacological actions remain unclear. In order to better understand the pharmacological actions of OMT therapy, this study aimed to determine the abundance of proteins and metabolites with proteomics and metabolomics. These findings highlighted OMT therapy that could exert anti-cancer effects in hemangiomas through multiple pharmacological actions, which may contribute to the development of alternative therapeutic options for patients with IH.

Materials and methods

Antibodies and reagents

Oxymatrine (Cat No. N1835) and the ferroptosis inhibitor Ferrostatin-1 (Fer-1, Cat No. A4371) were purchased from APExBIO (Boston, MA, USA). Mouse anti-GPX4 (Cat No. 67763-1-Ig) and Rabbit anti-ACSL4 (Cat. No. 22401-1-AP) were purchased from Proteintech (Rosemont, IL, USA). All other chemicals were purchased from Sigma‒Aldrich (St. Louis, MO, USA).

Cell culture and treatment

EOMA cells purchased from American Type Culture Collection (Manassas, VA, USA) were cultured in DMEM (Gibco, Gaithersburg, MD, USA) supplemented with 10% FBS, 100 mg/L penicillin, and 100 mg/mL streptomycin. The cells were treated with various concentrations of OMT (50 mM and 100 mM in culture medium) for 24–48 h. In rescue experiments, EOMA cells were pre-treated with Fer-1 (2 μM) for 1 h, followed by treatment with OMT (100 mM) for 48 h.

Cell viability

Cell viability was determined with a CCK-8 assay kit (Cat No. C0037; Beyotime, Shanghai, China). Approximately 2 × 10³ cells were cultured in 96-well culture plates for 24 h. After treatment, the cells were incubated with 10 μl of CCK-8 at 37 ℃ for 30 min. Absorbance was measured at 450 nm by a Multiskan Spectrum Thermo spectrophotometer.

Western blot analysis

The cells were lysed in RIPA lysis buffer on ice for 30 min (Cat No. P0013B, Shanghai, China). After centrifugation (10,000 g, 10 min), the supernatant was collected. The protein concentration was measured with a BCA protein assay kit (Thermo, USA). An equivalent amount of protein (50 μg) was subjected to 10% SDS-PAGE gel. The immunoblots were visualized by the chemiluminescent reagents, and were quantified by ChemiDoc XRS + system (Bio-Rad, Hercules, CA, USA).

Immunofluorescence (IF) staining

Cells were seeded in 12-well plates at 2 × 10⁴ cells/well. EOMA cells were fixed in 4% formaldehyde at room temperature for 15 min, followed by permeabilization with 0.1% TritonX-100 for 15 min. The cells were then incubated with primary antibodies against GPX4 (1:200) at 4 ℃ overnight. After 3 wishes with PBS, the cells were incubated with Alexa Fluor 488-conjugated secondary antibodies for 1 h at room temperature. Images were visualized using Leica DMI6000B fluorescence microscopy (Leica, Germany). Approximately 6–8 field/sample were captured for statistical analysis with LAS AF Lite image processing software (LAS AF Lite Version 4.2).

Assessment of mitochondrial membrane potential (MMP)

Cells were seeded in 12-well plates at 2 × 10⁴ cells/well. The MMP was determined with JC-1 probes (Cat No. C2006, Beyotime, Shanghai, China). MMP disruption was assessed using the ratio of aggregates to monomers. Images were visualized by a Leica DMI6000B microscope (Leica Microsystems, Wetzlar, Germany). Approximately 6–8 field/sample were captured for statistical analysis with LAS AF Lite image processing software (LAS AF Lite Version 4.2).

FerroOrange and Liperfluo staining

Cells were seeded in 12-well plates at 2 × 10⁴ cells/well. Lipid ROS formation and Fe²⁺ accumulation was detected with Liperfluo (Cat No. L248, DOJINGO, Japan) and FerroOrange (Cat No. F374, DOJINGO, Japan) according to the manufacturer’s instructions. Briefly, the cells were stained with FerroOrange (1 μM) or Liperfluo (10 μM) in medium at 37 ℃ for 30 min. After 3 washes with PBS, the images were visualized by a Leica DMI6000B microscope (Leica Microsystems, Wetzlar, Germany). Approximately 6–8 field/sample were captured for statistical analysis with LAS AF Lite image processing software (LAS AF Lite Version 4.2).

