Skip to main content
NCBI home page
Discover Oncology logoLink to Discover Oncology
. 2026 Mar 10;17:578. doi: 10.1007/s12672-026-04747-1

Integrated multi-target pharmacology of ginseng in acute myeloid leukemia through single-cell sequencing, molecular docking, network pharmacology, and in vitro experiments

Yan Zhang 1,#, Bo-bin Su 1,#, Chang-qing Jiao 1, Zhong-hui Wu 1, Jun-jie Zhou 1,✉, Jian Ge 1,2,✉
PMCID: PMC13086997  PMID: 41806205

Abstract

Background

Acute myeloid leukemia (AML) is a heterogeneous disease of the hematological system. It is characterized by the rapid clonal expansion of abnormally differentiated myeloid progenitor cells within a complex microenvironment. Ginseng has shown anticancer effects in AML. However, the function and mechanism of the major active component of ginseng in AML still remain to be elucidated. This study aims to explore the potential therapeutic targets and molecular mechanisms of ginseng in the treatment of AML.

Methods

We analyzed 10 AML patients treated with the uniform chemotherapy regimen using single-cell RNA sequencing (scRNA-seq). Then, we identified potential targets of ginseng and AML through a combination of network pharmacology and experimental validation using public databases like TCMSP and CTD. After that, we performed protein–protein interaction, gene ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses to uncover the underlying mechanisms of ginseng in AML. Ultimately, we verified the network pharmacology results via in vitro experiments.

Results

Our results indicated that in non-complete remission (non-CR) AML patients, ginseng comprises 14 anti-AML active compounds and 34 related target genes. KEGG pathway analysis revealed that the AGE-RAGE and NF-κB signaling pathways may play a critical role in modulating chemoresistance in AML treatment. Through network pharmacology and molecular docking, we identified several active compounds in ginseng, including ginsenoside-Rh4, gomisin B, and panaxadiol, which demonstrated strong binding affinities toward core AML targets. In vitro experiments demonstrated that the combination of panaxadiol and Ara-C exhibited synergistic anti-leukemic effects. Further analysis indicated that panaxadiol potentiates the anti-leukemic activity of Ara-C.

Conclusion

This study has elucidated the active compounds, potential targets, and signaling pathways of ginseng in overcoming AML chemoresistance. It offers new insights into the molecular mechanisms through which ginseng acts as a potential chemosensitizer of AML and lays a foundation for future research in this field.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-026-04747-1.

Keywords: Ginseng, Acute myeloid leukemia, Network pharmacology, Chemotherapy resistance, Panaxadiol

Introduction

Acute myeloid leukemia (AML) is a malignancy arising from clonal expansion of abnormal hematopoietic stem or progenitor cells, leading to impaired blood cell development, bone marrow failure, and eventually fatal outcomes [1]. Over the past 32 years, the global burden of AML has increased substantially, driven by rising incidence and mortality rates. Specifically, the global incidence of AML has increased steadily, rising from 79,372 cases in 1990 to 144,645 cases in 2021, according to the study by Zhou Y et al. [2]. Although initial induction therapy may induce remission in some patients, relapse is common, highlighting the urgent need for more effective and safer therapeutic strategies [3, 4].

The tumor microenvironment (TME) is a complex ecosystem comprising diverse immune cell populations, cancer-associated fibroblasts, endothelial cells, pericytes, and other resident cell types within the tumor tissue [5]. Tumor progression may be influenced by the ratio of pro- to anti-tumor inflammatory mediators [6, 7]. Recent studies have demonstrated that elements of the TME are critical for tumor development and metastasis. The immune system's ability to eliminate cancer cells can be impaired by processes such as neoantigen immunoediting and the accumulation of immunosuppressive myeloid cell populations [8]. Among the diverse immune cells in the TME, monocytes have emerged as key regulators of cancer development and progression [9]. Recent single-cell RNA sequencing (scRNA-seq) studies have demonstrated that monocytes play both pro-tumoral and anti-tumoral roles in cancer. Additionally, monocytes support angiogenesis, extracellular matrix (ECM) remodeling, immunosuppression, and the intravasation of cancer cells [5, 10, 11]. Accumulating evidence indicates that intra-tumor heterogeneity (ITH) within specific cellular subpopulations is a key contributor to drug resistance and disease relapse in AML [12, 13]. Recent years, much research has concentrated on T cells and the role of their immune checkpoint proteins [14–16]. The molecular determinants that regulate monocyte differentiation and function in AML are still poorly understood. Moreover, a comprehensive molecular investigation of monocytes in AML remains limited. Notably, the active compounds in Traditional Chinese Medicine (TCM) have recently emerged as a novel therapeutic approach for treating AML [17, 18].

