Aristolochic acids (AAs) mainly exist in plants of the Aristolochia species and Asarum belonging to the Aristolochiaceae family, which have been used for thousands of years to treat various diseases, including eczema, headaches, colds, chronic pain, infections, inflammatory diseases, and obstetrical and gynaecological diseases (Chen, 2011; Kwak et al., 2016; Hua et al., 2023). However, the clinical use of Aristolochia spp. has been limited by the discovery of aristolochic acid nephropathy (AAN) in the 1990s, a rapidly progressive renal interstitial fibrosis induced by AA exposure (Vanherweghem et al., 1993). Emerging evidence has proven that AA exposure is associated with urothelial carcinoma (Karanović et al., 2022; Liao et al., 2023). In addition, AAs are associated with genotoxicity and reproductive toxicity (You et al., 2023; Li et al., 2022; Sun et al., 2023). Recently, there has been significant controversy regarding whether AA exposure is related to liver cancer. Ng et al. have implied that AAs exposure is associated with liver cancer in Asia based on whole-exome sequencing data (Ng et al., 2017). However, Chen et al. found that AAs exposure is not the main cause of liver tumourigenesis in adulthood, drawing conclusions from combined retrospective and cohort studies, AA-DNA adduct detection, AA mutational signature analysis in patients with hepatocellular carcinoma (HCC), and mouse models (Chen et al., 2022). Although individual experiments have shown that intraperitoneal injection of AA I induces animal liver tumours, other studies, including long-term oral toxicity studies, did not observe AA-induced liver cancer occurrence (Lu et al., 2020; Chen et al., 2022; Fang et al., 2022; Wang et al., 2018; Liu et al, 2021). In addition to drug exposure, AAs can act as environmental pollutants, causing chronic renal disease. Balkan endemic nephropathy (BEN), only seen in certain geographical areas in the Balkan peninsula, is caused by the consumption of flour obtained from wheat contaminated with seeds of Aristolochia clematitis (Jelaković et al., 2019). AAs are released into the soil after Aristolochia plants wilt and decay. AAs can then accumulate and become enriched in crops through plant uptake, and their accumulation increases with growth time, indicating that AAs have the potential to cause chronic food poisoning through entering the food chain (Chan et al., 2016; Li et al., 2016; Gruia et al., 2018; Zhang et al., 2022). Currently, the public health risks caused by AAs have attracted worldwide attention. Despite the limitations placed on medicinal Aristolochia plant use in many countries, herbal medicines containing AAs remain accessible. However, the AAs content of these medicines and their potential toxicities remain unclear.

Kelisha capsule (KLS) is a traditional Chinese patented medicine consisting of Atractylodes lancea (Thunb.) DC., Piper longum L., Syzygium aromaticum (L.) Merr. & L.M. Perry., the root and rhizome of Asarum heterotropoides F. Schmidt, Angelica dahurica (Hoffm.) Benth. & Hook. f. ex Franch. & Sav., Acorus calamus var. angustatus Besser, Centipeda minima (L.) A. Braun, and Asch., borneol, realgar, cinnabar, artificial moschus, and dried Bufo gargarizans. It eliminates toxins and halts diarrhoea by regulating qi. Therefore, it is often used to treat acute diarrhoea and sunstroke in clinic (Wang, 1985; Wang, 1987; Xue, 2001). Clinically, KLS is administered orally, and the therapeutic dose range is 0.028–0.037 g/kg, usually for a treatment duration of 3–7 days. Studies have shown that KLS exhibits strong inhibitory effects on Dysentery bacillus and Escherichia coli (Li et al., 2007). In addition, KLS can relieve the fever caused by bacterial infections and reduce the side effects of E. coli endotoxins (Wang et al., 1986). Widespread application of broad-spectrum antibiotics has inhibited or killed many drug-sensitive E. coli strains, resulting in a large increase in drug-resistant strains with the R plasmid, which causes difficulties in clinical treatment. The elimination of resistance plasmids from host strains is an important means of restoring E. coli sensitivity to antibiotics. KLS exerts an obvious elimination effect on plasmid R in drug-resistant strains, suggesting that the combination of KLS and antibiotics is promising for reducing strain resistance in clinical treatment of E. coli infections (Bian et al., 2005). In addition to gastrointestinal disorders caused by bacterial infections, KLS significantly improved diarrhoea caused by non-infectious intestinal inflammation (Jin et al., 2017). KLS can alleviate acetic acid colitis in rats by reducing the levels of serum IL-1β, IL-6, and TNF-α in rats, increasing the level of serum epidermal growth factor, and repairing mucosal epithelium. Regulation of the Toll-like receptor 4/nuclear factor kappa B signalling pathway may be one of the mechanisms by which KLS reduces the inflammatory response of acetic acid colitis (Qin et al., 2018).

