| Pyrrolizidine alkaloids (PAs) are hepatotoxic and present in a wide variety of plant species. The consumption of PAs through the intake of either PA-containing natural products or PA-contaminated food stuffs leads to serious health problems in human. However, the severity of hepatotoxicity of different PAs varies significantly, and thus makes it difficult to set up the regulation for a universal threshold of toxic dose of individual PAs. In addition, the detailed mechanisms underlying PA-induced hepatotoxicity and the resultant hepatic sinusoid occlusion syndrome (HSOS) remain largely unknown, which results in the lack of specific methods for its diagnosis. Therefore, the aims of the present study included two aspects: the first is to develop a convenient in vitro model for the rapid assessment of the hepatotoxicity of different PAs; the second is to identify potential biomarkers for the diagnosis of P A intoxication in rats.;The results demonstrated that the developed in vitro model using MTT and BrdU assays under optimal incubation conditions was able to discriminate the cytotoxicity of three types of PAs. Among different PAs tested, clivorine, an otonecine-type PA representative, was significantly more toxic than retrorsine, a retronecine-type one, while platyphylline, a platyphylline-type one, was not toxic. Moreover, the cytotoxicity of alkaloid extract of Gynura segetum, a senecionine and seneciphylline-containing herb, was comparable to that of these two PAs tested individually, demonstrating that the developed model was also suitable for the assessment of PA-containing herb-induced cytotoxicity.;In the in vivo study, administration of retrorsine to rats produced hepatotoxicity in a dose-dependent manner: low dose induced a slight liver injury with an elevation of ALT, GSSG/GSH, MPO, and slight loss of SECs; medium dose caused a mild injury with enhanced elevation of ALT, MPO, RLW, GSH and GR, and pronounced damages in SECs and CVECs; whereas high dose resulted in severe damage evidenced by all aforementioned markers plus 1) significant increase in TB and LDH, 2) significant increase in GSH but no change in GSSG/GSH indicating the initiation of GSH depletion, 3) severe damages in SECs and CVECs, and 4) extensive sinusoid hemorrhage and lobular disarray. Using this developed rat model, a proteomic study was conducted to identify the modulations of hepatic proteins with a general trend of down-regulation at low dose and up-regulation at high dose, while remarkable variations were found at medium dose of retrorsine treatment. These proteins were found to belong to six clusters according to their functions, including proteins responsible for metabolism, related to cell organization, cell proliferation, and stress response, involved in thrombosis, as well as proteins related to several other functions. Among these modulated proteins, regulations of Cps 1 and Atp5b were consistently observed at three dosage regimens, and were verified by immunostaining and Western blot analysis.;In conclusion, for the first time, a convenient in vitro model for the assessment of the severity of cytotoxicity of different type of PAs has been developed. This method has a potential to be used for the quick screening the toxicity of PA-containing natural products. Furthermore, Cps 1 and Atp5b were found to have a potential to be developed as specific biomarkers for the diagnosis of PA intoxication in rats. |