A prominent herb in TCM is ginseng [19]. Ginseng contains a wide range of bioactive compounds, including alkaloids, phenols, polysaccharides, polyacetylenes, and saponins [20]. Ginseng exhibits a broad spectrum of biological activities, including regulation and strengthening of visceral functions, reinforcement of vascular function, and enhancement of vitality [21, 22]. It has been demonstrated that ginseng can downregulate microphthalmia-associated transcription factor of melanogenesis and melanoma growth [23]. Exosome-like nanoparticles produced from ginseng actively penetrate the blood–brain barrier and regulate the tumor microenvironment to produce anti-glioblastoma effects [24]. Compound K can induce apoptosis by increasing DNA damage in AML cells [25]. However, the mechanisms by which other ginseng-derived compounds exert therapeutic effects in AML remain unclear and require further investigation. In this study, we addressed this gap by integrating single-cell sequencing with network pharmacology.

ScRNA-seq provides a robust approach for profiling distinct cellular populations within complex tissues. By capturing the transcriptional landscape at single-cell resolution, it enables in-depth exploration of both cancer cell heterogeneity and their interactions with the surrounding tumor microenvironment, thereby elucidating mechanisms of drug resistance and informing personalized therapeutic strategies [26, 27]. In this study, we employed scRNA-seq to construct a transcriptomic map of AML samples, uncovering a range of cellular subtypes. This technique has proven especially valuable in characterizing the cellular composition and functional diversity within the TME [28, 29]. In order to analyze the molecular features of tumor-infiltrating monocytes and determine the monocyte marker genes, we conducted a thorough scRNA-seq of AML in this work. Furthermore, we integrated these findings with network pharmacology, a powerful methodology for deciphering novel drug targets and the systemic biological mechanisms of natural compounds [30].

In the present study, we integrated network pharmacology, scRNA-seq-based transcriptomic analysis, and molecular docking to elucidate the pharmacological mechanisms underlying the efficacy of ginseng in the treatment of AML.

Data acquisition

ScRNA-seq data corresponding to AML cohort GSE240618 was retrieved from the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) repository (http://www.ncbi.nlm.nih.gov/geo/). Target information for these ingredients was predicted using both the Comparative Toxicogenomics database (https://ctdbase.org) and SwissTargetPrediction (http://swisstargetprediction.ch/) database. In the present study, a total of 10 AML patients were enrolled in analysis. AML sample data were transformed into a Seurat-compatible object via the “CreateSeuratObject” function embedded within the Seurat package (implemented in R version 4.3.2).

scRNA-seq data integration and cell type identification

The top 2000 highly variable genes from the normalized expression matrix were used to conduct the principal component analysis (PCA). AML cell cluster analysis was performed using “FindNeighbors” and “FindCluster” functions. Meanwhile, we employed the SingleR package to annotate predicted cells for cell type, integrating classical marker genes and their reported functions from prior studies. To examine the features of each cellular subpopulation, we screened differentially expressed genes (DEGs) via the "FindAllMarkers" function.

Drug component screening and compound-target network

The compounds of “Ginseng” were collected from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, https://www.tcmsp-e.com/load_intro.php?id=43). Initially, a total of 190 ginseng-related compounds were retrieved from the database without any filtering. Subsequently, compounds were screened based on pharmacokinetic parameters, including oral bioavailability (OB) ≥ 30% and drug-likeness (DL) ≥ 0.18, to identify the potential bioactive compounds for further analysis. To illustrate the interactions between ginseng compounds and AML, a compound-target interaction network was generated using Cytoscape software (version 3.10.1, https://cytoscape.org/).

Pathway enrichment analysis

GO and KEGG pathway enrichment analyses were performed using the DAVID database (Database for Annotation, Visualization and Integrated Discovery) (http://david.ncifcrf.gov/). GO enrichment analysis, covering biological process (BP), cellular component (CC), and molecular function (MF), along with KEGG pathway enrichment analysis, was performed on the key targets. Genes for further investigation were selected based on a significance threshold of P < 0.05, including the top 6 enriched terms from the GO analysis and the top 10 enriched pathways from the KEGG analysis.

Molecular docking verification

To investigate the binding interactions between ginseng's bioactive compounds and the core target genes identified in this study, molecular docking techniques were applied. Structural data were collected from the RCSB Protein Data Bank (https://www.rcsb.org/) and the TCMSP database ( https://www.tcmsp-e.com/load_intro.php?id=43). AutoDockTools (version 4.2) (http://autodock.scripps.edu/) was used to predict binding conformations, and the resulting molecular interactions were visualized using PyMOL software (version 4.6.0.)

Reagents and antibodies

Panaxadiol (T2763) and Ara-C (T1272) were purchased from Taozhu (China). The antibodies against BAX (#5023), phospho-NF-κB p65 (#3033), and NF-κB p65 (#8242) were purchased from Cell Signaling Technology (CST). The antibody against BCL-2 (AB32124) was purchased from Abcam (UK). The antibody against caspase-3 (82202–1-RR) was purchased from Proteintech (Wuhan, China). The antibody against β-actin (AC004) was purchased from Abbkine Biotechnology Co., Ltd. (China). Dilution ratio of antibodies was carried out according to the instructions.

Cell lines and cell culture

The THP-1 and NOMO-1 cell lines were obtained from the Shanghai Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Both cell lines were routinely maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. The cells were cultured in a humidified atmosphere at 37 °C with 5% CO₂.