Allogeneic mismatched murine model for IH

BALB/c nude mice (female, 4-week-old) purchased from Shanghai Laboratory Animal Center (Shanghai, China), were maintained under standard conditions at 25 ± 0.5 ℃ with access to food and water ad libitum. After one-week acclimation, the mice were inoculated with EOMA cells (1 × 10⁷ suspended in 0.2 mL PBS) in the axilla. When tumors reached an average volume of 100 mm³, the mice were randomized into two groups (n = 4 mice/group). Group I (control), was treated with vehicle alone (0.9% saline, 100 µL; i.p.; 3 times/week), Group II was treated with OMT (50 mg/kg; i.p.; 3 times/week). Therapy continued for 21 days from randomization. Tumor volume was monitored every 5 days, calculated with the formula: V = L × W²/2 (V refers to volume, L refers to length, and W refers to width). After 21-day administration, the mice were sacrificed and the implants were removed. Blood vessels were counted in 5 fields/section, 2 sections/implant. Vessel density is expressed as vessels/mm². Hepatotoxicity indicators were determined, including alanine aminotransferase (ALT) and aspartate aminotransferase (AST). Nephrotoxicity indicators were determined, including blood urea nitrogen (BUN) and creatinine. Moreover, tumor samples were snap-frozen in liquid nitrogen and stored at − 80 ℃ until further use. The study was approved by the Xin Hua Hospital Animal Use Committee (XHEC-NSFC-2021–095). In accordance with ethical guidelines, all the maximal tumor burden did not exceed 1000 mm³.

Metabolomic profiles

After the addition of 500 μL precooled 80% methanol, the samples (50 mg) were homogenized by repeated freezing–thawing in liquid nitrogen. All chromatographic separations were performed using a Thermo Scientific UltiMate 3000 HPLC. Differential metabolites were defined as log2 fold change (FC) > 1 or < − 1, adjusted p value < 0.05, variable importance in the projection (VIP) > 1. Bioinformatics analyses were then performed by KEGG pathway classification (https://www.kegg.jp, accessed on12 July 2021). Gene-metabolite interaction network analysis was performed using MetaboAnalyst 6.0 (https://www.metaboanalyst.ca/MetaboAnalyst/ModuleView.xhtml).

Proteomic profiles

Tumor tissues (20 mg) were homogenized in 200 μL lysis buffer (4% SDS, 100 mM Dithiothreitol, 150 mM Tris–HCl pH 8.0). Samples were then incubated with trypsin (protein: enzyme = 50:1) for 14–16 h at 37 ℃. Separation was performed by Thermo Scientific EASY-nLC™ 1200 system at a flow rate of 300 nL/min. Partial least squares-discriminant analysis (PLS-DA) was conducted to discriminate the different variables between groups. Differentially expressed proteins (DEPs) were defined as log2 FC > 1 or < -1, adjusted p value < 0.05 (Benjamini–Hochberg correction). Bioinformatics analyses were then performed by KEGG pathway classification (https://www.kegg.jp, accessed on 12 July 2021), GO-term classification (david.abcc.ncifcrf.gov, accessed on 1 December 2021) and protein–protein interaction (PPI) networks (https://www.string-db.org). GSEA analysis was performed using the Java desktop software (http://software.broadinstitute.org/gsea/index.jsp).

Statistical analysis

Data are expressed as means ± SD, statistically analyzed and plotted using GraphPad Prism 9.0 (GraphPad Software Inc., San Diego, CA, USA). Variables were analyzed by Student’s t-test for two groups. p values < 0.05 were considered statistically significant.

Results

OMT suppressed the growth of hemangioma in vivo without significant adverse effects

As illustrated in the schematic diagram, the mice with visible hemangiomas (14 days post implanting) were randomly divided into Saline group or OMT group (Fig. 1A). Four mice were excluded from the final analysis because their tumor volumes exceeded 1000 mm³ (two from the control group and two from the OMT group), resulting in a final sample size of four mice per group for statistical comparison. Representative images of hemangioma implants showed significant reduction in the tumor volume from OMT-treated mice (Fig. 1B). When the tumors were removed on day 35, the average tumor volumes were 435.3 mm³ for Saline, and 192.8 mm³ for OMT, respectively (Fig. 1C). The average tumor weight was 599.1 mg for Saline, and 201.3 mg for OMT, respectively (Fig. 1D). Moreover, H&E staining revealed a trend toward less vessel formation in the implants from OMT group. The area of micro-vessels decreased by 69.6% upon OMT treatment (Fig. 1E). Conversely, no significant differences were found between groups in body weight (Fig. 1F), hepatotoxicity indicators (Fig. 1G) and nephrotoxicity indicators (Fig. 1H). Collectively, these results suggested that administration of OMT significantly suppressed the growth of hemangioma in vivo without significant adverse effects.

