Traditional Chinese medicine (TCM), effective alternative therapy for various diseases, has been widely used in China and other Asian countries for thousands of years (X. Liu et al., 2019). However, the complex components and unclear potential toxicity of TCM seriously hindered their clinical application (R. Liu et al., 2020; Zhong, Menon, Deng, Chen, & He, 2015). In recent years, several studies have demonstrated that some TCM could lead to chronic kidney disease (CKD) and renal fibrosis, which has been considered the final pathway for the development of CKD and was characterized by tubular epithelial–mesenchymal transition (EMT), interstitial myofibroblast activation, excessive extracellular matrix deposition, and fibrotic tissue remodeling.

Fructus Psoraleae (Buguzhi), the dry and mature fruit of plant Cullen corylifolium (L.) Medik. (Fabaceae) (syn. of Psoralea corylifolia L.), is a well-known TCM for external as well as internal treatment of various diseases, for instance, alopecia, inflammation, leukoderma, leprosy, psoriasis, and eczema from ancient times (Alam, Khan, & Asad, 2018). Regardless of its rich pharmacological activities, the adverse effects of Fructus Psoraleae have also been reported in recent years. For instance, Fructus Psoraleae decoction caused hepatotoxicity and nephrotoxicity in rats after 7 days of administration (A. Li et al., 2016). In vitro studies showed that Fructus Psoraleae could induce toxicity to human renal tubular epithelial cells by damaging cell membrane, inducing cell apoptosis, inhibiting intracellular DNA synthesis, blocking cell mitosis, and inhibiting cell proliferation (F. Jiang, Zhou, Wang, & Zhang, 2010). But the exact toxic component and molecular mechanism underlying Fructus Psoraleae-induced nephrotoxicity were not fully understood. Bakuchiol, as the main constituent of Fructus Psoraleae, was found to have a toxic effect on mouse liver and kidney at high doses (Hu et al., 2018; Z. J. Li et al., 2017; Y. Zhang, Liu, Wu, & Wang, 1981). However, our previous study demonstrated that bavachin (BV), another newly discovered toxic substance in Fructus Psoraleae, could cause obvious hepatotoxicity at very low doses, implying BV might be one of the main toxic components of Fructus Psoraleae-induced nephrotoxicity (Y. Yang et al., 2018).

EMT refers to the switching of cells from an epithelial state to a mesenchymal fibroblast state, accompanied by decreased intercellular adhesion and increased motility. There are three different subtypes of EMT, including type I EMT, type II EMT, and type III EMT. Type II EMT associated with fibrosis has commonly been considered the main fibrogenic process in renal fibrosis. More than 30% of the novel fibroblasts originate from local EMT (X. Liu et al., 2019). Furthermore, recent studies have identified EMT as a key determinant of the progression from adaptive to maladaptive repair in renal fibrosis (Lovisa et al., 2015; Qi et al., 2021). In general, the initiation of EMT aims to prevent epithelial stress response, but overactivated EMT significantly affects renal tubular epithelial cells uptake and secretion functions, induces G2 cell cycle arrest, and impairs tissue repair (Lovisa et al., 2015). Furthermore, EMT could cause interstitial fibrosis, inflammation, and immune recruitment by altering the epithelial secretome profile (Vincent & Fuxe, 2017; Wang et al., 2020). But so far, no direct evidence has focused on the relationship between BV and EMT.

Mechanistically, excessive oxidative stress and/or inflammatory reaction-induced EMT have been supposed to play critical roles in TCM-induced renal fibrosis (Cybulsky, 2017; Ji et al., 2019; Zhou et al., 2020a). Usually, the endoplasmic reticulum senses cellular stress and subsequently restores cellular homeostasis by triggering the ER stress and downstream unfolded protein response (UPR) (Cybulsky, 2017). In our previous study, we confirmed that BV-triggered ER stress participated in Fructus Psoraleae-induced hepatotoxicity (Yang et al., 2021). The latest articles showed that enhanced ER stress promoted EMT in human lens and lung epithelial cells (D. Liu et al., 2018; Zhou et al., 2020b). Also, there was crosstalk between the UPR and EMT under some conditions (Zhao, Zhang, Sun, & Brasier, 2021). Hence, we speculated that BV-triggered ER stress might contribute to Fructus Psoraleae-induced nephrotoxicity via EMT-mediated renal fibrosis.

In this present study, we first confirmed whether repeated BV treatment could induce renal fibrosis, ER stress, and EMT in vivo and in vitro and elucidated the possible molecular mechanism involved in BV-induced renal fibrosis, aiming to provide a specific intervention target for Fructus Psoraleae-caused nephrotoxicity.

