Abstract

   DNA methylation plays a critical role in myelodysplastic syndrome
   (MDS). Here, we aimed to observe the effects of polydatin (PD) on DNA
   methylation in MDS cells on a genome-wide scale and explore the
   underlying mechanisms, providing new evidence for PD as a novel
   hypermethylation agent. We used the Gene Expression Omnibus (GEO)
   online database to evaluate the DNA methylation characteristics of MDS
   patients. A Human Methylation 850 K BeadChip was used to evaluate the
   effects of PD on DNA methylation in SKM-1 cells. Western blotting (WB)
   was used to observe changes in the expression of related proteins.
   Cytoscape was used to determine the key genes that were hypermethylated
   by PD. The therapeutic effects were evaluated using flow cytometry
   experiment and a cell counting kit-8 (CCK-8) assay in vitro. Data from
   the GEO online database revealed that aberrant gene hypomethylation
   plays an important role in MDS. In MDS cells, 448 genes (71.91%) were
   hypermethylated following PD treatment. These hypermethylated genes are
   related to cancer-related signaling pathways. Moreover, key
   hypermethylated genes, including PIK3CA, ITPR3 and SPOPL, were
   identified, and these three genes are all oncogenes. Most importantly,
   PD decreased the protein expression of the above three oncogenes.
   Finally, we found that PD could inhibit the proliferation of MDS cells,
   arrest them in the S phase and induce their apoptosis. Our findings
   demonstrated that PD has therapeutic effects on MDS by inhibiting the
   protein expression of oncogenes via hypermethylation in vitro,
   indicating that PD may be a novel hypermethylation agent.

   Keywords: Myelodysplastic syndrome, Polydatin, Oncogene,
   Hypermethylation, Hypomethylation

   Subject terms: Cancer genomics, Cancer therapy, Haematological cancer

Introduction

   Myelodysplastic syndrome (MDS) is a category of malignant clonal
   disorders that stem from hematopoietic stem cells within the bone
   marrow. These conditions are defined by ineffective blood cell
   production, persistent cytopenia, and an elevated likelihood of
   progressing to acute myeloid leukemia (AML)^[38]1. In the past, MDS was
   considered to involve hypermethylation^[39]2. However, emerging studies
   suggest that aberrant hypomethylation is also a critical factor in MDS
   pathogenesis. For example, a previous study reported that aberrant
   hypomethylation of preferentially expressed antigens in the
   melanoma (PRAME) promoter was found in 19.6% of MDS patients, and these
   patients had a poor prognosis^[40]3. Moreover, SALL4 is instrumental in
   the progression of various cancers, functioning as an oncogene^[41]4.
   Lin et al. found that 21.7% (18/83) of MDS patients displayed abnormal
   hypomethylation of the SALL4 gene, which correlated with notably
   shorter survival rates^[42]5. In our earlier research, we discovered
   several hypomethylated genes in a group of five MDS patients in
   contrast to three healthy donors utilizing the Human Methylation 850 K
   Bead Chip^[43]6. However, a larger sample size is required to
   comprehensively investigate DNA methylation patterns in MDS patients,
   especially hypomethylated oncogenes.

   At present, there is still no cure for MDS except for hemopoietic stem
   cell transplantation, and most patients eventually die from fatal
   infection and bleeding caused by cytopenia and acute leukemia^[44]7.
   Several studies have emphasized the crucial role of abnormal DNA
   methylation in the pathologic development of MDS^[45]8. Over the last
   two decades, the use of the demethylation agents azacitidine and
   decitabine, which mainly target abnormal hypermethylation, has greatly
   progressed the treatment of MDS, especially high-risk MDS^[46]9.
   However, a real-world clinical study revealed that the median survival
   for high-risk MDS patients receiving azacytidine treatment was limited
   to 18 months^[47]10. Therefore, these agents fail to fully address the
   clinical needs of MDS patients, highlighting the urgent need for novel
   hypomethylation-targeting therapies to achieve improved therapeutic
   outcomes.

