Paeoniflorin alleviates cancer-related fatigue during chemotherapy for breast cancer by targeting EZH2/CCNE1 to regulate cell cycle and inflammatory microenvironment

doi:10.12122/j.issn.1673-4254.2026.08.11

Abstract

Abstract:

Objective To explore the molecular mechanisms and key pharmacodynamic basis of Baishao for alleviating cancer-related fatigue (CRF) during chemotherapy for breast cancer. Methods Bioinformatics analyses were used to explore the active ingredients and potential targets of Baishao, CRF-related targets, and the differentially expressed core genes between chemotherapy and non-chemotherapy groups. Four machine learning algorithms (Random Forest, SVM, XGBoost, and GLM) were used to screen the key feature genes to construct a diagnostic nomogram model with subsequent survival and immunohistochemical analyses using Kaplan-Meier Plotter and HPA databases. In an IL-17-induced Py230 breast cancer cell model of CRF, the regulatory effects of paeoniflorin (a major active ingredient of Baishao) on the core targets were validated using CCK-8 assay, qRT-PCR, Western blotting, and immunofluorescence staining. Results Thirteen active ingredients (including paeoniflorin, albiflorin, and kaempferol) and 475 drug targets of Baishao were identified, yielding 65 core intersection genes enriched in the MAPK signaling cascade, circadian rhythm, and cell cycle regulation and showing significant correlations with immune cells. The SVM model demonstrated the best diagnostic performance and identified PSMB8, EZH2, CCNE1, PSEN2, and CDK1 as the key feature genes. The SVM-based nomogram achieved a C-index of 0.911, showing excellent calibration and clinical net benefit. Molecular docking confirmed strong binding affinities between the core ingredients of Baishao and the key targets. Survival analysis linked high EZH2 and CCNE1 expressions to poor breast cancer prognosis. In IL-17-induced Py230 cells, chemotherapy resulted in significant upregulation of EZH2 and CCNE1 expressions, which were effectively reversed by paeoniflorin treatment with an efficacy comparable to specific positive inhibitors. Conclusion Paeoniflorin alleviates chemotherapy-induced CRF in breast cancer by targeting and inhibiting abnormal expressions of EZH2 and CCNE1 to regulate cell cycle and inflammatory microenvironment, highlighting the therapeutic potential of Baishao for attenuating toxicity and enhancing efficacy of chemotherapy.

Key words: Baishao, breast cancer, cancer-related fatigue during chemotherapy, integrative pharmacology, machine learning

Zhuang LI, Huan SHI, Jieting CHEN, Yuqi LIANG, Yingchao WU, Qian ZUO, Qianjun CHEN. Paeoniflorin alleviates cancer-related fatigue during chemotherapy for breast cancer by targeting EZH2/CCNE1 to regulate cell cycle and inflammatory microenvironment[J]. Journal of Southern Medical University, 2026, 46(8): 1835-1849.

Figures/Tables 15

Tab.1 Primer sequence for RT-qPCR

Gene	Forward primer (5'-3')	Reverse primer (5'-3')
GAPDH	GGAGTCCACTGGTGTCTTCA	GGAGTCCACTGGTGTCTTCA
EZH2	CTGGTGGAAGAGCTGGAAGA	TGGTGCAGAGGAATAGAGCC
CCNE1	AAGGAGCGGGACACCATGA	ACGGTCACGTTTGCCTTCC

Fig.1 Targets of Baishao and the disease. A: Botanical image of Paeonia lactiflora. B: Cross-section of Paeonia lactiflora root. C: Venn diagram.

Fig.2 "Herb-Compound-Core Target" network.

Fig.3 Protein-protein interaction (PPI) network diagram.

Fig.4 Differential expression analysis of the core genes. A: Differential gene expression. B: Heatmap of the core genes.

Fig.5 GO and KEGG enrichment dot plot.