Fig. 1.

OMT suppressed the growth of hemangioma in vivo without significant adverse events. A Schematic diagram of the murine IH model receiving OMT therapy. B Images of implants (n = 4/group) C Time course of tumor volume. Data were presented as mean ± SD (n = 4/group). **, p < 0.01 D Tumor weight of removed implants. Data were presented as mean ± SD (n = 4/group). ***, p < 0.001. E Representative H&E staining (a). Scale bars = 50 μm. Quantification of vessel area (b). Data were presented as mean ± SD (n = 4/group). **, p < 0.01. F Time course of body weight. Data were presented as mean ± SD (n = 4/group) G Hepatotoxicity indicators (serum levels of ALT and AST). Data were presented as mean ± SD (n = 4/group). H Nephrotoxicity indicators (serum levels of creatinine and BUN). Data were presented as mean ± SD (n = 4/group). ALT, Alanine aminotransferase; AST, Aspartate aminotransferase; BUN, Blood urea nitrogen

Proteomic profiles in the OMT-treated IH mice

The PLS-DA showed obvious separation of the identified proteins between groups (Figure S1). As indicated in volcano plots, a total of 340 down-regulated and 529 up-regulated DEPs were identified (Fig. 2A). In the GO-term enrichment analysis, the top 15 cluster annotations of Biological Process (BP) were presented, including “negative regulation of apoptotic process”, “proteolysis”, “mRNA processing”, “response to xenobiotic stimulus”, “translation” (Fig. 2B). The KEGG scatter plot indicated that most of the DEPs were enriched in “metabolic pathway”, “proteasome”, “purine metabolism”, “carbon metabolism”, “biosynthesis of amino acids” (Fig. 2C). Given the central role of apoptosis in mediating the therapeutic effect of OMT, apoptosis-related DEPs were presented. Notably, the proteins involved in “negative regulation of apoptotic process” significantly decreased in OMT group, including TRFC, ARB1, BCL2L1; Conversely, the proteins involved in “positive regulation of apoptotic process” significantly increased in OMT group, including RNPS1, DEGS1, AIFM1 (Fig. 2D). Furthermore, top 100 up-regulated DEPs were imported for PPI analysis, and 25 hub genes were identified, including ACTN2/3 (Fig. 3A). Enrichment analysis suggested that the interactions among top 100 up-regulated DEPs were mainly classified in the muscle contraction (Fig. 3B). As shown in the heatmap, the proteins involved in “muscle contraction” significantly increased in OMT group, including MYBPC2, ACTN3, TNNC2, along with the pericyte markers (αSMA, NG2) (Fig. 3C). We thus assumed that OMT might be able to accelerate onset of involution by targeting pericyte contractility. Top 100 down-regulated DEPs were imported for PPI analysis as well, and 25 hub genes were identified, including EPB42, NT5C3 (Fig. 3D). Enrichment analysis suggested that the interactions among top 100 down-regulated DEPs were mainly classified in the metabolism of purine and pyrimidine (Fig. 3E). As shown in the heatmap, most of the proteins involved in purine or pyrimidine metabolism decreased significantly in OMT group, including TYMS, NT5C2, PRPS1L3 (Fig. 3F). We thus assumed that suppressed growth of hemangioma upon OMT might be associated with disrupted metabolism of purine and pyrimidine. Taken together, these results suggested that the effect of OMT on hemangioma cells might be associated with multiple pathways, including apoptosis regulation, accelerating involution, and disrupting nucleotide homeostasis.

Fig. 2.

Proteomic profiles in the OMT-treated IH mice. A Volcano plots showing DEPs. B GO term enrichment analysis of DEPs. C KEGG pathway enrichment analysis of DEPs. D Heatmap of the key DEPs involved in apoptosis regulation

Fig. 3.