   Recently, emerging bioactive compounds derived from traditional Chinese
   herbal medicine have shown potential as a solution to overcome the
   limitations of current MDS treatments^[48]6. Polydatin (PD), a
   polyphenolic compound extracted from Polygoni Cuspidati Rhizoma et
   Radix, has been reported to inhibit the proliferation of multiple tumor
   cell types, such as osteosarcoma, colorectal cancer, and hepatocellular
   carcinoma cells^[49]11–[50]13. However, no studies have reported how PD
   affects DNA methylation in MDS cells. Here, we hypothesized that PD
   targets the aberrant hypomethylation in MDS. To prove this hypothesis,
   we began this study by utilizing the GEO database to examine DNA
   methylation patterns in MDS patients, using a sample size larger than
   that in our previous research. Then, we assessed the genome-wide
   changes in DNA methylation induced by PD treatment in these cells.
   Kyoto Encyclopedia of Genes and Genomes (KEGG) was used to analyze
   those PD-induced hypermethylated genes. Western blotting (WB) was
   carried out to observe the effects of PD on DNA methyltransferases
   (DNMTs) in both SKM-1 and MUTZ-1 cells. Furthermore, the key
   hypermethylated genes induced by PD were identified using Cytoscape
   3.10.2, and their protein expression levels were determined via WB.
   Finally, we performed assays using a Cell Counting Kit-8 (CCK-8) and
   flow cytometry to evaluate the impact of PD treatment on MDS cells.

Materials and methods

Reagents and cell lines

   PD was obtained from the National Institutes for Food and Drug Control.
   A stock solution was prepared by dissolving PD in DMSO. A Cell Counting
   Kit-8 (# E1CK000208-10, EnoGene, Nanjing, China) was used to determine
   cell viability. Detection of cell cycle distribution and apoptosis was
   performed using a Cell Cycle and Apoptosis Analysis Kit (# C1052) and
   an Annexin V-FITC Apoptosis Detection Kit (# p0013B) from Beyotime, in
   Hangzhou, China. The DNA Extraction Kit (# Y5-69,506) was purchased
   from Qiagen, Germany. EZ DNA methylation kit (# D5005) was from Zymo
   Research, USA. The protein extraction kit (# GPP1815) was obtained from
   Gene Pool, China. Specific primary antibodies and secondary antibodies
   were sourced from Abcam (Cambridge, USA), Proteintech (Rosemont, USA),
   and Bioss (Beijing, China). The MDS cell line SKM-1 was a gift from
   Professor Su-ning Chen at Soochow University. The MUTZ-1 cell line was
   sourced from Jennio Biotech Co., Ltd. (Guangzhou, China).

DNA methylation information acquisition and analysis

   The DNA methylation profiles of 17 MDS patients and 14 healthy donors
   were downloaded from GEO online database ([51]GSE124390, [52]GSE119617;
   [53]https://www.ncbi.nlm.nih.gov/). The human samples were all
   mesenchymal stromal cells (MSCs) and were analyzed using the Illumina
   Infinium Methylation EPIC BeadChip 850 k. The selected data passed the
   quality control. Gene annotation was performed via the R package
   Bioconductor.

DNA extraction and methylation detection

   The SKM-1 cells were divided into a DMSO control group (MDS cells
   treated with DMSO) and a 30 μmol/L PD group (MDS cells treated with
   DMSO and PD at 30 μmol/L). Following 72 h of treatment, cells were
   collected. Genomic DNA was immediately extracted using a Qiagen DNeasy
   Blood & Tissue Kit in accordance with the product’s guidelines. The
   Illumina Infinium Methylation EPIC BeadChip 850 k was subsequently used
   to detect the methylation status of the DNA samples. Briefly, the
   bisulfite DNA was conversed with an EZ DNA methylation kit according to
   the product’s instruction manual. Following amplification, fracture,
   precipitation and resuspension, the DNA was hybridized and washed. An
   SQ fluorescence scanner was used to scan the DNA. The degree of DNA
   methylation was then calculated from the raw signal values to obtain
   the original beta value matrix. The beta value is a commonly used index
   to measure a DNA methylation site ranging from 0–1. As the beta value
   approaches 1, DNA methylation increases; as it nears 0, methylation
   decreases. The β value was calculated using the following formula:
   graphic file with name 41598_2025_1867_Article_Equa.gif

Bioinformatics analyses

   The hypomethylated genes of MDS patients vs. healthy donors and
   hypermethylated genes induced by PD were analyzed via KEGG pathway
   enrichment using the DAVID platform for bioinformatics. The criterion
   was set as P<0.05, and the top 30 pathways and functions are listed.
   Protein-protein interaction (PPI) network construction was carried out
   by STRING ([54]https://string-db.org/), and then the key genes were
   obtained and visualized utilizing Cytoscape 3.10.2
   ([55]https://cytoscape.org/).

