1 Introduction
Breast cancer remains the leading cancer in incidence and mortality among women, with nearly 2.3 million new cases and 670 000 deaths globally in 2022 [
1]. Human epidermal growth factor receptor-2 (HER2) overexpression, observed in 20%–30% of breast cancer cases, is strongly associated with poor prognosis [
2,
3]. Anthracyclines and trastuzumab are essential treatments for HER2-positive breast cancer, significantly improving disease-free survival (DFS) and overall survival (OS) [
4,
5]. However, the sequential use of doxorubicin (DOX) and trastuzumab (TRZ) is linked to high cardiotoxicity rates, reaching up to 25% [
6–
8]. The mechanisms underlying DOX/TRZ-induced cardiotoxicity have not yet been fully elucidated. Despite the use of cardioprotective agents like dexrazoxane and β-blockers, current therapeutic strategies show limited efficacy, underscoring the need for novel interventions.
In recent years, traditional Chinese medicine (TCM) has played a growing role in preventing and treating antitumor drug-induced adverse reactions, particularly cardiotoxicity. For instance, Zhigancao Decoction improves DOX-induced declines in cardiac ejection function and reduces the incidence of electrocardiographic abnormalities in patients [
9]. Wenxin Granules, a proprietary Chinese medicine, can be combined with dexrazoxane to prevent anthracycline-induced cardiotoxicity in clinical settings. Current research primarily focuses on anthracycline-related cardiotoxicity, with limited studies investigating sequential strategies involving anthracyclines and molecularly targeted drugs.
Qizaobaoxin Decoction (QZBXD) is a clinical experience-based Chinese herbal formula developed with modifications from the classical Shengyu Decoction based on clinical symptoms of cardiotoxicity. QZBXD has been empirically used in our clinical practice to manage cardiotoxicity induced by DOX/TRZ. It consists of seven herbs:
Astragalus membranaceus,
Codonopsis Radix,
Cmnamomi Mmulus,
Angelica sinensis,
Salvia miltiorrhiza,
Chuanxiong Rhizoma, and Spine Date Seed. Many individual components of QZBXD, such as
Astragalus membranaceus,
Salvia miltiorrhiza, and
Chuanxiong Rhizoma, are known to exert diverse cardioprotective effects through various pathways, including autophagy inhibition, reduction of inflammatory cytokines, and resistance to oxidative stress [
10–
14]. Therefore, we hypothesized that co-administration of QZBXD could provide additional protection against DOX/TRZ-induced cardiotoxicity in a rat model. Despite its promising clinical application, QZBXD has a complex chemical composition, with its active substances and precise mechanisms of action remaining largely unclear, which significantly limits its broader clinical promotion and application. With the rapid advancement of multi-omics technologies, these approaches offer powerful tools for investigating the mechanisms of TCM. Based on this, we first developed a UPLC-Q-TOF/MS-based qualitative analysis method to preliminarily identify the chemical components of QZBXD. Through comprehensive assessments of myocardial injury markers, cardiac enzyme profiles, echocardiography, and pathological staining, the cardioprotective effects of QZBXD against DOX/TRZ-induced cardiotoxicity were clarified. Subsequently, 16S rDNA sequencing and untargeted metabolomics were employed to preliminarily investigate its underlying mechanisms.
2 Materials and methods
2.1 Chemicals and materials
DOX and TRZ were purchased from Main Luck Pharmaceuticals Inc. (Shenzhen, China) and Henlius Biotech, Inc. (Shanghai, China), respectively. Rat brain natriuretic peptide (BNP) and cardiac troponin I (cTnI) ELISA Kits were supplied by Enzyme-Linked Biotechnology Co., Ltd. (Shanghai, China). All remaining reagents and chemicals were obtained from commercial suppliers.
