3.5 Diagnostic utility of ICP placenta differential lipids
We constructed ROC curves and calculated AUCs to evaluate the probability of these lipids being diagnosed with ICP. As shown in Fig 5A-C, the AUCs of SM (d42:1), SM (d18:1/24:1) and PC (17:0/18:2) were 0.718, 0.768 and 0.748, respectively. And the AUCs of PE (18:0p/20:1), dMePE (20:1p/16:0) and PE (16:0p/22:1) were 0.735, 0.723 and 0.764 (Fig 5D-F). We also evaluate the diagnostic utility of 5 lipids identified from autophagy pathway analysis, results showed the AUCs of PE (16:0/18:1), PE (16:0/20:2), PE (18:0/22:4), PE (18:1/20:3) and PE (18:1/20:4) were 0.682, 0.649, 0.665, 0.659 and 0.657, respectively (Fig S4). Using multiple logistic regression analysis, we found that a good diagnostic model for ICP could be constructed by combining these lipids. Analysis results of combined biomarker data yielded an AUC of 0.904 for PE (16:0/20:2), SM (d42:1) and PC (17:0/18:2) (Fig 5G). The AUCs even reach 0.933 when combining PE (16:0/20:2), SM (d42:1), PC (17:0/18:2) and PE (16:0p/22:1) (Fig 5H). Therefore, the combination of these lipids could be reliable, and a novel biomarker in predicting the risk of ICP.
Discussion
Intrahepatic cholestasis of pregnancy is associated with increased risk of perinatal mortality and morbidity, however, the pathogenesis of ICP is still unclear. There is no effective treatment for ICP currently, once ICP is diagnosed, the only effective way to prevent adverse pregnancy outcomes is delivery27. As an emerging field that branches from metabolomics, more and more lipidomics proved that lipids have multiple functions and are involved in many biological processes. Previous studies suggested that placental lipid profiles, for example, LDL cholesterol and sphingolipids differed between normal pregnant women and ICP patients23, 28. Given that lipid metabolic dysregulation potentially plays a role in pregnancy, a better understanding of lipid metabolism could have significant clinical implications for the diagnosis and treatment of ICP.
In this study, we found ICP placenta has a significantly different lipid profile from normal pregnancy. A total of 61 lipids were found to be dysregulated in mild ICP placenta, and the differentially expressed lipids mainly include PE, PC, PS and SM species. We also identified 88 differential lipids between severe ICP placenta and normal pregnancy, and these lipids mainly include PE, PC, PS, SM and TG species. Among these lipids, we identified 44 lipids differentially expressed both in mild and severe ICP group compared with the control group, mainly of them were SM, PE, PC and PS species. Moreover, pathway analysis of the placentas of patients with ICP identified autophagy regulation, glycerophospholipid metabolism and GPI-anchor biosynthesis, suggesting these pathways might be related to the pathogenesis of ICP.
Autophagy is an evolutionally conserved process that targets cytoplasmic components such as organelles, protein aggregates or proteins for degradation into lysosomes to maintain cellular homeostasis under environmental stress29, 30. Autophagy plays a pivotal role in implantation, embryogenesis and maintenance of pregnancy31. A previous study demonstrated the expression of LC3-II is elevated in ICP human placenta and placentas of ICP rats, indicating autophagy is closely related to ICP25. Zhang et al used TCA-treated term placental villous for proteomic analysis, their results revealed protein expression difference in ICP mainly related to autophagy and cell metabolism32. Consistent with their result, Fang et al analyzed 10 ICP and 10 normal placental tissue through quantitative proteomics and their results showed the differentially expressed proteins mainly participated in autophagy, autophagosome formation and metabolism33. All of these indicate a close correlation between ICP and autophagy. In our lipidomic analysis, we also found the differentially expressed lipids mainly enriched in the autophagy pathway, and PE (16:0/18:1), PE (16:0/20:2), PE (18:1/20:3), PE (18:1/20:4) and PE (18:0/22:4) participated in this process. Phosphatidylethanolamine (PE) is the second most abundant phospholipid in the membranes of all mammalian cells34, 35. As a fundamental component of biological membranes, PE is essential for many cellular functions and the dysregulation of PE metabolism has been implicated in many diseases, such as nonalcoholic liver disease, atherosclerosis and obesity36, 37. In addition, PE is essential for the formation of LC3-II in the autophagolysosomal bilayer membrane. During autophagy, a cytosolic form of LC3-I is conjugated to phosphatidylethanolamine to form LC3- phosphatidylethanolamine conjugate, that is LC3-II, which is recruited to autophagosomal membranes. The autophagosomal marker LC3-II could reflect autophagic activity, and detecting LC3-II has become a reliable method for monitoring autophagy and autophagy-related processes38. Therefore, excessive accumulation of PE (16:0/18:1), PE (16:0/20:2), PE (18:1/20:3) and PE (18:1/20:4) in ICP placenta may lead to the formation of LC3-II and eventually the activation of autophagy pathway, which has been confirmed to be involved in the pathogenesis and development of ICP.
Through combined analysis with previous studies, we also identified SM (d42:1), SM (d18:1/24:1) and PC (17:0/18:2) are differentially expressed both in ICP placenta and plasma samples. And their good diagnostic utility was also accessed through the construction of ROC curves. Two of these three lipids were sphingolipids. The sphingolipid metabolic pathway produces bioactive metabolites like sphingosine 1-phosphate (S1P), ceramide and sphingosine, they can participate in various biological processes such as cell survival, growth, vascular integrity and inflammation. Sphingolipids have been proven to be related to pregnancy-associated complications such as preeclampsia, recurrent pregnancy loss and intrauterine growth restriction39-41. A previous study reported the median values of C16-Cer and C18-Cer were significantly increased in the plasma of ICP patients, and after treatment with UDCA for a week, the concentration of C16-Cer and C18-Cer decreased considerably42. Consistent with their finding, Sun et al also reported ICP is associated with disordered sphingolipid homeostasis23. Collectively, this evidence proved that sphingolipids may act as new diagnostic targets for ICP.
There are still several limitations in our study. First, further animal and cellular experiments are needed to elucidate the specific mechanisms of lipid changes. Second, many studies on lipidomics and ICP have been conducted with a small number of patients, which limits the generalizability of the findings. Larger studies are needed to confirm the role of lipids in ICP. Finally, the diagnosis of ICP is currently based on clinical symptoms and biochemical tests. However, there is a lack of standardized criteria for the diagnosis of ICP, which can lead to variability in patient selection and make it challenging to compare results across studies.
In conclusion, we used UHPLC to characterize the placental lipidomics profiling for women complicated with ICP. Our results confirmed there were significant changes in lipid profile in the mid and severe ICP placenta, the metabolic pathway was primarily associated with autophagy and glycerophospholipid metabolism. And ICP placenta is associated with dysregulation of sphingolipid homeostasis. Our research highlighted some important mechanisms involved in the onset of ICP and might provide new insight into the treatment and diagnosis of ICP.