Transcriptomics, proteomics, metabolomics and network pharmacology reveal molecular mechanisms of multi‐targets effects of Shenxianshengmai improving human iPSC‐CMs beating

Dear Editor, Bradycardia is one of common cardiovascular diseases; however, effective oral drugs in clinic are rare. Shenxianshengmai (SXSM) is one traditional Chinese medicine licensed by National Medical Products Administration for it. In the study, multi-omics approach will be used to deeply investigate its intracellular mechanism for better clinical application. To verify the purity of hiPSC-CMs, we utilized immunofluorescence and flow cytometry stained with cTnT or α-actinin as well as manual patch clamp to assess their electrophysiological properties (Figure S1A–C). Subsequently, different diluted concentrations of SXSM original solution were selected for the pre-treatment of human iPSC-CMs (Figure S1D–L). Finally, 1 h pre-treatment of SXSM with the concentration of 0.55 mg/mL was preliminarily chosen. As previous studies showed that the effects of SXSM were strongly associated with adrenoceptor-related pathways,1, 2 isoproterenol (ISO) was used to compare with SXSM regarding the effect on human iPSC-CMs (Figure S2A–M). With the treating time protracting, the SXSM group, instead of the ISO group, appeared more durable in the increase of beat rate (BR) and shortage of corrected field potential duration. First, to elucidate the myocardial pharmacological function of SXSM, the potential chemical components of SXSM were identified by UPLC-QTOF-MS/MS and GC–MS in comparison with the standard TCM chemicals (Figure 1A,B), which was followed by network pharmacology analysis. In total, 262 candidate compounds were identified by combining 252 components from UPLC-Q-TOF/MS (Table S1) and 28 components (Table S2) from GC–MS after removing duplicates and isomers. Among them, 63 compounds were confirmed by comparison with the controls, and chemical structures of some representative identified monomer components were shown (Figure 1C, Table S3). Subsequently, based on the STRING database, protein interaction networks of components and potential targets of diseases (arrhythmia, myocardial ischaemia) were constructed (Figure S3A,B, Table S4), and then a 150 core targets' cluster was generated (Table S5). GO-BP/MF analysis (Figure S3C,D) suggested that SXSM might increase the intracellular phosphorylation level and modulate the intracellular energy metabolic environment. In the top 15 metabolic pathways of node degrees (Table S6), phosphorylation-related signal transduction pathways were also identified. Subsequently, transcriptomics–proteomics–metabolomics revealed the molecular mechanism of SXSM improving the beating of human iPSC-CMs (Supplementary Document 1). In transcriptomics, 345 differentially expressed (DE) genes were detected between control (CTRL) and SXSM group (Figure S4A, p-value <0.05). By using GSEA based on GO database (gGSEA), the activity of the three calcium-related pathways was found to increase (Figure 2A, p-value <0.05). Besides calcium, other ions involved in the electrophysiological activity of human iPSC-CMs were also observed in gGSEA (Figure S4B). Moreover, the mobilization of energy metabolism indicated in network pharmacology was also corroborated by gGSEA analysis (Figure 2B, p-value <0.05). Meanwhile, structural units of myocardial contraction were also found to increase (Figure S4C). Consistent with the anti-apoptotic effect shown in the network pharmacology, the pathways of myocardial apoptosis were suppressed (Figure 2C, p-value <0.05). Among the significant down-regulation pathways of GSEA based on the KEGG database (kGSEA), arachidonic acid metabolism had a significant down-regulation trend (Figure 2D, p-value <0.05). Additionally, SXSM could not induce pathological and non-matching angiogenesis based on the Reactome database (Figure S4D,E). Between CTRL and SXSM, a few significantly differential alternative splicing (AS) events were detected (Figure S4F). Subsequently, KEGG pathway enrichment analysis for the genes obtained from each of the five AS forms was