2.2 QZBXD preparation
The herbal decoction QZBXD is composed of seven medicinal herbs, including Astragalus membranaceus, Codonopsis Radix, Cmnamomi Mmulus, Angelica sinensis, Salvia miltiorrhiza, Chuanxiong Rhizoma, and Spine Date Seed. The herbs were purchased from Sinopharm Beijing Huamiao Pharmaceutical Co., Ltd. After identification by Professor Hongzhu Li (Department of Pharmacy, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital), the extraction procedures for QZBXD were as follows: (1) the crude herb was mixed with deionized water (10-fold volume) and steeped for 30 min, then extracted with boiling water for 60 min; (2) the residue was extracted with boiling water (8-fold volume) for 60 min and filtered; (3) the two filtrates were mixed and concentrated so that the final liquid contained 1.72 g/mL of the raw herb.
2.3 Component analysis of QZBXD
Take 1 mL of the concentrate, add methanol dropwise to achieve the final methanol concentration of 50% (v/v), shake gently, and incubate overnight at 4 °C. Before analysis, centrifuge the sample at 4800 rpm for 10 min, then collect the supernatant and filter it through a 0.45 μm microporous membrane. The mobile phase contained eluent A (acetonitrile) and eluent B (0.1% formic acid in water). Detailed analysis procedures and standard information are provided in Table S1.
2.4 Animal experimental design and sample collection
Female Sprague-Dawley rats, aged 8 weeks and weighing 180–200 g, certificate number: SCXK (Beijing) 2019-0010, were procured from SiPeiFu Biological Technology Co., Ltd. (Beijing, China). All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee of the Chinese Academy of Medical Sciences Cancer Hospital. Prior to experiments, animals were housed in an environmentally controlled room at a constant temperature of 23 ± 1 °C, with a 12 h light/dark cycle. Thirty rats were randomly assigned to five groups (
n = 6 in each group). The animal model of cardiotoxicity was established based on previous studies [
15,
16]. Except for the control group, each rat received 6 intravenous doses of Dox (cumulative dose = 15 mg/kg), followed by 6 intraperitoneal doses of TRZ (cumulative dose = 20 mg/kg). Two cohorts of rats were administered QZBXD at doses of 30.7 g/kg and 15.4 g/kg daily, which were determined based on clinical dose conversion, with the low and high doses corresponding to 2 and 4 times the clinical dose, respectively. In addition, we chose dexrazoxane as a positive drug. The positive control drug, dexrazoxane, was administered intraperitoneally at a dose equivalent to 10 times that of the DOX solution, 1 h prior to each DOX administration. The schematic of the experimental protocol is depicted in Fig. S1. Before the end of the experiment, following echocardiographic assessment, all rats were euthanized under isoflurane anesthesia, and blood samples were collected from the abdominal aorta.
2.5 Echocardiography, determination of biochemical indicators and histopathology changes
Echocardiography was conducted using a VisualSonics (VINNO 6 VET) equipped with an 18-MHz linear probe. Measurements included the left ventricular internal dimension in diastole (LVIDd) and the left ventricular internal dimension in systole (LVIDs). Left ventricular ejection fraction (LVEF) and left ventricular fractional shortening (LVFS) were calculated using the appropriate formulas. Lactate dehydrogenase (LDH), creatine kinase isoenzyme (CK-MB), and α-hydroxybutyrate dehydrogenase (α-HBDH) levels in plasma from rats in each group were measured using a Beckman Coulter AU 5800 (Beckman Coulter, Brea, USA). Plasma concentrations of cTnI and BNP were assessed by ELISA according to the manufacturer’s instructions. Heart tissue samples were fixed in 10% neutral buffered formalin for at least 72 h, then paraffin-embedded, sectioned into 4-μm thick slices, and stained with hematoxylin and eosin (HE).
2.6 16S rDNA gene sequencing
Five samples were randomly collected from each group. In this experiment, the genomic DNA of the samples was extracted using the CTAB method, and the V3–V4 region of the 16S rDNA gene was amplified by PCR, followed by electrophoresis detection and purification. The amplified products were then used to construct a library, which was sequenced using the NovaSeq 6000 platform. The sequencing data were processed with fastp (v0.22.0) for quality control, FLASH (v1.2.11) for read merging, and vsearch (v2.22.1) to remove chimeric sequences. The final valid sequences were denoised using the DADA2 plugin (v1.26.0) in QIIME2, and species taxonomy was assigned using the naive Bayes classifier.