Western blotting

   The two cell lines were divided into three groups: DMSO control group
   (MDS cells treated with DMSO), 10 μmol/L group (DMSO and PD at
   10 μmol/L), and 30 μmol/L group (DMSO and PD and 30 μmol/L). Proteins
   extraction was conducted by a protein extraction kit following the
   product’s instructions. After performing SDS‒PAGE, proteins were
   transferred to a PVDF membrane and incubated with the primary
   antibodies: anti-Dnmt3b (Abcam/ab79822, 1:500), anti-Dnmt3a
   (Abcam/ab188470, 1:500), anti-Dnmt1 (Abcam/ab188453, 1:500),
   anti-PIK3CA (Bioss/bs2067R, 1:500), anti-ITPR3
   (Proteintech/20,729–1-Ig, 1:500), anti-SPOPL (Proteintech/17,740–1-Ig,
   1:500), and anti-Actin (Abcam/ab8226, 1:3000) /anti-GAPDH
   (Abcam/ab181602, 1:3000). Afterward, the membranes were incubated with
   goat anti-rabbit IgG HRP (Abcam/ab6721, 1:5000). The antigen-antibody
   complexes were detected using enhanced chemiluminescence (ECL) reagents
   (Thermo Fisher, USA) in a dark room. The total grayscale value of each
   protein band was quantified via Quantity One v.4.6.2.

CCK-8 cell viability assay

   SKM-1 and MUTZ-1 cells were grown and maintained in RPMI-1640 medium
   containing 10% fetal bovine serum (Gibco, California, USA) under
   standard culture conditions. Then the cells were seeded at a density of
   2 × 10^5 cells per well in 96-well plates. PD was added at final
   concentrations of 0, 25, 50, 100, and 200 μmol/L. Following 72 h of
   incubation, CCK-8 reagent (10μL) was added, followed by incubation for
   1 h. The absorbance at 460 nm was assessed utilizing a microplate
   reader (Biotek, Shanghai, China). The experiment was repeated three
   times, with five replicates each time, and the cell viability of each
   group was calculated.

Cell cycle distribution and apoptosis assays

   The exponentially growing phase cells were adjusted to a density of
   2 × 10^5 cells/mL. There were three groups of cells: a blank group (MDS
   cells only), a DMSO control group (MDS cells treated with DMSO), a
   10 μmol/L group (MDS cells treated with DMSO and PD at 10 μmol/L), and
   a 30 μmol/L group (MDS cells treated with DMSO and PD at 30 μmol/L).
   Following 72 h of incubation, the cells were harvested for cell cycle
   and apoptosis assays. The cells were fixed with 75% cold ethanol at
   4 °C overnight for cell cycle distribution detection, after which the
   ethanol was removed, and the mixture was stained with a mixture of
   propidium iodide (PI) staining solution for 30 min in the dark. For
   cell apoptosis detection, Annexin V-FITC binding solution, PI, and
   Annexin V-FITC were added to the tubes in sequence and incubated with
   the cells for 10 min in the dark. Apoptosis and the cell cycle were
   analyzed by a flow cytometer (Beckman Coulter, USA). Replicate
   experiments were conducted to confirm the findings. Wright staining was
   performed to observe SKM-1 cell morphology.

Statistical analysis

   T-test and one-way ANOVA with the Bonferroni post-hoc test were
   performed to compare cell cycle distribution, apoptosis distribution
   and protein expression among different groups, considering p < 0.05 as
   statistically significant. For the methylation data, Bayesian and
   linear regression approaches were applied, differences in DNA
   methylation among different groups for SKM-1 cells were evaluated using
   a threshold of ± 0.06, and a p value <0.01 was set as the significance
   criterion. Statistical analyses were performed on data from at least
   three independent experiments.