Protein–protein interaction analysis in the OMT-treated IH mice. A Top 25 hub genes in the PPI analysis of top 100 up-regulated DEPs (Generated by Cytoscape 3.10 with cytohubba using MCC method). B Enrichment strength evaluation of PPI among the top 100 up-regulated DEPs. The number at the top of the bar represents the quantity of proteins involved. C Heatmap of differential proteins involved in muscle contraction. D Top 25 hub genes in the PPI analysis of top 100 down-regulated DEPs (Generated by Cytoscape 3.10 with cytohubba using MCC method). E Enrichment strength evaluation of PPI among the top 100 down-regulated DEPs. The number at the top of the bar represents the quantity of proteins involved. F Heatmap of differential proteins involved in purine and pyrimidine metabolism

Metabolomic profiles in the OMT-treated IH mice

As indicated in volcano plots, a total of 18 down-regulated and 20 up-regulated metabolites were identified (Fig. 4A). These metabolites belong to “Lipids and lipid-like molecules” (n = 16), “Nucleosides, nucleotides, and analogues” (n = 5), “Organic acids and derivatives” (n = 9) and other (n = 8) (Fig. 4B). KEGG pathway enrichment analysis revealed that most of the differential metabolites could be clustered into “cGMP-PKG signaling pathway”, “mTOR signaling pathway”, “FoxO signaling pathway” and “PI3K-Akt signaling pathway” in addition to “Pyrimidine metabolism” and “Purine metabolism” (Fig. 4C). The top 20 differential metabolites were presented in the heatmap, including Adenosine 5'-monophosphate (AMP), Guanosine 5'-monophosphate (GMP), S-Adenosyl-L-methionine (SAM) and S-Adenosyl-L-homocysteine (SAH), two of the key molecules for carbon transfer (Fig. 4D). Moreover, identification of gene-metabolite interaction network suggested that disrupted metabolic homeostasis were largely attributed to nucleotides and analogues (e.g., SAH, AMP, GMP), as well as genes involved in pyrimidine and purine metabolism (e.g., NT5C2/3) (Fig. 5). An overview of dysregulated pyrimidine and purine metabolic process plotted in the KEGG map “PYRIMIDINE METABOLISM” (Figure S2) and “PURINE METABOLISM” (Figure S3). Taken together, these results indicated that disrupted homeostasis of pyrimidine and purine metabolism may play a central role in mediating the therapeutic effect of OMT therapy.

Fig. 4.

Metabolomic profiles in the OMT-treated IH mice. A Volcano plots showing differential metabolites. B Classification of identified differential metabolites. C KEGG pathway enrichment analysis of differential metabolites. D Hierarchical clustering of top 20 differential metabolites

Fig. 5.

Combined analysis of proteomics and metabolomics. Gene-metabolite interaction network analysis (MetaboAnalyst 6.0)

Identification and experimental validation of ferroptosis in response to OMT therapy

GSEA analysis suggested that OMT treatment significantly altered the expression pattern of the genes involved in “iron ion homeostasis” and “positive regulation of reactive oxygen species metabolic process”, two of the hallmarks of ferroptosis (Fig. 6A). Among the ferroptosis-related genes, the expression of GCLC, GCLM, GPX4, GSS and SLC7A5 decreased significantly upon OMT treatment, whereas HMOX1, LPCAT3 and PRNP substantially increased (Fig. 6B). In addition to the in vivo findings, the role of ferroptosis in mediating the therapeutic effect of OMT therapy was also experimentally validated in vitro. As shown, OMT treatment dose-dependently reduced the viability of EOMA cells (Fig. 6C), as well as the antioxidant capacity of glutathione (Fig. 6D). Moreover, the protein levels of GPX4 were substantially decreased, while the protein levels of ACSL4 were significantly increased upon OMT treatment (Fig. 6E). Consistently, Liperfluo and FerroOrange staining demonstrated that the lipid ROS formation and Fe²⁺ accumulation was substantially promoted by OMT treatment in a dose-dependent manner (Fig. 6F). In the rescue experiments, reduced cell viability upon OMT treatment was evidently alleviated by Fer-1 (Fig. 7A). Moreover, the results of western blot suggested that OMT-induced loss of and GPX4 was remarkably abrogated by ferroptosis antagonism with Fer-1. Nevertheless, no significant changes were found in ACSL4 by Fer-1 (Fig. 7B). Consistently, the results of IF staining basically mirrored the changes in GPX4 expression upon OMT and Fer-1 treatment (Fig. 7C). Furthermore, JC-1 aggregates indicating normal MMP was significantly reduced by OMT treatment, while JC-1 monomers indicating disrupted MMP was substantially elevated. Nevertheless, reduced ratio of aggregates/monomers was evidently attenuated by Fer-1, suggesting that OMT-induced mitochondrial dysfunction was dependent on ferroptosis-related pathways (Fig. 7D). Collectively, these results validated that ferroptosis may play a fundamental role in mediating the therapeutic effect of OMT therapy for hemangiomas.