2.7 Plasma metabolomics analysis and metabolite identification
The sample was thawed on ice and vortexed for 10 s. Then, 50 μL of the sample was combined with 300 μL of extraction solution (acetonitrile:methanol = 1:4, v/v). The mixture was vortexed for 3 min and centrifuged at 12 000 rpm (4 °C) for 10 min. After centrifugation, 200 μL of the supernatant was collected, placed at −20 °C for 30 min, and then centrifuged at 12 000 rpm for 3 min (4 °C). A 180 μL aliquot of the supernatant was transferred for LC-MS analysis.
The LC-ESI-MS/MS system (LC-30A, Shimadzu; TripleTOF 6600+, SCIEX) was used for analysis. A C18 column (Waters ACQUITY UPLC BEH C18, 1.8 µm, 2.1 mm × 100 mm) was employed, with the column temperature maintained at 40 °C. The product ion scan parameters were set as follows: mass range, 50–1000 Da. The gradient program is provided in Table S2. Information-dependent acquisition (IDA) mode using Analyst TF 1.7.1 software (SCIEX, Concord, ON, Canada) was used for data acquisition. The source parameters were set as follows: source temperature, 550 °C; ion spray voltage floating (ISVF), 5000 V in positive mode or –4500 V in negative mode; ion source gas 1 (GS1), gas 2 (GS2), and curtain gas (CUR) were set to 35, 50, and 35 psi, respectively.
2.8 Data processing and analysis
The mass data were normalized using MetaboAnalyst 6.0 and then the data matrix was introduced into SIMCA-P 14.1 software (Umetrics, Umea, Sweden) for multivariate analysis. For two-group analysis, differential metabolites were determined by variable importance in projection (VIP) values (VIP > 1) and P-values (P < 0.05, Student’s t-test). Subsequently, the differential metabolites were introduced into MetaboAnalyst 6.0 to explore the metabolic pathways according to the Rattus norvegicus pathway library.
2.9 Correlation analysis
To verify the association between two consecutive variables, correlation analysis was applied to analyze the relevance between gut microbiota and the differential metabolites selected in this study as well as the cardiotoxicity indicators. Given the non-normal distribution of the original data, Spearman rank correlation coefficient was used for correlation analysis. P < 0.05 indicates correlation between variables.
2.10 Statistical analysis
All data are expressed as the mean ± standard deviation (SD). Statistical analyses between two groups were performed using Student’s t-tests. One-way ANOVA with Dunnett’s post hoc multiple comparisons was used for multiple comparisons (GraphPad Prism 9.0 software, San Diego, CA, USA). P < 0.05 was considered statistically significant.
3 Results
3.1 Identification of QZBXD
The QZBXD chemical composition database was constructed with reference to databases such as CNKI, PubMed, and SciFinder. The database includes key information such as the component names, structural formulas, relative molecular weights, and characteristic fragments of the components. Representative total ion chromatograms in positive and negative ion modes are shown in Fig. 1A. The precise molecular weights and fragmentation data of the compounds under positive and negative ion modes, obtained using SCIEX OS-Q1.6 software, were compared with the established database of QZBXD constituents and relevant literature. The cleavage processes of the compounds were deduced based on both first-order and second-order mass spectrometry data for individual compounds to identify the constituents. A total of 83 compounds were preliminary identified (as shown in Table S3), including 25 flavonoids, 23 organic acids, 16 phthalides, 8 saponins, 6 diterpenes, 2 alkaloids, and 3 other types of compounds.