Western blot

Protein lysates were extracted from cells according to standard procedures, and Western blotting was carried out as outlined in previous studies [31]. Protein bands were visualized and captured with the LiCOR infrared imaging system (LI-COR Biosciences, Lincoln, NE, USA), and the intensity of these signals was analyzed using ImageJ software (NIH, Bethesda, MD, USA).

RNA extraction and real-time PCR

Total RNA was extracted using TRIzol reagent (Invitrogen, USA) from cultured cells in accordance with the manufacturer's instructions. As previously mentioned [31], quantitative real-time PCR (qRT-PCR) was carried out utilizing a CFX96 real-time RT-PCR detection system (Bio-Rad, USA). The primer sequences are provided in Supplemental Table 1.

Cell apoptosis

Cell apoptosis was detected using Annexin V-FITC and PI double staining (Vazyme) and analyzed by flow cytometry (CytoFLEX, BECKMAN).

Statistical analysis

Statistical analysis was carried out using GraphPad Prism 8.0 and R software (version 4.2.2). Student’s t-test or one-way ANOVA was used based on data distribution and experimental design. The significance levels were set as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.

Results

Cellular landscape of AML

A total of 10 AML single-cell samples were included in this study from GSE240618 dataset. After the removal of low-quality cells, normalization, integration, and PCA, 24,251 cells were divided into 18 clusters (Fig. 1A). Then, based on the single-cell sequencing results and previous literature, by annotating these subclusters for analysis, we identified 9 cell types (Fig. 1B), according to the expression of canonical markers (Fig. 1C, D). Meanwhile, Fig. 1C shows the important marker genes for each cell type. The samples were split into two groups: six samples from AML patients were in the CR group, and four samples from AML patients were in the non-CR group. To ascertain the features of the illness, these groups were examined. The CR and non-CR groups' UMAP results agreed with the findings of the comprehensive analysis (Fig. 1E, G). To prevent comparison mistakes caused by variations in the number of cells between the CR and non-CR groups, the proportion of cells in each group was shown (Fig. 1F).

Fig. 1.

Fig. 1

Open in a new tab

Cellular landscape of AML. A UMAP plot colored by various cell clusters; B UMAP plot colored by cell types after annotation; C Dot plot of selected marker genes for each cell type. Dot size represents the percentage of cells expressing each gene, while the dot color represents the level of expression; D Feature plots of selected marker genes for each cell type. The color scale shows the log1p-normalized expression levels of the genes; E UMAP plots of the CR group (n = 6) and non-CR group (n = 4); F Bar plot depicting the proportions of the nine main cell types in the CR and non-CR groups; G UMAP plots of individual samples.

Heterogeneity of monocyte phenotypes and functions between the CR and non-CR groups

Based on the subpopulation, we observed disease-associated alterations in the monocytes. After removing low-quality cells, performing normalization, integration, and PCA, the monocytes detected in this study were further classified into 12 subclusters after dimensionality reduction (Fig. 2A), and gene expression patterns in different cell subgroups were demonstrated (Fig. 2B, C). According to the expression of classical markers, we identified 12 cell types (Fig. 2D). To enable a clearer intergroup comparison, we performed single-cell subcluster analysis on CR and non-CR samples separately. Based on canonical cell markers, the CR group was partitioned into 7 cell clusters, while the non-CR group was divided into 9 cell clusters. Notably, the number of monocytes in the non-CR group was higher than that in the CR group (Fig. 2E). Meanwhile, we also separately listed each individual sample within the two groups to analyze disease characteristics, and we found that the absolute and relative abundance in the non-CR group was significantly higher than that in the CR group (Fig. 2F).

Fig. 2.

Fig. 2

Open in a new tab

Heterogeneity of monocyte phenotypes and functions between the CR and non-CR groups. A UMAP plot of monocytes colored by cell clusters; B UMAP plot colored by monocyte subpopulations after annotation; C Feature plots of selected marker genes for each cell type. The color scale indicates the log1p-normalized expression levels of the genes; D Dot plot of selected marker genes for each cell type. Dot size represents the percentage of cells expressing each gene, while the dot color represents expression level; E UMAP plots of the CR and non-CR groups; F UMAP plots of individual samples.

Monocyte subpopulation and pseudotime trajectory analysis in non-CR AML

We observed disease-related changes in the monocytes based on the subpopulation and pseudotime trajectory analysis. The monocytes of the non-CR group detected in this study were further classified into nine subclusters after a dimensionality reduction (Fig. 3A). We used pseudotime techniques to reconstruct the differentiation trajectories of monocytes in order to better understand the function of each monocyte subcluster in non-CR AML. Figure 3B shows the distribution patterns of 14 monocyte subclusters following pseudotime analysis. Following that, seven monocyte states were determined using various time nodes (Fig. 3C). The branch of State 1 was chosen as the starting point in accordance with the trajectory schematic generated by the algorithm (Fig. 3D). In a heatmap (Fig. 3E), the genes exhibiting notable variations in expression along the pseudotime axis in the monocytes were grouped into four modules according to their expression patterns. The pseudotime kinetics of CXCL8, STAT1, F3, TNFAIP3, and PLAU in the seven monocyte states are shown in Fig. 3F.