Fig. 6.

Identification of ferroptosis-like responses upon OMT therapy. A GSEA analysis (Iron ion homeostasis/Positive regulation of reactive oxygen species metabolic process). B Heatmap of ferroptosis-related genes. C Cell viability. EOMA cells were treated with various doses of OMT for 24–48 h. Data were presented as mean ± SD of biological triplicates. *, p < 0.05, **, p < 0.01 compared to control (0 mM). D GSH/GSSG ratio. EOMA cells were treated with various doses of OMT for 48 h. Data were presented as mean ± SD of biological triplicates. **, p < 0.01. E Representative immunoblots (a) and relative folds (b) of ferroptosis-related proteins. EOMA cells were treated as described above. Data were presented as mean ± SD of biological triplicates. *, p < 0.05, **, p < 0.01. F Liperfluo and FerroOrange staining for intracellular lipid ROS and Fe²⁺ in EOMA cells. EOMA cells were treated as described above. Scale bar = 25 μm. Data were presented as mean ± SD (n = 6–8 fields/sample). *, p < 0.05, **, p < 0.01. Three independent experiments were performed that showed similar results

Fig. 7.

Experimental validation of OMT-induced ferroptosis in vitro. A Cell viability. EOMA cells were pre-treated with the ferroptosis inhibitor ferrostatin-1 (Fer-1, 2 uM) for 1 h, followed by treatment with OMT (100 mM) for 48 h. Data were presented as mean ± SD of biological triplicates. *, p < 0.05, **, p < 0.01. B Representative immunoblots (a) and relative folds (b) of GPX4 and ACSL4. Data were presented as mean ± SD of biological triplicates. *, p < 0.05. C Representative staining (a) and fluorescence intensity (b) of GPX4. EOMA cells were treated as described above. Scale bar = 50 um. Data were presented as mean ± SD (n = 6–8 fields/sample). *, p < 0.05. Three independent experiments were performed that showed similar results. D JC-1 assay (a) and the ratio of aggregates/monomers (b). EOMA cells were treated as described above. Mitochondrial membrane potential was determined by JC-1 monomers (green) and aggregates (red). Scale bar = 100 um. Data were presented as mean ± SD (n = 6–8 fields/sample). Three independent experiments were performed that showed similar results

Discussion

Risk factors for IH include preterm birth, low birth weight, and multiple gestations. A study by Drolet et al. found a 40% increased risk of IH for every 500 g decrease in birth weight [11]. Other risk factors include progesterone use, preeclampsia, older maternal age, placenta previa, and in vitro fertilization [12]. Although the precise pathogenesis of IH remains incompletely understood, recent research has identified multiple signaling pathways, such as hypoxic signaling (e.g., HIF1a pathway) in the early phase of IH development, angiogenesis signaling (e.g., VEGFA/VEGFR axis) cascades for endothelial cell proliferation [13]. As reported, approximately 10% of IH cases show resistance to PRN therapy, and 10–15% of patients are susceptible to recurrences after discontinuation of PRN therapy [14]. Under these circumstances, alternative therapeutic options could be considered, including intralesional pingyangmycin, intravenous vincristine, and oral steroids [15]. Additionally, a few pre-clinical studies have suggested the potential medicinal value of phytochemicals in treating IH. For example, Cai et al.'s study demonstrated that 15,16-dihydrotanshinone I, an active component of S. miltiorrhiza, can significantly inhibit the growth of hemangioma through pro-apoptotic and anti-angiogenic pathways [16].