3.2 Flavonoids
The flavonoid components in QZBXD primarily originate from Astragalus membranaceus, which has a wide range of pharmacological activities and is the most abundant component in QZBXD, potentially acting as an active ingredient. Taking calycosin-7-O-β-D-glucoside as an example, it was identified by its quasi-molecular ion peak at m/z 447.1285 ([M + H]+). The ion at m/z 285.0768 ([M + H-Glc]+) was formed by the loss of a glycosyl fragment, followed by the further loss of CH3 and CH3OH, resulting in the characteristic fragments at m/z 270.0521 ([M + H-Glc-CH3]+) and m/z 253.0487 ([M + H-Glc-CH3OH]+). The putative fragmentation pathway of salvianolic acid B is shown in Fig. 1B.
3.3 Organic acids
A total of 23 organic acids were identified in QZBXD, most of which are phenolic acids derived from Salvia miltiorrhiza. Salvianolic acid B produces a quasi-molecular ion peak at m/z 717.1505 ([M-H]–). The fragmentation process likely begins with the loss of a C9H10O5 group, producing the ion at m/z 519.0976 ([M-H-C9H10O5]–). From this intermediate, the loss of a second C9H10O5 group yields the ion at m/z 321.0417 ([M-H-2C9H10O5]–); notably, this ion may also arise from the simultaneous loss of two C9H10O5 groups directly from the parent ion. The m/z 321.0417 species then further loses a C2H2O group to yield m/z 279.0317 ([M-H-2C9H10O5-C2H2O]–). In a parallel pathway originating from m/z 519.0976, the elimination of a C9H8O4 group leads to m/z 339.0528 ([M-H-C9H10O5-C9H8O4]–), followed by the subsequent loss of CO2 to produce m/z 295.0629 ([M-H-C9H10O5-C9H8O4-CO2]–) (Fig. 1C).
3.4 Diterpenes
Tanshinone IIA, a significant diterpene compound in Salvia miltiorrhiza, was easily ionized in positive ion mode with proton adduction, yielding m/z 295.1330 ([M + H]+). The primary fragmentation involves the loss of a water molecule, generating a fragment ion at m/z 277.1230 ([M + H-H2O]+). This is followed by the loss of a carbon monoxide molecule, producing the characteristic fragment at m/z 249.1282 ([M + H-H2O-CO]+). Further fragmentation includes the loss of a methyl group (-CH3), resulting in a fragment ion at m/z 234.1025 ([M + H-H2O-CO-CH3]+) (Fig. 1D).
3.5 Phthalides
Phthalide compounds, which are major active components of Angelica sinensis and Chuanxiong Rhizoma, primarily include phthalide monomers and dimers. Taking senkyunolide, a major component in Chuanxiong Rhizoma, as an example, it is prone to be deprotonated as [M + H]+, as shown the peak of m/z 193.1222. The peak of m/z 175.1121 represented the fragment ion ([M + H-H2O]−), produced by the loss of a water molecule from the parent ion. A further loss of CO2 molecule occurred, resulting in the product ion at m/z 147.1162 (M + H-H2O-CO]−) (Fig. 1E).
3.6 Effect of QZBXD on DOX/TRZ-induced cardiotoxicity
We evaluated the protective effects of QZBXD on cardiotoxicity induced by DOX and the sequential administration of TRZ using biomarkers, echocardiography, and histologic examination. Fig. 2A shows the comparison of body weight across the groups throughout the experiment. Compared with the control group, rats in the model group exhibited poor weight gain throughout the experiment, particularly during the early stages. Additionally, ascites occurred in some rats in the model group, a sign of heart failure (HF). However, compared with the model group, the rats treated with QZBXD experienced slower weight loss.
Fig. 2B presents the levels of cTnI, BNP, CK-MB, LDH, and α-HBDH in serum across the groups. DOX/TRZ exposure significantly increased these levels compared to the control group (P < 0.01). However, oral treatment of QZBXD with 30.7 g/(kg·d) of significantly attenuated the increases in serum cTnI, BNP, CK-MB, LDH, and α-HBDH levels.