One component of KLS, Asarum, contains aristolochic acid analogues (AAAs) (Michl et al., 2017). Aristolochic acid I (AA I), aristolochic acid II (AA II), aristolochic acid IIIa (AA IIIa), aristolochic acid IVa (AA IVa), and aristolactam I (AL I) are the common components extracted from the Aristolochia species. Among AAAs, AAs I and II have shown obvious nephrotoxicity both in vitro and in vivo and are mainly responsible for nephrotoxicity and carcinogenicity (Debelle et al., 2008; Tian et al., 2022). AA I and AA II are reduced to aristolactams during metabolism, and the aristolactams subsequently react with purines in DNA to form AA-DNA adducts of dA-AL-I, dG-AL-I, dA-AL-II, and dG-AL-II (Attaluri et al., 2010; Bárta et al., 2021). The formation of AA-DNA adducts gives rise to a specific A:T to T:A transversion mutation in the TP53 region, which results in nephrotoxicity and carcinogenicity. Therefore, AA-DNA adducts are used as biomarkers to assess AAs exposure and AA-induced upper urothelial cancer (Hollstein et al., 2013; Stiborová et al., 2017). DNA damage induced by AA II is relatively lower than that induced by AA I (Attaluri et al., 2010; Qu et al., 2022). AA I has been reported to be absorbed, distributed, and eliminated more rapidly than AA II. Moreover, the reductive metabolic efficiency of AAI is higher than that of AA II. Differences in in vivo metabolism may be an important reason for the stronger genotoxic and nephrotoxic effects of AA I than those of AA II (Chiang et al., 2023). The co-exposure of AA I and AA II can impact both the metabolism of the aristolochic aids, and increase the formation of AA I and DNA adducts, which leads to increased AA I genotoxicity (Dedı Ková et al., 2020; Bárta et al., 2021). The cytotoxicity and acute toxicity of AA IIIa, AA IVa, and AL I are much weaker than those of AA I, which may be due to the substitution of the toxic 8-methoxy and 10-nitro groups in AA I (Balachandran et al., 2005; Liu et al., 2021). Although AA IVa can form adducts with DNA, it does not cause significant renal interstitial fibrosis or carcinogenicity in vivo (Wan et al., 2021; Xian et al., 2021). AL I is an active component of Aristolochia species and is also an important metabolite of AA I. Although AL I has slightly lower cytotoxicity than AA I, its bioaccumulation, glutathione depletion efficiency, and DNA adduct formation ability are relatively higher than those of AA I (Au et al., 2023). Renal injury induced by AL I has been observed in oral subacute nephrotoxicity assays; however, the histopathological damage induced by AL I is much less than that induced by AAI (Wang et al., 2022). Inhibition of the mitochondrial iron overload-mediated antioxidant system may assist in ALI-induced ferroptosis in renal tubular epithelial cells (Deng et al., 2020).

As AA-DNA adducts persist for an exceptionally long period in renal tissues and lead to irreversible nephrotoxicity, strict control of AAs intake is important to prevent AA-induced nephrotoxicity (Yang et al., 2007; Schmeiser et al., 2014). The AAs content in KLS and their potential toxicity are unknown, which poses certain safety risks for clinical use. Therefore, the aims of this study were to quantitatively detect the content of AA I, AA II, AA IIIa, AA IVa, and AL I in Asarums and KLS and to systematically assess the long-term in vivo toxicity of KLS. Ultra-performance liquid chromatography coupled with tandem mass spectrometry (UPLC-MS/MS) was used to determine AAAs in Asarums and Kelisha capsules. Endpoint measurements, including clinical observations, body weights, blood biochemistry, haematology, and histopathology, of rats administered KLS for one month were analysed to systematically evaluate the potential toxicity of KLS.