Fig.6 Expression characteristics and co-expression interaction network of core genes. A: Circos plot of the core genes. B: Chord diagram of correlations among the core gene. C: Correlation heatmap of the core genes.

Fig.7 Infiltration characteristics of the immune microenvironment and immune correlation analysis of the core targets. A: Stacked bar plot of immune cell infiltration proportions. B: Boxplot for comparison of immune cell abundance between CRF and control groups. C: Correlation heatmap between the core genes and immune cells. *P<0.05.

Fig.8 Construction of the predictive model based on the key feature genes and performance validation of the nomogram. A: Comparison of ROC curves for diagnostic models. B: Nomogram for predicting the risk of CRF. C: Calibration curve. D: Decision curve analysis.

Fig.9 Molecular docking analysis of the active ingredients and key targets. A: Binding energy heatmap. B: Visualization of molecular docking results.

Fig.11 Comparison of immunohistochemical stainingof the core proteins between breast cancer tissues andnormal breast tissues (n=3, scale bar=100 μm).**P<0.01,*** .P<0.001,

Fig.12 Boxplot for comparison of the core gene mRNA expressions between breast cancer and normal breast tissues (tumor=1085, normal=291). *P<0.05.

Fig.13 Effects of paeoniflorin on viability of Py230 cells and mRNA expressions of the core genes (n=3). A: Relative cell viability detected by CCK-8 assay after treatment with different concentrations of paeoniflorin for 24 h. B: Relative mRNA expression levels of EZH2 in each group. The Model group was induced by chemotherapeutic agents, and GSK126 served as a positive EZH2 inhibitor. C: Relative mRNA expression levels of CCNE1 in each group. SCH727965 served as a positive CCNE1 inhibitor. *P<0.05, **P<0.01.

Fig.14 Comparison of protein expression levels in breast cancer cells in each group (n=3). *P<0.05, **P<0.01, ***P<0.001.

Fig.15 Regulatory effects of paeoniflorin on protein expressions and localization of the core targets CCNE1 and EZH2 in breast cancer cells (scale bar=20 μm). A: CCNE1 expression levels in breast cancer cell groups. B: EZH2 expression levels in breast cancer cell groups.