Echocardiographic measurements were performed to assess changes in myocardial geometry and function in response to DOX/TRZ exposure and QZBXD treatment in rats (Fig. 2C and 2D). The echocardiography results showed that the higher dosage of QZBXD significantly rescued the DOX/TRZ-induced reduction in LVEF and LVFS (P < 0.01). Additionally, the higher dosage of QZBXD (30.7 g/kg) reversed the increase in LVIDs induced by DOX and TRZ (P < 0.05). As shown in Fig. 3, heart tissue structure in rats from the control group was clear, with cardiomyocytes evenly arranged and well-shaped. However, significant pathological changes, including disordered cardiac muscle arrangement and interstitial edema, were observed in rats treated with DOX and TRZ. These conditions were significantly improved in the QZBXD treatment group.
3.7 Effect of QZBXD on host metabolism in DOX/TRZ-induced cardiotoxicity
To reveal the metabolic phenotypes that may be involved in DOX/TRZ-induced cardiotoxicity, we performed metabolic profiling of plasma from each group. The overlaid QC chromatograms are presented in Fig. S2, which expanded on the reproducibility of this method and its results. The metabolic profiles were analyzed using the UHPLC system in both positive and negative ion modes. Principal component analysis (PCA) indicated significant differences in plasma metabolites among the 3 groups in both positive and negative ion modes (Fig. 4A). The tendency to separate the control group, the model group, and the QZBXD group was observed in both PCA score plots, which indicated metabolite difference in the above groups. Specifically, the control group was clustered together and relatively far from the model group; the QZBXD group also showed a trend of separation from the model, which indicated the trend of changing metabolomics profile after administration.
To observe the separation trend among different groups, we established an Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA) model. As shown in the OPLS-DA scores plots for data generated (Fig. 4B and 4C), the 3 groups could be well distinguished both in positive and negative ion modes. The parameters of the OPLS-DA and the results of 200 permutation tests indicated that the model is credible. The results in negative ion mode were shown in supplementary materials (Fig. S3).
The results of univariate analysis are also displayed in the form of volcano plots (Fig. 5A). In total, 175 metabolites were upregulated and 391 were downregulated in the model group compared with the control group, while 404 were upregulated and 239 were downregulated in the QZBXD group compared with the model group. As the Venn diagrams show, 299 common metabolites were obtained (Fig. 5B). Among them, 269 metabolites showed opposite trend in the model group versus the control group and in the QZBXD group versus the model group, suggesting that these metabolites may be associated with the protective effect of QZBXD closely.
To further understand the roles and functions of the above differential metabolites, pathway analysis was performed using MetaboAnalyst 6.0. With P < 0.05, impact > 0.1 as the screening condition, the significantly impacted pathways were tryptophan metabolism and arachidonic acid metabolism (Fig. 5C). Tryptamine, 4-hydroxy-L-tryptophan, L-formylkynurenine, 8,9-epoxyeicosatrienoic acid, 20-HETE, L-tryptophan, leukotriene A4, arachidonic acid, 5-methoxyindoleacetate, and indoleacetaldehyde were enriched. Information on the above differential metabolites is given in Table 1, and the raw intensity values obtained from the assays in each group are shown in Fig. S4.
3.8 Correlation analysis of efficacy indicators with differential metabolites
Correlation analysis of 10 key differential metabolites obtained from the screening with the efficacy indexes showed that all differential metabolites were correlated with the efficacy indexes. Specifically, 4-hydroxy-L-tryptophan, L-formylkynurenine, 20-HETE, L-tryptophan, leukotriene A4, arachidonic acid, 5-methoxyindoleacetate, and indoleacetaldehyde were negatively correlated with LVIDs, CK-MB, BNP, LDH, cTnI, and α-HBDH, while positively correlated with LVEF and FS%. In contrast, tryptamine, 8,9-epoxyeicosatrienoic acid were positively correlated with LVIDs, CK-MB, BNP, LDH, cTnI, and α-HBDH, and negatively correlated with LVEF and FS%. These findings suggest that elevated levels of tryptamine and 8,9-epoxyeicosatrienoic acid may be associated with decreased cardiac function.
3.9 QZBXD alters intestinal microbe diversity in DOX/TRZ-induced cardiotoxicity rats
To explore the potential involvement of the gut microbiota in mediating DOX/TRZ-induced cardiotoxicity and the regulation of microbiota composition by QZBXD, we performed 16s rDNA gene sequencing.