Results

Abnormal gene hypomethylation is common in MDS patients

   To observe the methylation characteristics of MDS patients, we compared
   MSCs derived from seventeen MDS patients and fourteen healthy donors
   utilizing human methylation chip data from the GEO database. Numerous
   differentially methylated sites (DMSs) were identified in MDS patients,
   and the top 1000 DMSs are shown in a heatmap (Fig. [56]1a). We
   identified 5491 aberrantly hypomethylated sites and 14,759 aberrantly
   hypermethylated sites (Fig. [57]1b). KEGG pathway analysis revealed
   that several pathways related to cancer, such as PI3K-AKT signaling
   pathway and MAPK signaling pathway (Fig. [58]1c). Furthermore, PPI
   analysis revealed that the key hypomethylated genes in MDS included
   PIK3CA, RAP1, and PKT (Fig. [59]1d).

Fig. 1.

   [60]Fig. 1
   [61]Open in a new tab

   The 850 K BeadChip DNA methylation data revealed an aberrant
   methylation pattern in 17 MDS patients compared with 14 healthy donors.
   (A) Heatmap of the top 1000 DMSs ranked by differential values,
   comparing MDS patients to healthy donors. (B) Volcano plot of the DMSs
   between MDS patients and healthy donors. Red (up) represents the
   hypermethylated sites, and blue (down) represents the hypomethylated
   genes. (C) KEGG pathway enrichment analysis of differentially
   hypomethylated genes in MDS patients compared with healthy donors. (D)
   PPI network analysis of differentially hypomethylated genes in MDS
   patients; the size and intensity of the nodes correspond to the degree
   of interaction, with larger and darker nodes indicating higher degree
   values.

PD induces hypermethylation in SKM-1 cells on a genome-wide scale

   We utilized the Infinium Human Methylation 850 K BeadChip to assess
   whole-genome methylation changes in SKM-1 cells following treatment
   with PD. The heatmap indicated a significantly higher genome-wide
   methylation level in the PD-treated group compared to the DMSO control
   group (Fig. [62]2a). The volcano plot shows the DMSs between the two
   groups (Fig. [63]2b). We found that the methylation of 623 genes was
   notably altered by PD treatment (Tables [64]S1 and [65]S2), among which
   448 genes (71.91%) were hypermethylated. Several pathways related to
   cancer, such as PI3K-AKT signaling pathway, mTOR signaling pathway, and
   the cGMP-PKG signaling pathway, were enriched in these hypermethylation
   genes (Fig. [66]2c).

Fig. 2.

   [67]Fig. 2
   [68]Open in a new tab

   The 850 K BeadChip DNA methylation data revealed the effects of PD on
   gene methylation in SKM-1 cells. (A) Heatmap depicting the median
   methylation level at CpG sites between the DMSO control group and the
   PD-treated group, with red representing the PD treatment group (P1–P3)
   and blue representing the DMSO control group (C1–C3). (B) Volcano plot
   illustrating CpG methylation in genes of the PD group compared with
   those of the DMSO control group. The red vertical dotted line
   delineates ± 0.06 in the Beta value between the two groups, and the
   horizontal dotted line represents a p value threshold of 0.01.
   Hypermethylated genes (448) are shown in red, and hypomethylated genes
   (175) are shown in green. (C) The top 30 enriched pathways of the
   hypermethylated genes induced by PD treatment.

PD increases the protein expression of DNMTs in SKM-1 and MUTZ-1 cells

   Following PD treatment, the protein expression levels were examined in
   MDS cells, including DNMT1, DNMT3a, and DNMT3b (Fig. [69]3a). For SKM-1
   cells, PD treatment at 30 μmol/L significantly increased the
   expressions of DNMT1, DNMT3a, and DNMT3b (p < 0.05), whereas the
   expressions of DNMTs did not change after PD treatment at 10 μmol/L
   when comparing with those in DMSO control group respectively
   (p > 0.05). (Fig. [70]3b). In MUTZ-1 cells (Fig. [71]3c), PD treatment
   at 30 μmol/L significantly increased the expression of DNMT1 and DNMT3b
   (p < 0.05), whereas no change existed of the expression of DNMT3a in
   both PD treatment groups versus DMSO control group (p > 0.05).