Fig. 3.

Fig. 3

Open in a new tab

Monocyte subpopulation and pseudotime trajectory analysis in non-CR AML. A UMAP plot colored by various cell clusters; B Monocle pseudotime trajectory showing the progression of fourteen monocyte subclusters; C The monocytes subcluster trajectory was separated into seven cell states; D Monocle pseudotime trajectory indicating the starting and terminal points of the differentiation process; E The DEGs (in rows, q-value < 10 − 10) along the pseudotime were hierarchically clustered into four modules. The top annotated GO terms in each cluster were provided; F Kinetic expression profiles of selected genes along the pseudotime. Each dot represents a single cell; colors indicate different clusters, and the y-axis represents the log-normalized expression level.

Cell–cell communication between monocytes and other cell types in AML

We conducted a cell-to-cell communication analysis by tracking the ligand–receptor interactions in order to further examine the impact of monocytes on other cells in AML. The ligand-receptor interactions between monocytes and other major cells were found to be stronger in the CR group than in the non-CR group (Fig. 4A, B). This pattern potentially suggests that there is less contact between non-CR tumor cells and TME adaptive responses. We observed significant differences in the frequency and strength of cellular communications between the CR and the non-CR groups as we progressed into signal pathway analysis, as shown in Fig. 4C. Furthermore, as illustrated in Fig. 4D, we highlighted notable changes in ligand-receptor pairings within the non-CR cohort. In order to identify possible new molecular targets for therapeutic intervention, these molecular interactions imply that intercellular communication plays crucial regulatory roles.

Fig. 4.

Fig. 4

Open in a new tab

Cell–cell communication between monocytes and other cell types in AML. A Comparison of the number and strength of cell communications between CR and non-CR groups; B Interaction net count plot of CR and non-CR groups. Line thickness represents the number of interactions between the two cell types; C Differences in the number and strength of signaling pathway-related cell communications between CR and non-CR groups; D Key ligand-receptor pairs with significant changes in CR versus non-CR tissues.

Identification of bioactive components and comprehensive pathway analysis of ginseng against AML

In the TCMSP, 22 active compounds were filtered with OB ≥ 30% and DL ≥ 0.18. Then use the SwissADME in order to select effective compounds characterized by high GI and satisfying at least two drug-likeness criteria. The effective compounds of ginseng were listed in Supplemental Table 2. A total of 776 target genes for these active components were identified via the SwissTargetPrediction and CTD databases, with detailed information presented in Supplemental Table 3. To find the targets of ginseng in AML cells, as shown in Fig. 5A, this network consists of 791 nodes, including one representing ginseng, 14 for its active compounds, and 776 for their corresponding target proteins. The gene chip dataset GSE240618 was retrieved from the GEO database, and a total of 607 upregulated AML-related targets were screened and obtained from these databases. Volcano maps of DEGs were shown in Fig. 5B. There are 34 shared targets between AML and ginseng, including targets of compounds in ginseng, according to the Venn diagram of AML and ginseng targets displayed in Fig. 5C. Figure 5D displays the "drugs-active compounds-potential targets" network, which consists of 34 targets and 1 herb. Additionally, compounds interact with targets more frequently, indicating that ginseng exerts therapeutic effects against AML by acting on a variety of targets with many compounds. The clusterProfiler R package was utilized to assess 34 intersecting targets in order to further investigate the molecular processes of ginseng for the treatment of AML. On the basis of the 34 targets arranged by degree value, GO enrichment analysis was carried out. BP, MF, and CC were the three categories of items that were obtained. A total of 395 BP terms, 39 MF terms, and 9 CC terms were significantly enriched. The top 6 enriched terms were visualized in Fig. 5E. A total of 12 KEGG pathways were identified. In this study, 10 significantly enriched KEGG pathway terms are shown in Fig. 5F. Figure 5G presented the string diagram of the top three pathways, each pathway was interacted by the common targets, indicating that ginseng could treat AML through orchestrating multiple pathways.

Fig. 5.

Fig. 5

Open in a new tab

Identification of bioactive components and comprehensive pathway analysis of ginseng against AML. A The “Drug-Active Component-Target Gene” network. The red triangle represents Ginseng, the orange rectangles denote active compounds, and the light yellow rectangles indicate target genes; B Volcano plot of differential gene expression between monocyte groups, highlighting significantly upregulated and downregulated genes based on the thresholds of p value < 0.05 and |log₂fold change (avglog2FC)|≥ 1; C Venn diagram visualizing the common targets between drug targets and disease-lesioned cell-related targets; D The “Drug-Active Component-Intersecting Gene-Disease” network. The red triangle represents ginseng, the orange rhombuses indicate active compounds, the light yellow rectangles denote intersecting genes, and the red triangle represent AML; E GO enrichment analysis displaying the top 6 representative BPs, MFs, and CCs with the lowest p values (p < 0.05); F KEGG pathway enrichment analysis identifying 10 significant pathways (p < 0.05) related to the intersecting genes were screened out; G Chord diagram analysis of the five key pathways involved in signal transduction.