References 37

[1]	Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025[J]. CA Cancer J Clin, 2025, 75(1): 10-45. doi：10.3322/caac.21871
[2]	中国抗癌协会癌症康复与姑息治疗专业委员会,中国临床肿瘤学会肿瘤支持与康复治疗专家委员会．癌症相关性疲乏诊断与治疗中国专家共识[J]. 中华医学杂志, 2022, 102(3): 180-9． doi：10.3760/cma.j.cn112137-20210811-01789
[3]	马永芳. 基于IMB模型的结肠癌化疗期患者癌因性疲乏护理方案的构建和应用研究[D]. 北华大学, 2025.
[4]	梁粲. 百笑灸联合参芪扶正注射液治疗肠癌化疗患者癌症相关性疲乏的临床研究[D]. 南京中医药大学, 2024.
[5]	Feng XL, Li ZH, Guo WH, et al. The effects of traditional Chinese medicine and dietary compounds on digestive cancer immunotherapy and gut microbiota modulation: a review[J]. Front Immunol, 2023, 14: 1087755. doi：10.3389/fimmu.2023.1087755
[6]	孙心悦, 王宽宇, 王钢, 等. 扶正消岩颗粒通过调控AKT1/BAD/BCL-2通路改善乳腺癌化疗期癌因性疲乏[J]. 南方医科大学学报,2025, 45(12): 2646-57.
[7]	王秋艳, 王世新, 隋方宇, 等. 白芍活性成分、药理作用及成分变化的影响因素研究进展[J]. 中草药, 2025, 56(5): 1817-29. doi：10.7501/j.issn.0253-2670.2025.05.030
[8]	Zhang P, Wang ZX, Qiu HX, et al. Machine learning applied to serum and cerebrospinal fluid metabolomes revealed altered arginine metabolism in neonatal sepsis with meningoencephalitis[J]. Comput Struct Biotechnol J, 2021, 19: 3284-92. doi：10.1016/j.csbj.2021.05.024
[9]	李鑫, 方崇锴, 黄越, 等. 参芪固本方调控IL-17信号通路改善癌因性疲乏的作用机制[J]. 中国药房, 2025, 36(14): 1722-9.
[10]	朱潇旭. 基于纹状体A2AR-ERK-NF-κB通路探讨逍遥散对肝郁脾虚证抑郁大鼠的神经保护作用机理[D]. 湖北中医药大学, 2022.
[11]	徐方宁, 刘清珍, 张悦, 等. 基于TRPV1-NLRP3通路探讨芍药苷缓解神经病理性疼痛的作用机制[J]. 医学研究与战创伤救治, 2025, 38(10): 1035-41.
[12]	谢晓燕, 张娟, 向勇, 等. 芍药苷可逆转疼痛共病抑郁大鼠海马内脑源性神经营养因子含量的减少[J]. 中华中医药学刊, 2022, 40(12): 181-5, 308.
[13]	Li CC, Liu B, Xu JY, et al. Phloretin decreases microglia-mediated synaptic engulfment to prevent chronic mild stress-induced depression-like behaviors in the mPFC[J]. Theranostics, 2023, 13(3): 955-72. doi：10.7150/thno.76553
[14]	牛永强, 江涛, 秦雪梅, 等. 蛋白质组学视角下的中药抗抑郁作用研究进展[J]. 中草药, 2025, 56(23): 8816-24.
[15]	Ye YH, Pei HR, Cao XL, et al. The study of a novel paeoniflorin-converting enzyme from Cunninghamella blakesleeana [J]. Molecules, 2023, 28(3): 1289. doi：10.3390/molecules28031289
[16]	Zeng J, Xu H, Fan PZ, et al. Kaempferol blocks neutrophil extracellular traps formation and reduces tumour metastasis by inhibiting ROS-PAD4 pathway[J]. J Cellular Molecular Medi, 2020, 24(13): 7590-9. doi：10.1111/jcmm.15394
[17]	He YL, Li J, Gong SH, et al. BNIP3 phosphorylation by JNK1/2 promotes mitophagy via enhancing its stability under hypoxia[J]. Cell Death Dis, 2022, 13(11): 966. doi：10.1038/s41419-022-05418-z
[18]	Arunachalam S, Nagoor Meeran MF, Azimullah S, et al. α-bisabolol attenuates doxorubicin induced renal toxicity by modulating NF-κB/MAPK signaling and caspase-dependent apoptosis in rats[J]. Int J Mol Sci, 2022, 23(18): 10528. doi：10.3390/ijms231810528
[19]	Montero P, Milara J, Pérez-Leal M, et al. Paclitaxel-induced epidermal alterations: an in vitro preclinical assessment in primary keratinocytes and in a 3D epidermis model[J]. Int J Mol Sci, 2022, 23(3): 1142. doi：10.3390/ijms23031142
[20]	Ekedahl H, Isaksson S, Ståhl O, et al. Low-grade inflammation in survivors of childhood cancer and testicular cancer and its association with hypogonadism and metabolic risk factors[J]. BMC Cancer, 2022, 22(1): 157. doi：10.1186/s12885-022-09253-5
[21]	Bower JE. Cancer-related fatigue: mechanisms, risk factors, and treatments[J]. Nat Rev Clin Oncol, 2014, 11(10): 597-609. doi：10.1038/nrclinonc.2014.127