The rarefaction curve was constructed using the microbial alpha diversity (α-diversity) index of each sample at different sequencing depths to reflect the microbial diversity at various sequencing quantities. The results showed that the curve tended to flatten, suggesting that the amount of sequencing data was sufficient to capture most of the microbial diversity in the sample (Fig. S5). The ACE and Chao1 indices were used to reflect community species richness, and the observed_otus index was used to calculate diversity. The results showed that, compared with the control group, the Chao1 index, ACE index, and observed_otus index were significantly lower in the model group (P < 0.01), whereas these indices were significantly higher in the QZBXD group compared with the model group (P < 0.05) (Fig. S6).
By comparing and analyzing the species diversity among different microbial communities, beta diversity (β-diversity) analysis was performed to explore the similarity or differences in community composition. Principal coordinate analysis (PCoA) and non-metric multidimensional scaling (NMDS) analysis revealed a clear division among the control, model, and QZBXD groups, indicating that significant changes in gut microbiota composition occurred after QZBXD administration in DOX/TRZ-treated rats (Fig. 6A).
3.10 QZBXD alters the composition of intestinal flora at the multispecies level in DOX/TRZ treated rats
To study the specific changes in bacterial communities, we drew the cluster histograms to show the changes in intestinal flora at the phylum, family, and genus levels in each sample, as well as in the subgroups (Fig. 6B–6D). As shown in the figures, at the phylum level, Firmicutes and Bacteroidetes were dominant phyla in all groups. The relative abundance of Firmicutes increased and that of Bacteroidetes decreased in model group compared with the control group. Besides, the Firmicutes/Bacteroidetes relative abundance ratio of model group increased in our study, and QZBXD was able to decrease it, which was consistent with the control group. At the family level, compared with the control group, DOX/TRZ reduced the relative abundance of Bacteroidaceae while the relative abundance of Oscillospiraceae, Lachnospiraceae and Ruminococcaceae was significantly increased. Compared to model group, QZBXD increased the relative abundance of Bacteroidaceae and decreased the relative abundance of Oscillospiraceae, Lachnospiraceae and Ruminococcaceae. At the genus level, DOX/TRZ caused the relative abundance of Ruminococcus and Oscillibacter to significantly increase, while Bacteroides decreased. QZBXD treatment reduced Ruminococcus and Oscillibacter, while Bacteroides increased. The above results indicated that QZBXD treatment largely restructured the bacterial flora, and Bacteroides, Ruminococcus, and Oscillibacter may be the key regulators of the cardioprotective effects of QZBXD.
To further confirm which bacterium was altered by QZBXD treatment and in turn affected disease progression against DOX/TRZ-induced cardiotoxicity, we performed high-dimensional class comparisons using linear discriminant analysis (LDA) of effect size (LEfSe) and detected marked differences in bacterial community predominance between the different groups. The cladogram showed that 50 taxa were present in the 3 experimental groups, with significant variations in the compositions of intestinal bacterial communities among them. Furthermore, according to LDA (α = 0.05; LDA score > 3.5), the top 5 groups enriched by DOX/TRZ were Lachnospirales, Lachnospiraceae, Ruminococcus, o_unidentified Clostridia and g_Oscillibacter. QZBXD may play a cardioprotective role through Bacteroidales, Bacteroidia, Bacteroidota, Prevotellaceae, f_Muribaculaceae, g_unidentified Clostridia exerted a cardioprotective effect. Interestingly, in the control group, Bacteroides as well as Bacteroidaceae, were also significantly enriched. Collectively, these results showed that DOX/TRZ treatment significantly altered the gut microbiota diversity and composition, and QZBXD may prevent cardiotoxicity by changing the above situation (Fig. 7A and 7B).