In addition, the top ten important genes of each index were obtained through analyses of maximal clique centrality (MCC), maximum neighborhood component (MNC), and degree (number of nodes). The core genes (Fig. 6A) were then obtained by taking the intersection of these genes (Venn diagram analysis). Together with KEGG and GO enrichment analyses, the most relevant signaling pathways were selected: the AGE-RAGE signaling pathway and the NF-κB signaling pathway. The compound-target-pathway network was also constructed. As shown in the figure, the network comprises one node representing AML, five nodes corresponding to key targets, two nodes for signaling pathways, and one node for the drug (Fig. 6B). In Fig. 6C, D, numerous routes influence apoptosis, DNA damage, cell survival, and other processes. Core genes are dark blue, while the additional genes are green.

Fig. 6.

Fig. 6

Open in a new tab

PPI network analysis and identification of central targets of ginseng against AML-associated cells. A Identification of hub genes by intersecting results from MCC, MNC, and Degree algorithms; B The “Drug-Pathway-Common Gene-Disease” network. The red triangle represents ginseng, the blue hexagons denote pathways, the yellow rectangles indicate genes, and the orange diamond represents AML; C AGE-RAGE signaling pathway. The core targets of ginseng and AML are highlighted in blue; D NF-κB signaling pathway. The core targets of ginseng and AML are highlighted in blue.

Molecular docking validation

Using molecular docking studies, an analysis was conducted to investigate the molecular interactions between the fourteen active compounds of ginseng and potential target genes. The main active ingredients showed strong binding affinities with these targets, indicating robust binding interactions between these compounds and their targets. In particular, panaxadiol displayed the strongest binding affinity to the respective targets, with values of −9.2 kcal/mol for CXCL8, −7.5 kcal/mol for STAT1, -10.0 kcal/mol for F3, −7.8 kcal/mol for TNFAIP3, and −7.8 kcal/mol for PLAU. Docking scores are shown in Fig. 7A. According to the study's findings, panaxadiol may serve as a pivotal bioactive component. As shown in Fig. 7B-F, target proteins can bind to panaxadiol via hydrophobic interactions and hydrogen bonds. Based on the docking studies, we effectively predicted the binding of ginseng to target proteins.

Fig. 7.

Fig. 7

Open in a new tab

Molecular docking validation. A Heatmap illustrating the binding affinities of molecular docking; B–F Detailed molecular docking visualization and binding affinities between panaxadiol and its targets: B CXCL8 (–9.20 kcal/mol), C STAT1 (–7.50 kcal/mol), D F3 (–10.00 kcal/mol), E TNFAIP3 (–7.80 kcal/mol), and F PLAU (–7.80 kcal/mol)

In vitro validation of the synergistic effects of panaxadiol and Ara-C

To investigate whether ginseng enhances chemosensitivity in AML, NOMO-1 and THP-1 cells were co-treated with panaxadiol and cytarabine (Ara-C). The combination treatment significantly potentiated apoptosis compared with either single-agent treatment, as assessed by Annexin V/PI staining and flow cytometry (Fig. 8A, B). Subsequently, we analyzed the mRNA expression levels of pro-apoptotic proteins, anti-apoptotic proteins, and related target genes. qRT-PCR analysis revealed that the combination treatment significantly downregulated BCL2 and F3 expression while markedly upregulating BAX, CXCL8, STAT1, TNFAIP3, and PLAU (Fig. 8C). Western blot analysis showed increased BAX and cleaved Caspase-3 and decreased BCL2 in the combination group (Fig. 8D). To further elucidate the mechanism underlying panaxadiol-enhanced cytarabine-induced apoptosis, Western blot analysis showed that p-p65 levels were significantly reduced in the combination group compared to the single-agent treatments (Fig. 8D). In conclusion, panaxadiol potentiates the pro-apoptotic efficacy of cytarabine, thereby promoting AML cell death.

Fig. 8.

Fig. 8

Open in a new tab

In vitro validation of the synergistic effects of panaxadiol and Ara-C. A, B Synergistic effects of panaxadiol on Ara-C–induced cell death in NOMO-1 and THP-1 cells. Cells were treated with Ara-C (1 μM), panaxadiol (50 μM), or their combination for 48 h. DMSO-treated cells served as the control group; C Relative mRNA expression of BCL2, BAX, CXCL8, STAT1, F3, TNFAIP3, and PLAU in NOMO-1 and THP-1 cells as detected by RT-qPCR; D Western blotting of BCL-2, BAX, cleaved caspase-3, p65, and p-p65 proteins in NOMO-1 and THP-1 cells following combination treatment. Data are presented as mean±SD, *p < 0.05, **p < 0.01, ***p < 0.001, ns, not significant (Student’s t–test).