One aspect requires comments here regarding the dosages of OMT used in this study. This study used a relatively high dose of OMT (50 mg/kg) in vivo, because this dosage was adopted to maintain consistency with our previously established IH model receiving OMT therapy [10]. Moreover, this experimental design allowed us to investigate whether high-dose OMT would induce hepatic or renal toxicity, thereby providing preliminary toxicological data on the safety profile of OMT therapy. As for the in vitro experiments, a high dose of OMT (50–100 mM) was used in the EOMA cells. In previous studies, we determined the appropriate dosage of OMT in patient-derived primary hemangioma-derived endothelial cells (HDECs), with a maximum concentration of 2 mg/ml (approximately 10 mM) [9]. However, the present study utilized EOMA cells, which are murine hemangioendothelioma cells that represent a higher grade of malignancy and may exhibit greater resistance to conventional chemotherapeutic agents. In order to achieve better inhibitory efficacy, we applied a higher dose of OMT for EOMA cells in this study.

One strength of this study is the use of proteomics instead of transcriptomics. To date, although some previous studies using transcriptomic profiling have been reported on the pathogenesis of IH as well as the pharmacological actions of PRN, currently there are no published studies on IH using proteomic profiling. Given the close relationship between IH and autophagy that can predominantly regulate gene expression at the protein level (e.g., autophagic degradation), we considered that proteomics could provide more unique information [17]. OMT is a quinolizidine alkaloid that may be a promising compound as a novel anti-cancer agent. Numerous studies have uncovered the diverse ways by which OMT can abrogate tumorigenesis, including triggering apoptosis, disturbing metabolism, and suppressing epithelial-to-mesenchymal transition [18, 19]. In this study, numerous genes involved in apoptosis regulation altered significantly upon OMT treatment, suggesting that apoptosis may play a central role in mediating the therapeutic effect of OMT on hemangioma cells.

In addition to apoptosis regulation, the anti-cancer effect of OMT might be attributed to disrupted metabolism of purine and pyrimidine as well. Compared to normal cells, cancer cells exhibit elevated levels of purines and enzymes involved in the de novo purine biosynthetic pathway to sustain rapid proliferation. In this study, we found significantly reduced abundance of AMP, one of the most important purine derivatives to cancer biology. Reduced AMP levels might mediate the anti-cancer effect of OMT through multiple pathways. (1) Energy metabolism: AMP is involved in energy metabolism pathways, particularly as a component of adenosine triphosphate (ATP) which serves as a primary energy currency in cells. (2) Nucleotide synthesis: Cancer cells have high demands for nucleotides to support DNA and RNA synthesis, and thus AMP availability is crucial for sustaining these processes in rapidly dividing cancer cells. (3) Signaling transduction: AMP also serves as a precursor for the synthesis of cyclic adenosine monophosphate (cAMP), an important second messenger molecule involved in various cellular signaling pathways. Previous studies have proven that cAMP signaling is involved in the proliferation of hemangioma-derived endothelial cells that express β-adrenergic receptors (β-ARs). Notably, PRN is a nonselective β-AR antagonist that can lower cAMP levels, thereby suppressing proliferation and promoting apoptosis [20, 21]. These findings may have potential clinical significance, as OMT can exert similar effects to PRN by reducing cAMP levels. However, given no evidence indicating that OMT can block β-ARs, OMT therapy might avoid a series of adverse events caused by the inhibition of β-AR signaling (e.g., central nervous system, or gastrointestinal disorders). Additionally, aberrant metabolism of amino acids and one-carbon units (e.g. SAM, SAH) in cancer cells could play fundamental roles in sustaining proliferative signaling, drug resistance, and invasion and metastasis [22]. In this study, we also noted significantly reduced abundance of SAM and SAH, which might be an alternative explanation for the underlying mechanism of OMT therapy against IH.

Progression of IH slows dramatically in the “involution phase”, which typically lasts for four to six years as fibro-fatty tissue replaces the cellular elements [23]. Pericytes are contractile cells that wrap around endothelial cells in blood vessels, regulating vessel stability and function [24]. During involution, pericytes can actively participate in the regression of blood vessels by inducing vessel constriction and subsequent vessel collapse [25]. Lee et al.’s study demonstrated that the pharmacological actions of PRN include increased pericyte contractility, suggesting that pericytes may serve as a key therapeutic target for accelerating IH involution [26]. In this study, along with increased abundance of pericyte markers (αSMA, NG2), we also found significantly increased expression of proteins involved in muscle contraction (e.g. MYBPC2, ACTN3, TNNC2). These results suggest two possibilities: first, OMT may elicit similar effect to PRN that promotes IH involution by targeting pericyte contractility; alternatively, it is possible that OMT may recruit more stromal cells rather than pericytes themselves. Nevertheless, the interaction between pericytes and endothelial cells in IH is complex and dynamic, and thus further studies are required for understanding the molecular mechanisms by which OMT accelerates IH involution.