Fig. 3.

   [72]Fig. 3
   [73]Open in a new tab

   PD increased the protein expression of DNA methyltransferases (DNMTs),
   including DNMT1, DNMT3a and DNMT3b, in SKM-1 and MUTZ-1 cells. (A) The
   expression levels of DNMT1, DNMT3a and DNMT3b following PD treatment
   for 48 h. (B) and (C) Grayscale values of DNMT1, DNMT3a and DNMT3b in
   SKM-1 and MUTZ-1 cells. *p < 0.05, n = 3.

PD inhibits the protein expression of PIK3CA, ITPR3 and SPOPL in SKM-1 cells

   Among the 448 hypermethylated genes induced by PD, hub hypermethylated
   genes including PIK3CA, SPOPL and ITPR3, were identified and marked in
   red using Cytoscape 3.10.2 (Fig. [74]4a).

Fig. 4.

   [75]Fig. 4
   [76]Open in a new tab

   PD inhibited the protein expression of PIK3CA, ITPR3 and SPOPL in SKM-1
   cells. (A) A PPI network constructed via the STRING database and
   visualized by Cytoscape. PIK3CA, ITPR3 and SPOPL were identified as the
   hub genes in red. (B) Protein expression bands of PIK3CA, ITPR3 and
   SPOPL following PD treatment for 48 h. Grayscale values of PIK3CA (C),
   ITPR3 (D) and SPOPL (E). *p < 0.05, **p < 0.01.

   DNA methylation is known to inhibit gene transcription. In this study,
   the expression levels of the oncogenes PIK3CA, ITPR3, and SPOPL were
   examined in SKM-1 cells following PD treatment (Fig. [77]4b). PD
   treatment at 10 and 30 μmol/L significantly decreased the expression of
   PIK3CA (p < 0.01) and ITPR3 (p < 0.05; Fig. [78]4c & d). The expression
   of SPOPL remained unchanged after PD treatment at 10 μmol/L, whereas
   treatment at 30 μmol/L significantly inhibited SPOPL expression
   compared with that in the DMSO control group (p < 0.05) (Fig. [79]4e).

PD inhibits the proliferation of SKM-1 and MUTZ-1 cells

   Figure [80]5a illustrates the chemical structure of PD, which was
   provided by the National Institutes for Food and Drug Control. SKM-1
   and MUTZ-1 cells were incubated with PD at 0 to 200 μM for 72 h.
   Figure [81]5b shows the dose-dependent inhibitory effect of PD on the
   viability of the two cell lines.

Fig. 5.

   Fig. 5
   [82]Open in a new tab

   PD reduces the viability of SKM-1 and MUTZ-1 cells. (A) Chemical
   structure of PD. (B) Dose–response curves for the proliferation of
   SKM-1 and MUTZ-1 cells after treatment with PD for 48 h. The error bars
   represent the means ± SEMs, n = 3.

PD induces S-phase cell cycle arrest in SKM-1 and MUTZ-1 cells

   Cell cycle analysis demonstrated that PD treatment altered the cell
   cycle distribution of SKM-1 and MUTZ-1 cells (Fig. [83]6a, b). The cell
   cycle distribution of the two cell lines in DMSO control group was
   similar to that in the blank group (p > 0.05). When PD was added at
   different concentrations, the proportion of cells in the S phase
   significantly changed in MUTZ-1 cells (control: 33.70 ± 2.07%;
   10 µmol/L: 47.93 ± 3.02%; 30 µmol/L: 61.57 ± 2.20%) and cells in the
   G0/G1 phase significantly decreased (control: 57.27 ± 2.75%; 10 µmol/L:
   39.77 ± 1.62%; 30 µmol/L: 25.43 ± 0.39%) (p < 0.0001; Fig. [84]6c). In
   SKM-1 cells, treatment with two concentrations of PD resulted in a more
   pronounced effect on the cell arrest in the S phase (p < 0.0001;
   Fig. [85]6d). These results indicate that PD induces S-phase arrest,
   partially blocking DNA replication and cell cycle progression.

