Lysosomal membrane stability (LMS) has been used in various organisms as a very sensitive biomarker of stress. However, despite the abundance of data about regulation of the autophagic process in mammals, in the invertebrates there is only limited mechanistic understanding. Marine mussels (Mytilus galloprovincialis Lam.) are bivalve molluscs, widely used as models in ecotoxicology and as environmental bioindicators of sea water quality. In order to elucidate this fundamental process, in the present study, mussels were exposed for 3 days to a “priority”, ubiquitous environmental contaminant, benzo[a]pyrene (B[a]P) at different concentrations (i.e. 5, 50, 100 μg/L seawater). B[a]P accumulated in lysosomes of digestive tubule epithelial cells (digestive cells) and in enlarged lipid-rich lysosomes (autolysosomes) as detected by immunofluorescence and UV-fluorescence. B[a]P also activated the autophagic process with a marked decrease of LMS and concurrent increase in lysosomal/cytoplasmic volume ratio. Dephosphorylation of mTOR contributes to increased lysosomal membrane permeability and induced autophagy. B[a]P induced a decrease in phosphorylated (active form) mTOR. The probable role of mTOR in cell signalling and the regulation of the cellular responses to the contaminants has been also confirmed in a field study, where there was significant inactivation of mTOR in stressed animals. Statistical and network modelling supported the empirical investigations of autophagy and mTOR; and was used to integrate the mechanistic biomarker data with chemical analysis and DNA damage.

溶酶体膜稳定性（LMS）已在各种生物中用作压力的非常敏感的生物标记。然而，尽管有大量关于哺乳动物自噬过程调控的数据，但在无脊椎动物中，对机理的了解还很有限。海洋贻贝（Mytilus galloprovincialis Lam。）是双壳软体动物，被广泛用作生态毒理学的模型和海水质量的环境生物指标。为了阐明这一基本过程，在本研究中，贻贝暴露于3天的“优先”，普遍存在的环境污染物，不同浓度的苯并[ a ]（（B [ a ] P）（即5、50， 100μg/ L海水）。B [ a通过免疫荧光法和紫外荧光法检测到，] P在消化管上皮细胞（消化细胞）的溶酶体和扩大的富含脂质的溶酶体（自体溶酶体）中积累。B [ a ] P还激活了自噬过程，LMS明显降低，同时溶酶体/细胞质体积比增加。mTOR的去磷酸化有助于增加溶酶体膜通透性和诱导自噬。B [ a] P诱导磷酸化（活性形式）mTOR减少。在现场研究中也证实了mTOR在细胞信号转导中的可能作用以及细胞对污染物的反应的调节，其中在应激动物中mTOR明显失活。统计和网络建模支持自噬和mTOR的实证研究。并用于将机械生物标记数据与化学分析和DNA损伤整合在一起。