Discussion

AML is the most common type of acute leukemia in adults. The current standard treatment regimen is Ara-C in combination with other chemotherapy drugs, but survival rates are substantially low for patients over 60 years old and those with relapsed or refractory disease. The five-year overall survival rates are 4–18% and approximately 10% respectively [32]. The clinical and pathophysiological mechanisms underlying the failure to achieve complete remission following AML treatment remain poorly understood. TCM has demonstrated significant potential in recent years for prolonging survival time and improving quality of life in the treatment of AML [18]. Ginsenosides are one of the bioactive constituents of ginseng and often exhibit potent anticancer properties. They play an important role in enhancing chemotherapeutic sensitivity and inducing tumor cell apoptosis, either alone or in combination with anticancer drugs [33, 34]. A series of studies also investigated the anti-proliferative and apoptotic effects of panaxadiol on human colon cancer, breast cancer, and liver cancer [35–37]. However, the role of panaxadiol in combination with Ara-C remains unexplored in AML cells.

In our study, we mapped the deep differentiation profile of monocytes in patients with AML in CR and non-CR. It is based on the GSE240618 single-cell transcriptomics dataset. Furthermore, previous research has shown that monocytes are involved in AML chemoresistance [38]. Pseudotime trajectory analysis revealed a differentiation arrest of non-CR monocytes. Consistent with our findings, research by Abe et al. demonstrated that monocytes can enhance the immune response of NK cells [39]. This study used network pharmacology to investigate how the active compounds in ginseng interact with potential AML targets. The study integrated multiple database resources and we performed molecular docking of the identified active compounds into core human AML targets. Additionally, we found that the combination of panaxadiol and Ara-C exhibits additive or synergistic anti-leukemic activity in various AML cells. Integrating the core targets from the constructed protein–protein interaction network and in vitro experiments demonstrated that panaxadiol exerts a synergistic effect with Ara-C in inducing apoptosis of AML cells, potentially mediated through the NF-κB signaling pathway.

In conclusion, we characterized tumor microenvironment heterogeneity in CR and non-CR AML patients and discovered the significant potential of ginseng and its active compounds, especially ginsenosides like panaxadiol, to target abnormal monocyte differentiation and overcome AML resistance. We evaluated the effects of Ara-C and panaxadiol on the inhibition of AML cell proliferation and the induction of apoptosis. Cell viability and apoptosis assays demonstrated that the combination of Ara-C and panaxadiol achieved efficacy comparable to that of higher concentrations of Ara-C monotherapy. To overcome the current challenges and maximize ginseng's clinical potential in treating AML, we need to actively explore effective combination therapy regimens. Cell viability and apoptosis assays indicated that Ara-C combined with panaxadiol achieved efficacy comparable to that of higher-dose Ara-C alone. This combination therapy may represent a promising strategy for the clinical treatment of Ara-C-resistant AML patients. However, current mechanistic insights largely rely on database-based predictions and in vitro experiments; therefore, further validation in appropriate in vivo models is required.

Conclusion

In conclusion, our study utilized the GEO dataset to identify AML disease targets and constructed a single-cell transcriptomic atlas to explore the roles of monocytes. Through network pharmacology and molecular docking approaches, we systematically investigated the bioactive constituents, potential targets, and underlying mechanisms of ginseng in AML treatment, identifying several promising constituents, including ginsenoside-Rh4, gomisin B, and panaxadiol. Although the complex composition of ginseng makes it difficult to pinpoint a single therapeutic component, existing research supports its efficacy against AML. This study provides a foundation for future research into the multi-compound, multi-pathway properties of ginseng and its potential for drug screening in AML.

Supplementary Information

Additional file 1. (17.8KB, docx)
Additional file 2. (18.4KB, docx)
Additional file 3. (18.2KB, xlsx)
Additional file 4. (154.2KB, docx)

Acknowledgements

We acknowledge the Gene Expression Omnibus (GEO) database, the RCSB Protein Data Bank, and the TCMSP database for providing the publicly available data used in this study.

Author contributions

Yan Zhang: Data curation (equal); formal analysis (equal); writing – original draft (lead). Bobin Su: Data curation (equal); formal analysis (equal); methodology (equal). Changqing Jiao: Data curation (equal). Zhonghui Wu: Data curation (equal). Junjie Zhou: Formal analysis (equal). Jian Ge: Conceptualization (lead); funding acquisition (lead); supervision (lead); writing – review and editing (equal). All authors have read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 82370175 and 81200371); Key Research and development Project of Anhui Province (Grant No. 201904a07020057); Research Foundation of Anhui Medical University (Grant No.2020xkj166); Research Foundation of Anhui Institute of Translational Medicine (Grant No. 2021zhyx-C32); Health Research Program of Anhui (AHWJ2024BAa10003); Postgraduate Innovation Research and Practice Program of Anhui Medical University (Grant No. YJS20240019).

Data availability

The gene expression datasets generated and analysed during the current study are available in the NCBI Gene Expression Omnibus (GEO) repository under accession number GSE240618 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE240618). The three-dimensional structures of the core target proteins were downloaded from the RCSB Protein Data Bank (PDB) ([https://www.rcsb.org/] (https://www.rcsb.org/)) under the following PDB IDs: 4XDX (CXCL8), 7NUF (STAT1), 1BOY (F3), 5LRX (TNFAIP3), and 1OWD (PLAU). The molecular structures of the active ingredients of ginseng were retrieved by name from the TCMSP database (https://old.tcmsp-e.com/tcmsp.php). All patient information is provided in Supplemental Table 4, and all other supporting data are included within the article and its supplementary materials.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Yan Zhang and Bo-bin Su have contributed equally to this work.

