Bridging the gap between lipid dyshomeostasis and brain disorders : role of prosaposin and progranulin

Abstract

Brain tissue is the second-highest lipid-containing tissue after adipose in the human body. Lipid homeostasis maintenance requires the coordination of multiple cellular organelles. Pathogenesis of brain disorders, especially Parkinson’s disease (PD), usually involves different dysfunctional pathways that intimately link to lipid homeostasis. Indeed, lipid dyshomeostasis, especially disruption in sphingolipid metabolism, is a critical factor of several common brain disorders, including PD and schizophrenia, and rarer, Gaucher’s disease and Niemann-Picks disease, and probably underlies the dysfunctional pathways contributing to these diseases. Up till now, explanations of various lipid alterations in different disease conditions are lacking, let alone a factor that unifies the principles underlying lipid dyshomeostasis across different diseases. However, recently two lysosomal proteins, prosaposin (PSAP) and progranulin (PGRN), displayed the potential to lend mechanisms to widespread lipid changes across different diseases, which may be a good example to bridge the gap between lipid dyshomeostasis and brain disorders. Both PSAP and PGRN are pleiotropic proteins that show lipid metabolism modulatory functions and neuroprotective effects in the brain and link to different brain disorders. Therefore, the current thesis aims to decipher the role of PSAP and PGRN in three brain conditions, PD, L-DOPA-induced dyskinesia (LID), and schizophrenia.

Regarding PD, in Paper I, we report that PSAP levels correlate mainly with core motor symptoms, while PGRN correlates with non-motor symptoms. In mice, PSAP deficiency in DA neurons (cPSAPDAT) causes behavior defects, DA neurotransmission impairment, and synaptic plasticity disruption. By contrast, mice with PSAP-deficiency in serotonin neurons (cPSAPSERT) show normal behaviors and intact serotonin neurotransmission. Spatial lipidomics unraveled systemic lipid changes in the whole brain, with relation to mitochondrial and peroxisomal functions, in cPSAPDAT mice. On the contrary, cPSAPSERT mice only displayed contained lipid accumulation in the dorsal raphe nucleus (DRN). Metabolic analyses demonstrate that the difference in de novo synthesis of NAD+ between DA and serotonin neurons caused divergent lipid changes in these two mouse lines. cPSAPDAT mice are more vulnerable to α-syn toxicity due to exacerbated aggregation of p-Ser129 α-synuclein that can be reversed by PSAP overexpression, compared to control mice. PSAP overexpression protected wild-type mice against both α-syn- and 6-OHDA-induced toxicity. Consistently, PSAP delivered by encapsulated-cell biodelivery (ECB) devices in the striatum protected rats against α-syn-toxicity. In all, PSAP critically modifies the pathogenesis of PD, especially lipid dyshomeostasis, and serves as a potential therapeutic target for PD. The unfolded protein response (UPR) has long been associated with PD and serves as an important regulator of lipid homeostasis. In Paper II, the unfolded protein response (UPR) related proteins and their transcription levels have been quantitatively confirmed to be changed in PD brains. Meanwhile, UPR is not affected in the periphery of PD patients, which indicates that peripheral UPR is independent of its central counterpart.

Regarding LID and schizophrenia, sphingolipid changes have been found in both diseases, and PSAP and PGRN have been genetically linked to schizophrenia. In Paper III and IV, consistent with these facts, PSAP levels are found to be elevated in LID animals. PSAP differentially modulates lipid metabolism in striatal and non-striatal DRD1 neurons and affects basic physiological functions more of the latter. Meanwhile, PSAP deficiency decreases the susceptibility of striatal DRD1 neurons to L-DOPA-induced malfunction that presents as LID. In contrast with PSAP in LID, PSAP and PGRN are decreased in postmortem cingulate tissue from schizophrenia patients. Moreover, PSAP and PGRN downregulation in the cingulate induces widespread brain immediate early gene (IEG) changes and schizophrenic behaviors in mice, which provide evidence for a causative role of PSAP and PGRN in schizophrenia. Therefore, PSAP, together with PGRN, may take part in the pathogenesis of both LID and schizophrenia, though in opposite ways.

All in all, this thesis sheds light on the role of PSAP and PGRN in multiple brain disorders and proposes that PSAP and PGRN may serve as a bridge between lipid dyshomeostasis and brain disorders.