performed (Figure 2E), with the result of enriched pathways directly related to spliceosome, heart rate elevation, energy, calcium, phosphorylation and anti-apoptosis. In proteomics, 190 DE proteins were detected between CTRL and SXSM group (p-value <0.05 and fold change > 1.1 or <(1/1.1)) (Figure S5A–F). Subsequently, several pathways worthy of attention in the KEGG analysis of DE proteins were detected (Figure 2F, p-value <0.05). In metabolomics, 1106 metabolites by combining positive (739) and negative (367) ion patterns were identified, including 13 main classes and other additional compounds (Figure S6A). Between CTRL and SXSM group, 138 metabolites (pos 110, neg 68) with significant differences (fold change > 1.2 or <(1/1.2) VIP > 1) were detected (Figure S6B). The DE metabolites were selected for KEGG and SMPDB analysis of Metabolite Set Enrichment Analysis and Metabolomic Pathway Analysis (Figure 2G). Moreover, a joint analysis between network pharmacology and three omics was performed to elucidate molecular pathways (Supplementary Document 2). A two-by-two analysis in the 'gene–protein–metabolite' dimension was conducted: interaction network for AS events and proteins (Figure 3A), protein–metabolite correlation network analysis (Figure 3B) and Orthogonal Projections to Latent Structures (O2PLS)3-6 for DE genes (p-value <0.05) and DE metabolites (Figure 3C, Figure S6C,D). Finally, multichannel validation of SXSM enhancing human iPSC-CMs beating was performed. KN93 significantly reversed the effect of 0.55 mg/mL SXSM on BR elevation of human iPSC-CMs at the concentration of 300 nM (106.0% ± 2.5%) and further inhibited Amp (80.5% ± 2.8%) (Figure 4A,B). Moreover, AIP pre-treatment had no significant effect on either the BR or Amp changes of human iPSC-CMs induced by SXSM (Figure 4C,D). Meanwhile, pCaMK of human iPSC-CMs was decreased after 0.55 mg/mL SXSM 1 h treatment (Figure 4E). NCLX inhibitor CGP37157 ≥10 μM post-treatment could suppress the elevation of BR of human iPSC-CMs due to 0.55 mg/mL SXSM treatment for 1 h (Figure 4F,G). Furthermore, the inhibition of CaMKII by SXSM contributed to myocardial protection and the prevention of oxidative stress due to Ca2+ overload.7 Collectively, this suggests that 0.55 mg/mL SXSM treatment mainly triggers Ca2+-CaM-AC-PKA, a pathway that activates the calcium pool (Figure 4H). Additionally, SXSM may influence the whole process of human iPSC-CMs' action potential at least through Nav1.5, Kir2.1 and Kv11.1 (Figure S7). In sum, increased intracellular calcium cycling and intracellular calcium, and accelerated excitatory contractile coupling in human iPSC-CMs by SXSM, which could coordinately lead to their faster beating. Meanwhile, SXSM promoted TCA cycling and oxidative phosphorylation through adequate mobilization of the glycolytic pathway, amino acid metabolism and fatty acid β-oxidation to better supply energy to human iPSC-CMs. Combined with the enhanced antioxidant pathway and reduced apoptosis in SXSM-treated human iPSC-CMs, SXSM implicated a cardioprotective effect (Figure 4I). In addition, the long-term mechanism of SXSM and its protective effect on hypoxia and reoxygenation are also worthy of further exploration (Figure S8). As science and technology advance, the more precise mechanism of SXSM improving heartbeat will be demonstrated. This may offer alternative therapy options for arrhythmia patients who cannot have a pacemaker implanted. We thank Help Stem Cell Innovation Co., Ltd (Nanjing, China), Shanghai Standard Technology Co., Ltd., Shanghai Applied Protein Technology and Buchang Pharmaceutical (Shandong, China) for their technological support of the study. Data Analysis and Visualization are supported by the Medical Research Data Center of Fudan University, and some images were performed by using the OmicShare tools, a free online platform for data analysis (https://www.omicshare.com/tools). This work was supported by the National Natural Science Foundation of China (82170334, 81870182). There is no conflict of interest involved in this article. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.