[22]	Boyle CC, Bower JE, Eisenberger NI, et al. Stress to inflammation and anhedonia: Mechanistic insights from preclinical and clinical models[J]. Neurosci Biobehav Rev, 2023, 152: 105307. doi：10.1016/j.neubiorev.2023.105307
[23]	Wang MM, Zhong B, Li M, et al. Identification of potential core genes and pathways predicting pathogenesis in head and neck squ-amous cell carcinoma[J]. Biosci Rep, 2021, 41(5): BSR20204148. doi：10.1042/bsr20204148
[24]	Matthews HK, Bertoli C, de Bruin RAM. Cell cycle control in cancer[J]. Nat Rev Mol Cell Biol, 2022, 23(1): 74-88. doi：10.1038/s41580-021-00404-3
[25]	Wang J, Zhou H. Mitochondrial quality control mechanisms as molecular targets in cardiac ischemia-reperfusion injury[J]. Acta Pharm Sin B, 2020, 10(10): 1866-79. doi：10.1016/j.apsb.2020.03.004
[26]	龙腾飞, 张红雁, 吴爱林, 等. 基于CGGA数据库的325例脑胶质瘤患者PSMB8表达水平与临床特征的相关性研究[J]. 安徽医科大学学报, 2021, 56(11): 1833-7. doi：10.19405/j.cnki.issn1000-1492.2021.11.031
[27]	Pudova EA, Pavlov VS, Guvatova ZG, et al. Machine learning models for cancer research: a narrative review of bulk RNA-seq applications[J]. Int J Mol Sci, 2025, 26(24): 12081. doi：10.3390/ijms262412081
[28]	Wang FL, Zain AM, Ren YJ, et al. Navigating the microarray landscape: a comprehensive review of feature selection techniques and their applications[J]. Front Big Data, 2025, 8: 1624507. doi：10.3389/fdata.2025.1624507
[29]	Jiang L, Xu C, Bai Y, et al. Autosurv: interpretable deep learning framework for cancer survival analysis incorporating clinical and multi-omics data[J]. NPJ Precis Oncol, 2024, 8(1): 4. doi：10.1038/s41698-023-00494-6
[30]	Zahedi R, Ghamsari R, Argha A, et al. Deep learning in spatially resolved transcriptomics: a comprehensive technical view[J]. Brief Bioinform, 2024, 25(2): bbae082. doi：10.1093/bib/bbae082
[31]	He J, Song YJ, Li GP, et al. Fbxw7 increases CCL2/7 in CX3CR1hi macrophages to promote intestinal inflammation[J]. J Clin Investig, 2019, 129(9): 3877-93. doi：10.1172/jci123374
[32]	Yoo KH, Oh S, Kang K, et al. Loss of EZH2 results in precocious mammary gland development and activation of STAT5-dependent genes[J]. Nucleic Acids Res, 2015, 43(18): 8774-89. doi：10.1093/nar/gkv776
[33]	Lv Q, Xing Y, Liu J, et al. Lonicerin targets EZH2 to alleviate ulcerative colitis by autophagy-mediated NLRP3 inflammasome inactivation[J]. Acta Pharm Sin B, 2021, 11(9): 2880-99. doi：10.1016/j.apsb.2021.03.011
[34]	Das A, Ranadive N, Kinra M, et al. An overview on chemotherapy-induced cognitive impairment and potential role of antidepressants[J]. Curr Neuropharmacol, 2020, 18(9): 838-51. doi：10.2174/1570159x18666200221113842
[35]	Liu YC, Feng N, Li WW, et al. Costunolide plays an anti-neuroinflammation role in lipopolysaccharide-induced BV2 microglial activation by targeting cyclin-dependent kinase 2[J]. Molecules, 2020, 25(12): 2840. doi：10.3390/molecules25122840
[36]	Ma JF, Meng QL, Zhan JP, et al. Paeoniflorin suppresses rheumatoid arthritis development via modulating the circ-FAM120A/miR-671-5p/MDM4 axis[J]. Inflammation, 2021, 44(6): 2309-22. doi：10.1007/s10753-021-01504-0
[37]	Kim KH, Roberts CWM. Targeting EZH2 in cancer[J]. Nat Med, 2016, 22(2): 128-34. doi：10.1038/nm.4036