Fig. 6.

   [86]Fig. 6
   [87]Open in a new tab

   PD induces S-phase arrest in SKM-1 and MUTZ-1 cells (n = 3). (A) and
   (B)Representative flow cytometry images. (C) and (D) Percentages of
   SKM-1 and MUTZ-1 cells in each phase of the cell cycle, respectively,
   after treatment with PD (0, 10, or 30 µmol) for 48 h. The values are
   expressed as the means ± SEMs. *p < 0.05, ****p < 0.0001.

PD induces apoptosis in SKM-1 and MUTZ-1 cells

   PD treatment notably increased the proportion of apoptotic cells
   (Fig. [88]7a). For example, the percentage of viable apoptotic SKM-1
   cells increased from 0.26 ± 0.03% (DMSO control) to 3.83 ± 0.03% (PD at
   30 μmol/L) (p < 0.0001), and change of nonviable apoptotic cells was
   similar (Fig. [89]7b). In MUTZ-1 cells, nonviable apoptosis increased
   significantly from 2.7 ± 1.11% (DMSO control) to 6.07 ± 0.65% with
   10 μmol/L PD treatment (p < 0.05) and further increased to 9.80 ± 1.16%
   with 30 μmol/L PD (p < 0.0001; Fig. [90]7c). The pathological
   morphologies of SKM-1 cells in each group were shown in Fig. [91]7d. In
   both the 10 μmol/L group and 30 μmol/L group, apoptotic SKM-1 cells
   were identified. Wright staining revealed that the apoptotic cell had a
   reduced size, pyknotic nuclei, and was accompanied by apoptotic bodies.

Fig. 7.

   [92]Fig. 7
   [93]Open in a new tab

   PD increases the proportions of apoptotitic SKM-1 and MUTZ-1 cells
   (n = 3). (A) Representative flow cytometry images. Apoptotic cell
   proportions in SKM-1 (B) and MUTZ-1 cells (C) following treatment with
   PD (0, 10, and 30 µmol/L) for 48 h. The values are expressed as the
   means ± SEMs. *p < 0.05, ****p < 0.0001. PI, propidium iodide. (D)
   Wright-stained images highlighting apoptotic cells (indicated by “↓”)
   observed at 10×100 magnification.

Discussion

   In this study, genome-wide DMGs were identified in 17 MDS patients
   compared with 14 healthy donors utilizing data from the GEO online
   database and revealed several aberrantly hypomethylated genes.
   Bioinformatic analysis revealed that various cancer related pathways
   such as the PI3K-AKT signaling pathway and the MAPK signaling pathway
   were significantly enriched in these hypomethylated genes. The PPI
   network revealed hub hypomethylated genes including GNAS, PIK3CA, and
   RAP1. These pathways and hub genes play critical roles in MDS. For
   example, activation of the PI3K-AKT signaling pathway has been shown in
   CD34 + cells from MDS patients^[94]14. Taken together, the data from
   the GEO database strongly indicate that drugs targeting aberrant
   hypomethylation, distinct from present hypomethylating agents, may be
   essential for future MDS treatment, particularly in patients
   unresponsive to currently employed hypomethylating therapies.

   Targeting aberrant hypermethylation with DNA hypomethylating agents
   (HMAs) has been a significant advancement in the treatment of MDS.
   However, the failure of HMAs remains common, and there are no curative
   therapies outside of allogeneic hematopoietic stem cell
   transplantation^[95]15,[96]16. Considering the key role of abnormal
   hypomethylation in MDS, novel drugs for DNA hypomethylation-targeted
   therapy may be needed to improve the clinical efficacy of MDS
   treatment. Interestingly, the present study indicated that PD acts as a
   methylation-promoting agent. Human methylation chip analyses identified
   623 DMGs following PD treatment, of which 448 genes (71.91%) were
   hypermethylated. These hypermethylated genes were associated primarily
   with pathways such as the PI3K-AKT signaling pathway, cGMP-PKG
   signaling pathway, and glutamatergic pathway. DNA methylation involves
   the covalent addition of a methyl group from S-adenosylmethionine (SAM)
   to cytosine, which is mainly mediated by DNMTs^[97]17. To further
   investigate the mechanism of PD hypermethylation, WB was performed to
   assess the effect of PD on three DNMTs protein expression in two MDS
   cell lines. In SKM-1 cells, compared with DMSO treatment, PD treatment
   effectively increased the expression of DNMT1, DNMT3a and DNMT3b. In
   MUTZ-1 cells, PD treatment significantly elevated DNMT1 and DNMT3b
   protein levels. These findings suggest that PD promotes genome-wide DNA
   methylation by increasing DNMTs expression, highlighting its potential
   as an innovative hypermethylation agent.