Ferroptosis is a new form of regulated cell death characterized by iron dependency and the accumulation of lipid peroxides, leading to oxidative damage and cell membrane rupture [27]. It involves metabolic pathways related to iron, lipids, and amino acids, contributing to the production of lipid ROS [28, 29]. Accumulating evidence has revealed the potential of ferroptosis in cancer treatment owing to its ability to target therapy-resistant tumors, making it a potential therapeutic avenue in oncology. For instance, Zheng’s group reported that utilizing of LGR4-mAb augments ferroptosis when co-administrated with chemotherapeutic agents, demonstrating a potential opportunity to specifically eliminate drug-resistant colorectal cancer cells [30]. Moreover, ferroptosis also influences the tumor immune microenvironment. For instance, Dai et al. reported that depletion of Gpx4 in pancreatic adenocarcinomas promotes oxidative DNA damage and the release of 8-OHG via TMEM173/STING pathway, leading to macrophage M2 polarization and expression of immunosuppressive signals [31]. In recent years, ferroptosis-related strategies have emerged as a promising direction in the therapeutic landscape for IH. Liu et al. found that hemangioma cells could promote macrophage ferroptotic resistance through upregulating expression of GPX4, which is required for the progression of IH [32]. Moreover, Geng et al. developed a novel therapeutic approach to hemangiomas, and found that the combination of photothermal therapy and ferroptosis therapy exhibited a superior synergistic effect in damaging vascular structures [33]. In this study, we found that OMT could dose-dependently reduce the expression of GPX4, and promote the production of ROS and Fe²⁺ accumulation. Moreover, ferroptosis inhibitor Fer-1 could reverse the loss of GPX4, and significantly alleviate the toxic effect on EOMA cells. This present study was first to demonstrate that OMT could inhibit IH in a ferroptosis-depend way in vitro, and thus OMT may serve as a promising alternative option beyond PRN for treating IH.

Nevertheless, a limitation of this study should be noted that only one cell line (EOMA) was utilized throughout the experimental design. According to Kong et al.’s review, IH models are primarily categorized into two classes according to their cellular origin: models originating from hemangioma stem cells (HemSCs) and models originating from hemangioma endothelial cells (HemECs) [34]. The advantages of the HemEC model include rapid proliferation (ability to grow rapidly after implantation into nude mice) and similar pathophysiological characteristics (ability to generate tumors with similar morphology, physiology and histopathology to human IH). In contrast, the HemSC models possess multidirectional differentiation capacity, which can provide more cellular-level insights into the pathogenesis and biological behaviors of IH. EOMA cells are HemEC that lack stem-cell-like characteristics, and thus cannot fully represent human IH biology.

Conclusion

In conclusion, this study identified multiple pharmacological actions in a murine model of IH receiving OMT therapy. First, apoptosis regulation was involved in the OMT therapy. Next, OMT could disturb purine and pyrimidine metabolism, reduce the concentration of key purine and pyrimidine derivatives. Furthermore, OMT could inhibit IH in a ferroptosis-depend way in vitro. Consequently, these findings suggested that treatment with OMT can suppress the growth of hemangiomas in nude mice through multiple mechanisms, which may provide new insights to the management of IH.

Author contributions

Data curation, Formal analysis, Investigation, Methodology, Yang Wang and Jingyu Peng; Investigation, Mingke Qiu; Investigation, Yuxin Dai; Investigation, Shuqing Wang; Conceptualization, Funding acquisition, Supervision, Jingmin Ou; Conceptualization, Data curation, Formal analysis, Supervision, Validation, Writing—original draft, Writing—review & editing, Junkai Yan.

Funding

This work was supported by the National Natural Science Foundation of China (82170402 and 82370402).

Availability of data and materials

The data presented in this study are available upon request from the corresponding author. The data are not publicly available due to privacy concerns.

Declarations

Ethics approval and consent to participate

The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Xin Hua Hospital Animal Use Committee (XHEC-NSFC-2021-095). Animal studies were strictly complied with the 3Rs (Replacement, Reduction, Refinement). In accordance with ethical guidelines, the maximal tumor burden was limited to ≤ 1000 mm³. Not applicable as this study did not involve human participants.

Consent for publication

All authors have read the manuscript and approved the final version.

Competing interests

The authors declare no competing interests.