3.11 Correlation analysis of efficacy indicators and intestinal flora
At the phylum level, Firmicutes and Bacteroidetes are correlated with all efficacy indicators. Specifically, Firmicutes is positively correlated with LDH, CK-MB, α-HBDH, BNP, cTnI, and LVIDs, while negatively correlated with LVEF and FS%. In contrast, Bacteroidetes exhibits the opposite trend. At the family level, Lachnospiraceae, Oscillospiraceae, Ruminococcaceae, Bacteroidaceae, and Tannerellaceae are correlated with all efficacy indicators. Similarly, at the genus level, Bacteroides, Ruminococcus, Oscillibacter, and Parabacteroides are correlated with all efficacy indicators. Among them, Ruminococcus and Oscillibacter are positively correlated with LDH, CK-MB, α-HBDH, BNP, cTnI, and LVIDs, but negatively correlated with LVEF and FS%. Conversely, Bacteroides and Parabacteroides show negative correlations with these indicators and positive correlations with LVEF and FS%. These results suggest that QZBXD may improve cardiac function in rats by increasing the abundance of beneficial bacteria and reducing the abundance of harmful bacteria (Fig. 8).
3.12 Correlation analysis of metabolomics and intestinal flora
At the phylum level, Firmicutes and Bacteroidetes were correlated with all differential metabolites. Specifically, Firmicutes showed positive correlations with tryptamine and 8,9-epoxyeicosatrienoic acid, and negative correlations with 4-hydroxy-L-tryptophan, L-formylkynurenine, 20-HETE, L-tryptophan, leukotriene A4, arachidonic acid, 5-methoxyindoleacetic acid, and indoleacetaldehyde. Bacteroidetes exhibited the opposite pattern, with negative correlations for tryptamine and 8,9-epoxyeicosatrienoic acid, and positive correlations for the remaining metabolites. At the family level, Lachnospiraceae, Oscillospiraceae, Ruminococcaceae, and Bacteroidaceae were correlated with all metabolites. Lachnospiraceae, Oscillospiraceae, and Ruminococcaceae showed positive correlations with tryptamine and 8,9-epoxyeicosatrienoic acid, while being negatively correlated with other metabolites. In contrast, Bacteroidaceae were negatively correlated with tryptamine and 8,9-epoxyeicosatrienoic acid, while showing positive correlations with the other metabolites. At the genus level, Bacteroides, Ruminococcus, and Oscillibacter were correlated with all metabolites. Ruminococcus and Oscillibacter showed positive correlations with tryptamine and 8,9-epoxyeicosatrienoic acid and negative correlations with the other metabolites, while Bacteroides showed an opposite trend. Overall, these results suggest that QZBXD-induced changes in the gut microbiota may alter tryptophan and arachidonic acid metabolism, potentially contributing to improved cardiac function (Fig. 9).
4 Discussion
The sequential of anthracyclines and the molecularly targeted drug TRZ remains a cornerstone in the treatment of HER2-positive breast cancer. However, its clinical utility is significantly limited by pronounced cardiotoxicity. Increasing clinical evidence suggests that TCM offers unique advantages in mitigating chemotherapy-induced cardiotoxicity. QZBXD, a clinically applied TCM formula, has demonstrated cardioprotective potential in practice. Nevertheless, due to the complexity of multi-component formulations and the lack of clearly defined mechanisms of action, the broader clinical application of TCM remains challenging. The identification and pharmacological characterization of bioactive components are essential for elucidating the therapeutic basis of TCM and promoting its modernization.
In this study, we established a qualitative analysis method for the chemical components of QZBXD using UPLC-Q-TOF-MS/MS in both positive and negative ion modes. A total of 83 compounds were preliminarily identified, including flavonoids, saponins, alkaloids, organic acids, diterpenes, and phthalides. Among these, components from
Astragalus membranaceus and
Salvia miltiorrhiza were particularly notable. Triterpenoid saponins (e.g., astragaloside I–IV) and flavonoids from
Astragalus have demonstrated cardioprotective effects via anti-oxidative, anti-apoptotic, and homeostasis-regulating pathways [
17–
22]. Salvianolic acid B, a major organic acid in
Salvia miltiorrhiza, has shown strong free radical-scavenging ability and protects against DOX-induced cardiotoxicity by regulating calcium homeostasis and mitochondrial function [
23]. These results suggest that multiple compounds in QZBXD may contribute synergistically to its protective effects. However, we acknowledge that our current study did not perform in-depth pharmacokinetic profiling or dose-response analysis of these active constituents. Future research will aim to isolate and characterize key bioactive components, define their pharmacodynamics and pharmacokinetics, and optimize their dosing strategies.