Contributor Information

Jun-jie Zhou, Email: zhoujj_0924@163.com.

Jian Ge, Email: gejian@ahmu.edu.cn.

References

  • 1.Venugopal S, Sekeres MA. Contemporary management of acute myeloid leukemia: a review. JAMA Oncol. 2024;10(10):1417–25. [DOI] [PubMed] [Google Scholar]
  • 2.Zhou Y, Huang G, Cai X, et al. Global, regional, and national burden of acute myeloid leukemia, 1990–2021: a systematic analysis for the global burden of disease study 2021. Biomark Res. 2024;12(1):101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Chakraborty S, Park CY. Pathogenic mechanisms in acute myeloid leukemia. Curr Treat Options Oncol. 2022;23(11):1522–34. [DOI] [PubMed] [Google Scholar]
  • 4.Thol F, Döhner H, Ganser A. How I treat refractory and relapsed acute myeloid leukemia. Blood. 2024;143(1):11–20. [DOI] [PubMed] [Google Scholar]
  • 5.De Visser KE, Joyce JA. The evolving tumor microenvironment: from cancer initiation to metastatic outgrowth. Cancer Cell. 2023;41(3):374–403. [DOI] [PubMed] [Google Scholar]
  • 6.Lv B, Wang Y, Ma D, et al. Immunotherapy: reshape the Tumor Immune Microenvironment. Front Immunol. 2022;13:844142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Deepak KGK, Vempati R, Nagaraju GP, et al. Tumor microenvironment: challenges and opportunities in targeting metastasis of triple negative breast cancer. Pharmacol Res. 2020;153:104683. [DOI] [PubMed] [Google Scholar]
  • 8.Ren X, Zhang L, Zhang Y, et al. Insights gained from single-cell analysis of immune cells in the tumor microenvironment. Annu Rev Immunol. 2021;39:583–609. [DOI] [PubMed] [Google Scholar]
  • 9.Olingy CE, Dinh HQ, Hedrick CC. Monocyte heterogeneity and functions in cancer. J Leukoc Biol. 2019;106(2):309–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Song H, Chen L, Pan X, et al. Targeting tumor monocyte-intrinsic PD-L1 by rewiring signaling and enhancing STING agonist therapy. Cancer Cell. 2025;43(3):503-518.e10. [DOI] [PubMed] [Google Scholar]
  • 11.Fendl B, Berghoff AS, Preusser M, et al. Macrophage and monocyte subsets as new therapeutic targets in cancer immunotherapy. ESMO Open. 2023;8(1):100776. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Miao P, Yu J, Chen Z, et al. Establishment and verification of a TME prognosis scoring model based on the acute myeloid leukemia single-cell transcriptome. Sci Rep. 2024;14(1):19811. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Cerrano M, Duchmann M, Kim R, et al. Clonal dominance is an adverse prognostic factor in acute myeloid leukemia treated with intensive chemotherapy. Leukemia. 2021;35(3):712–23. [DOI] [PubMed] [Google Scholar]
  • 14.Havel JJ, Chowell D, Chan TA. The evolving landscape of biomarkers for checkpoint inhibitor immunotherapy. Nat Rev Cancer. 2019;19(3):133–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Mazziotta F, Biavati L, Rimando J, et al. CD8+ T-cell differentiation and dysfunction inform treatment response in acute myeloid leukemia. Blood. 2024;144(11):1168–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Vadakekolathu J, Rutella S. Escape from T-cell-targeting immunotherapies in acute myeloid leukemia. Blood. 2024;143(26):2689–700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Nie ZY, Zhao MH, Cheng BQ, et al. Tanshinone IIA regulates human AML cell proliferation, cell cycle, and apoptosis through miR-497-5p/AKT3 axis. Cancer Cell Int. 2020;20:379. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Gao X, Zuo X, Min T, et al. Traditional Chinese medicine for acute myelocytic leukemia therapy: exploiting epigenetic targets. Front Pharmacol. 2024;15:1388903. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Valdés-González JA, Sánchez M, Moratilla-Rivera I, et al. Immunomodulatory, anti-inflammatory, and anti-cancer properties of ginseng: a pharmacological update. Molecules. 2023;28(9):3863. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Chen W, Balan P, Popovich DG. Review of ginseng anti-diabetic studies. Molecules. 2019;24(24):4501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Huang J, Liu D, Wang Y, et al. Ginseng polysaccharides alter the gut microbiota and kynurenine/tryptophan ratio, potentiating the antitumour effect of anti-PD-1/PD-L1 immunotherapy. Gut. 2022;71(4):734–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Yang Y, Nan Y, Du Y, et al. Ginsenosides in cancer: proliferation, metastasis, and drug resistance. Biomed Pharmacother. 2024;177:117049. [DOI] [PubMed] [Google Scholar]
  • 23.Hwang SJ, Bang HJ, Lee HJ. Ginsenoside Re inhibits melanogenesis and melanoma growth by downregulating microphthalmia-associated transcription factor. Biomed Pharmacother. 2023;165:115037. [DOI] [PubMed] [Google Scholar]