   Hypermethylation in promoter regions suppresses gene expression,
   whereas hypomethylation can lead to gene overexpression. In this study,
   we applied Cytoscape analysis to determine the core hypermethylated
   genes induced by PD. The results revealed that PIK3CA, ITPR3 and SPOPL
   were key genes in the network, all of which are classified as
   oncogenes. PIK3CA is a key molecule in the PI3K-AKT signaling pathway
   and plays a pivotal role in tumor cell proliferation, apoptosis,
   migration and blood vessel generation. PIK3CA is well known as an
   oncogene and recognized as a potential therapeutic target in multiple
   cancers^[98]18–[99]20. Similarly, ITPR3 is also a driver oncogene,
   frequently overexpressed in cancers like glioblastoma, squamous cell
   carcinoma, colorectal cancer and cholangiocarcinoma. High expression of
   ITPR3 is associated with adverse clinical outcomes, making it an
   attractive therapeutic target in cancer treatment^[100]21–[101]24.
   Notably, WB analysis revealed that PD treatment inhibited the protein
   expression level of these three oncogenes in SKM-1 cells. These
   findings suggest that PD promotes the DNA methylation of oncogenes,
   subsequently downregulating their expression, which may represent the
   primary mechanism underlying its therapeutic effects.

   In malignant diseases, tumor cells are abnormally active, apoptosis is
   inhibited, and the overexpression of oncogenes can promote
   proliferation and inhibit apoptosis. For example, PIK3CA can regulate
   cell survival and proliferation by influencing cell apoptosis^[102]25.
   Mutations in PIK3CA often lead to aberrant activation of the PI3K/AKT
   signaling pathway, which contributs to cancer progression by
   suppressing apoptosis and promoting cell survival^[103]26. ITPR3
   interacts with proteins such as PTEN and STAT3 to regulate endoplasmic
   reticulum–mitochondria Ca^2+ flux, preventing excessive calcium
   transfer, which can induce apoptosis^[104]27. SPOPL overexpression in
   glioma stem cells promotes tumorigenicity and stemness by activating
   the Notch signaling pathway^[105]28. Our CCK-8 assay results
   demonstrated PD could notably inhibit the proliferation of MDS cell
   lines in a dose‒dependent manner. Moreover, PD treatment significantly
   decreased the G0/G1 phase cell proportion while increasing that in the
   S phase. Additionally, apoptosis detection revealed that PD treatment
   markedly increased the proportion of apoptotic SKM-1 and MUTZ-1 cells.
   These findings suggest that PD may have therapeutic effects by
   downregulating the expression of PIK3CA, ITPR3 and SPOPL.

   In summary, our findings demonstrated that hypomethylated genes play a
   significant role in pathological progression in MDS patients. PD
   induced the methylation of PIK3CA, ITPR3 and SPOPL, decreased the
   protein expression of these oncogenes, and induced apoptosis in MDS
   cells. Furthermore, increased protein expression of DNMTs was observed
   in SKM-1 and MUTZ-1 cells, indicating that PD may be a novel
   hypermethylation agent. In the future, clinical trials are needed to
   confirm the improved survival rates and therapeutic efficacy of PD in
   MDS patients.

Supplementary Information

   [106]Supplementary Information 1.^ (697KB, pdf)
   [107]Supplementary Information 2.^ (15.9KB, xlsx)
   [108]Supplementary Information 3.^ (10.8KB, xlsx)

Acknowledgements