The gut microbiota is a vital component of the complex human micro-ecosystem [
24]. Our findings also indicate that the gut microbiota plays a crucial role in the cardioprotective effects of QZBXD. Recent studies emphasize the importance of the “gut-heart axis” in cardiovascular diseases, with imbalances in gut microbiota composition contributing to DOX-induced cardiotoxicity [
25–
27]. For example, key polyphenolic compounds from yellow wine can regulate DOX-induced gut dysbiosis by reducing Escherichia-Shigella and increasing beneficial bacteria like Ralstonia and Rikenellaceae [
28]. Similarly, emodin has been shown in animal studies to enhance Bacteroidetes abundance and reduce Verrucomicrobiota, reshaping gut microbiota for cardioprotection [
29]. Components of QZBXD, especially those from
Astragalus membranaceus and
Salvia miltiorrhiza, have been proven to regulate gut microbiota [
30–
34]. Our study found that QZBXD alters the gut microbiota, increasing
Bacteroides while decreasing
Ruminococcus and
Oscillibacter.
Bacteroides is increasingly recognized for its role in cardiovascular health [
35,
36], while
Ruminococcus is linked to metabolic syndrome and cardiometabolic diseases [
37]. Studies have shown that changes in fecal and plasma metabolic profiles in chronic HF patients are closely linked to gut dysbiosis, characterized by a reduction in
Faecalibacterium prausnitzii and a significant increase in
Ruminococcus [
38]. Additionally, patients with atrial fibrillation exhibit an overgrowth of
Ruminococcus in their gut microbiota compared to healthy individuals [
39].
Metabolites form the basis of biological phenotypes, providing a more intuitive and effective way to understand various biological processes and mechanisms [
40]. Our study demonstrates that QZBXD intervention significantly alters the metabolic profiles of cardiotoxic rats. Correlation analysis reveals that, at the genus level,
Ruminococcus and
Oscillibacter are closely associated with key metabolites, suggesting that QZBXD may regulate these microbes to influence tryptophan and arachidonic acid metabolism. Pathway enrichment analysis of differential metabolites suggests that QZBXD primarily mitigates DOX/TRZ-induced cardiotoxicity by modulating tryptophan metabolism and arachidonic acid pathways. Tryptophan is an essential amino acid for humans, serving as a core component for protein synthesis and a critical substrate to produce various key molecules. Two case-control studies in the PREDIMED trial found that plasma metabolites involved in tryptophan degradation via the kynurenine pathway are associated with HF [
41]. The arachidonic acid pathway forms a metabolic network closely linked to the processes of inflammation onset and resolution. Researchers conducting targeted metabolomics of arachidonic acid in plasma samples from acute HF patients found significant metabolic changes. The arachidonic acid score serves as an indicator of myocardial remodeling severity and HF risk [
42].
Nevertheless, it is important to emphasize that the observed microbiota and metabolic changes are correlative, but direct causal relationships have not yet been established. In future studies, further investigation using fecal microbiota transplantation and targeted metabolite interventions will be essential to better understand the potential causal connections between gut microbiota, metabolic pathways, and cardiac function.
In summary, our integrated analysis demonstrated that QZBXD alleviates DOX/TRZ-induced cardiotoxicity, potentially through modulation of gut microbiota and regulation of tryptophan and arachidonic acid metabolism. This study provides experimental evidence supporting the potential of QZBXD as a cardioprotective adjunct to chemotherapy. However, further research, including pharmacokinetic evaluations, dose optimization, mechanistic investigations, and validation of the underlying mechanisms, is necessary to fully elucidate its therapeutic potential and support clinical translation.
PDF
(5905KB)