Since thousands of years ago, Fructus Psoraleae has been widely used clinically for various disease therapy, for instance, spermatorrhea, vitiligo, and lumbago. Recently, increasing studies have reported Fructus Psoraleae could cause obvious toxic effects on the liver and kidney (Nam et al., 2005; Y. Zhang et al., 1981). But compared with hepatotoxicity, little research focused on Fructus Psoraleae-induced nephrotoxicity and its underlying mechanism. In addition, our previous studies found that being one of the most important components, BV could induce evident adverse outcomes at a very low dose and was presumed to be relevant to Fructus Psoraleae-induced nephrotoxicity (Jia-yan et al., 2021; Y. Yang et al., 2018). In this study, we first confirmed that long-term treatment with BV could induce apparent EMT and renal fibrosis in zebrafish kidneys. Mechanistically, BV treatment led to ROS generation and the activation of the Bip/eIF2α/CHOP signaling pathway and subsequently caused ER stress, UPR, and cytoplasmic Ca²⁺ overloading, both of which in turn regulated the expressions of classical signals involved in EMT and fibrosis, ultimately resulting in renal fibrosis in renal tubular epithelial cells. Furthermore, for the first time, ginsenoside Rb1 was verified to function as a protective agent against BV-induced EMT and renal fibrosis in vivo and in vitro through inhibiting ROS generation and Bip/eIF2α/CHOP toxicity pathway and following Ca²⁺ overload.

Renal fibrosis is considered the common final pathway of almost all nephropathies and is characterized by oxidative/inflammatory responses, fibroblast activation, excessive deposition of extracellular matrix, and fibrotic tissue remodeling (Nastase, Zeng-Brouwers, Wygrecka, & Schaefer, 2018). Of a wide range of insults, several Chinese herbs had been reported in relevant to drug-induced nephropathy, including Aristolochiae manshuriensis Kom. and Fructus Psoraleae (B. Yang et al., 2018). Different from the former, the exact toxic component and toxicity model of action involved in Fructus Psoraleae-induced renal injuries were not entirely elucidated. Consistent with previous studies and our earliest hypothesis, this study first identified that repeated treatment with low-dose of BV, one of the main toxic ingredients in Fructus Psoraleae, caused significant renal interstitial and tubular epithelial deposition of collagen fibers in zebrafish kidneys. The latest evidence showed that adverse effects on tubular epithelial cells acted as a driving force that initiated and promoted interstitial injuries and consequent fibrosis upon exposure and response to various insults (H. Li et al., 2019). Generally, AQP1, one of the water channel aquaporin isoforms highly expressed in the renal tubular epithelium, participates in the maintenance of body water and sodium homeostasis (J. Li et al., 2018; Su J. et al., 2020). N-cadherin, the adhesion marker of mesenchymal cells, expresses in fibrous, nervous, and musculoskeletal tissues, but does not largely exist in the epithelium (Noronha et al., 2021). Interestingly, our results presented an obvious reduction of AQP1 expression in the renal tubular epithelial cells of BV-treated zebrafish, indicating that long-term BV treatment induced an ongoing loss of normal structure-specific properties of tubular epithelium. Also, enhanced N-cadherin content was closely related to the acquisition of mesenchymal properties in the renal tubular epithelium, namely, EMT. Therefore, low-dose of BV-induced renal fibrosis could be EMT-dependent.

Moreover, in vitro results showed that cellular expressions of cadherins were dramatically disturbed in BV-treated HK2 cells. Apart from the enhancement of N-cadherin, the expression of E-cadherin, which regulated epithelial tissue integrity, was substantially inhibited by BV. This phenomenon was usually called E- to N-cadherin switch and is considered as the hallmark of EMT. Altered cadherin expression is accompanied by the suppression of epithelial tight junctions or zonula occludens-related proteins (ZO1 and Claudin1) and the increases in E-cadherin–related transcriptional repressors (ZEB1 and slug), together with confirmed BV-triggered EMT in human renal tubular epithelium. Mechanistically, many upstream signaling pathways, including TGF-β, Wnt, and Notch1, were well-documented to regulate the developmental EMT and renal fibrosis (Edeling, Ragi, Huang, Pavenstadt, & Susztak, 2016; J.; Su W. et al., 2020). Except for the Wnt/β-catenin signal, BV indeed dose-dependently upregulated canonical TGFβ1/Smad 3 and Notch1/NICD signaling pathways. Meanwhile, significant fibroblast activation in BV-treated tubular epithelium was evidenced by increasing the expressions of α-SMA and vimentin, both of which formed bundles of stress fibers and filament. However, BV caused distinct depletions in several tubular epithelial cells’ specific transporters, including AQP1, SLC22A6, and ATP1A1 (known as the Na⁺/K⁺-ATPase pump). Hence, we further demonstrated that BV treatment could promote EMT and fibroblast-like properties’ acquisition in renal tubular epithelial cells through TGFβ1 and Notch1-mediated fibrotic signaling responses, all of which eventually caused renal fibrosis.