  • 24.Kim J, Zhu Y, Chen S, et al. Anti-glioma effect of ginseng-derived exosomes-like nanoparticles by active blood-brain-barrier penetration and tumor microenvironment modulation. J Nanobiotechnology. 2023;21(1):253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Qi W, Yan X, Xu X, et al. The effects of cytarabine combined with ginsenoside compound K synergistically induce DNA damage in acute myeloid leukemia cells. Biomed Pharmacother. 2020;132:110812. [DOI] [PubMed] [Google Scholar]
  • 26.Li F, Zhang H, Huang Y, et al. Single-cell transcriptome analysis reveals the association between histone lactylation and cisplatin resistance in bladder cancer. Drug Resist Updat. 2024;73:101059. [DOI] [PubMed] [Google Scholar]
  • 27.Lei Y, Tang R, Xu J, et al. Applications of single-cell sequencing in cancer research: progress and perspectives. J Hematol Oncol. 2021;14(1):91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Guo S, Liu X, Zhang J, et al. Integrated analysis of single-cell RNA-seq and bulk RNA-seq unravels T cell-related prognostic risk model and tumor immune microenvironment modulation in triple-negative breast cancer. Comput Biol Med. 2023;161:107066. [DOI] [PubMed] [Google Scholar]
  • 29.Zhang Y, Zhang Z. The history and advances in cancer immunotherapy: understanding the characteristics of tumor-infiltrating immune cells and their therapeutic implications. Cell Mol Immunol. 2020;17(8):807–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Zhao L, Zhang H, Li N, et al. Network pharmacology, a promising approach to reveal the pharmacology mechanism of Chinese medicine formula. J Ethnopharmacol. 2023;309:116306. [DOI] [PubMed] [Google Scholar]
  • 31.Xie SS, Hou R, Gao L, et al. Insulin-like growth factor-binding protein 7 and cadmium-induced hepatorenal fibrosis. J Am Soc Nephrol. 2025;36(9):1716–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Wysota M, Konopleva M, Mitchell S. Novel therapeutic targets in acute myeloid leukemia (AML). Curr Oncol Rep. 2024;26(4):409–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Chung KS, Cho SH, Shin JS, et al. Ginsenoside Rh2 induces cell cycle arrest and differentiation in human leukemia cells by upregulating TGF-β expression. Carcinogenesis. 2013;34(2):331–40. [DOI] [PubMed] [Google Scholar]
  • 34.Kim AD, Kang KA, Kim HS, et al. A ginseng metabolite, compound K, induces autophagy and apoptosis via generation of reactive oxygen species and activation of JNK in human colon cancer cells. Cell Death Dis. 2013;4(8):e750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Wang Z, Li MY, Zhang ZH, et al. Panaxadiol inhibits programmed cell death-ligand 1 expression and tumour proliferation via hypoxia-inducible factor (HIF)-1α and STAT3 in human colon cancer cells. Pharmacol Res. 2020;155:104727. [DOI] [PubMed] [Google Scholar]
  • 36.Xu L, Zhang X, Xiao S, et al. Panaxadiol as a major metabolite of AD-1 can significantly inhibit the proliferation and migration of breast cancer cells: in vitro and in vivo study. Bioorg Chem. 2021;116:105392. [DOI] [PubMed] [Google Scholar]
  • 37.Xu L, Xiao S, Lee JJ, et al. In vivo metabolites of panaxadiol inhibit HepG-2 cell proliferation by inducing G1 arrest and ROS-mediated apoptosis. J Agric Food Chem. 2022;70(37):11560–71. [DOI] [PubMed] [Google Scholar]
  • 38.Pei S, Pollyea DA, Gustafson A, et al. Monocytic subclones confer resistance to venetoclax-based therapy in patients with acute myeloid leukemia. Cancer Discov. 2020;10(4):536–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Abe S, Asahi T, Hara T, et al. Hematopoietic cell-derived IL-15 supports NK cell development in scattered and clustered localization within the bone marrow. Cell Rep. 2023;42(9):113127. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Additional file 1. (17.8KB, docx)
Additional file 2. (18.4KB, docx)
Additional file 3. (18.2KB, xlsx)
Additional file 4. (154.2KB, docx)

Data Availability Statement

The gene expression datasets generated and analysed during the current study are available in the NCBI Gene Expression Omnibus (GEO) repository under accession number GSE240618 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE240618). The three-dimensional structures of the core target proteins were downloaded from the RCSB Protein Data Bank (PDB) ([https://www.rcsb.org/] (https://www.rcsb.org/)) under the following PDB IDs: 4XDX (CXCL8), 7NUF (STAT1), 1BOY (F3), 5LRX (TNFAIP3), and 1OWD (PLAU). The molecular structures of the active ingredients of ginseng were retrieved by name from the TCMSP database (https://old.tcmsp-e.com/tcmsp.php). All patient information is provided in Supplemental Table 4, and all other supporting data are included within the article and its supplementary materials.

Articles from Discover Oncology are provided here courtesy of Springer

RESOURCES