Oxidative responses always conduce to toxic effects induced by xenobiotics, including BV. Our previous work reported that BV caused ROS generation and finally promoted obvious cytotoxicity in vivo and in vitro (Y. Yang et al., 2018). Notably, ROS could increase the expression of slug, which was the transcriptional repressor of E-cadherin, by increasing the phosphorylation of extracellular signal-regulated kinase (ERK) (Z. Zhang et al., 2020), then downregulated cellular E-cadherin content, and changed the epithelial architecture via relegating cell-cell adhesion (Noronha et al., 2021). Also, overwhelming ROS-induced oxidative damages could activate HIF-1α and NF-κB, both of which drove the occurrence and progression of EMT following fibrosis (Gloire & Piette, 2009). In addition, ER, an elegant organelle that controls protein synthesis, folding, and trafficking, was reported to be vulnerable to excessive ROS (Zeeshan, Lee, Kim, & Chae, 2016). Oxidative stress or damage caused disruption of ER function and resulted in the accumulation of newly synthesized misfolded protein in the ER, known as ER stress. Subsequently, enhanced ER chaperone (Bip) binding to misfolded proteins could lead to the release of ER stress sensors, including IRE1α, PERK, and ATF6, and activate the following UPR to restore homeostasis via ER-associated degradation (ERAD) system (Cybulsky, 2017). Overt ER stress and following UPR were confirmed to play a critical role in BV-induced hepatotoxicity. But unlike unselective activation of the three classical ER stress sensors in BV-treated hepatocytes, only the Bip/PERK/eIF2a/ATF4 signaling pathway mediated BV-induced ER stress and UPR in HK2 cells. Furthermore, as the executors of ER stress-initiated mitochondrial apoptosis, CHOP and Ca²⁺ signaling could cause a profibrotic microenvironment by promoting apoptosis (Song, Gou, & Zhang, 2016; Zhong et al., 2021). Also, elevated Bip expression had been found to participate in EMT-associated metastasis (Kopp, Larburu, Durairaj, Adams, & Ali, 2019). Not surprisingly, when we reduced the levels of ROS and cytoplasmic Ca²⁺, BV-induced EMT and profibrotic signaling were alleviated, respectively. Altogether, ROS-induced ER stress could be the key modulator of BV-caused EMT and renal fibrosis.

Ginseng, one of the favorable natural medicines which have a long history of clinical use in east Asia for thousands of years, has been found to possess antioxidative, anti-inflammation, and anti-fibrosis properities (Mancuso & Santangelo, 2017). As one of the main effective ingredients in ginseng, ginsenoside Rb1 has shown excellent therapeutic effects on nervous and cardiovascular system diseases (Lin et al., 2022). In terms of pharmacological mechanisms, ginsenoside Rb1 limited ROS generation by suppressing the activity of NADH dehydrogenase (L. Jiang et al., 2021). On the other hand, ginsenoside Rb1 had been verified to target multiple pathways, including Nrf2/ARE, STAT3, and TGFβ1/Smad3 signaling pathways, to regulate redox balance, inhibit apoptosis, and attenuate cardiac fibrosis (Zheng et al., 2017; Wu, Huang, Bell, & Yu, 2018; Peng et al., 2021). Like the aforementioned efficacy, ginsenoside Rb1 not only inhibited BV-induced ROS generation but also decreased overactivated ER stress following UPR and cytoplasmic Ca²⁺ release. In particular, BV-caused sustained activation of EMT progress and renal fibrosis were significantly ameliorated by ginsenoside Rb1 after long-term combination administration. Hence, ginsenoside Rb1 was proposed to be a valuable ingredient in the therapy of BV-caused renal fibrosis.

In summary, this study first confirmed that low dose of BV could cause obvious EMT-related renal fibrosis in vivo and in vitro, which was associated with the activation of the TGFβ1/Smad3 and Notch1/NICD signaling pathway. During this progress, elevated ER stress and cytoplasmic Ca²⁺ overload, both of which were evoked by the ROS-triggered Bip/eIF2a/CHOP signaling pathway, play critical roles in BV-induced renal fibrosis (Figure 6). Furthermore, we discovered for the first time that ginsenoside Rb1 exerted an outstanding effect against BV-induced renal fibrosis via suppressing ROS-triggered ER stress following EMT. Hence, this research helped to understand the nephrotoxicity caused by low-dose of BV and its toxic model of action and provided evidence that ER stress could be an alternative target and ginsenoside Rb1 could be a promising natural ingredient for the treatment of BV- or Fructus Psoraleae-induced renal fibrosis.