Alterations In Lipid Metabolism In The Nervous System By Ethanol

National Institute On Alcohol Abuse And Alcoholism

Investigators

Linked publications, trials & patents

Abstract

We have previously established that the metabolism of docosahexaenoic acid (DHA) to N-docosahexaenoylethanolamine (synaptamide) is a significant endogenous mechanism for promoting neurogenesis, neuritogenesis, synaptogenesis and anti-inflammatory actions. We further demonstrated that the orphan G-protein coupled receptor 110 (GPR110, ADGRF1), an adhesion GPCR (aGPCR), is the synaptamide target receptor, triggering cAMP production with low nM potency. We established that synaptamide is the only small-molecule endogenous ligand thus far identified for an aGPCR a newly emerging class of GPCR. We also established the in vivo significance of this mechanism in injury/inflammation models, suggesting a therapeutic potential for targeting GPR110. During this review period, we continued to characterize the structure and function of GPR110 with particular attention to the modulation by ethanol. We found that GPR110 signaling along with its developmental function is recapitulated in neural cells derived from human induced pluripotent stem cells (iPSCs), presenting an opportunity to study the role of GPR110 activation in human neurodevelopment and neuro-dysfunction. GPR110 is expressed in iPSCs and neural progenitor cells (NPCs) derived from iPSCs, and GPR110 ligands dose-dependently increased cAMP in both cells. To investigate the neurodevelopmental implications of GPR110 signaling, we generated a GPR110 mutant iPSC line with a mutation at F663S presumably associated with schizophrenia based on the Swedish schizophrenia exome-sequencing study. This mutant was functionally inactive as GPR110 ligands did not increase cAMP due to its impaired expression in the plasma membrane. During differentiation, the F663S mutant showed a lower number of MAP2-positive neurons compared to the WT while maintaining a higher number of nestin-positive NPC cells, indicating suppression of neurogenesis. The neurons derived from the F663S mutant NPCs also showed retarded neurite outgrowth evidenced by significantly reduced total neurite length and number of branches per neuron. The multielectrode array (MEA) technique that monitors spontaneous electrophysiological activity of neurons indicated significantly decreased neuronal activity in the F663S mutant neurons, with lower spiking rates and network bursts (synchronized spiking events). GPR110 ligands, synaptamide or A8, promoted neurite outgrowth and accelerated spontaneous electrophysiological activity and synchrony in WT but not in the F663S mutant neurons. Moreover, the F663S mutation caused strikingly abnormal neuronal morphology with thicker and shorter neurites. While the WT GPR110 localized in the periphery of the soma and neurites in WT, the mutant GPR110 was contained in the nucleus and soma. These data indicate that GPR110 plays a crucial role not only in stimulating neurogenesis, neurite outgrowth and synaptic function but also in proper development of neuronal morphology during differentiation of hNPCs. Abnormal development of neurons observed with the functionally inactive F663S GPR110 mutant suggests a potential link between GPR110 dysfunction and schizophrenia. We found that human NPC cells can synthesize synaptamide from its precursor fatty acid, DHA. When treated with 13C-labeled DHA, NPCs generated significant amounts of 13C-labeled synaptamide. The media contained more than 95% newly synthesized synaptamide, indicating that this metabolite is mostly released after synthesis. Endogenous synaptamide was also found in the media, but at a much lower level. The presence of a fatty acid amide hydrolase inhibitor (FAAHI) did not affect the level of endogenous or 13C-labeled synaptamide significantly, indicating that the hydrolysis of synaptamide by FAAH is not active in the early stage of neuronal development. Previously, we have established that DHA-derived synaptamide is a potent suppressor of neuroinflammation in an LPS-induced model, by enhancing cAMP/PKA signaling and inhibiting NF-kB activation through GPR110 activation in both brain and periphery. We have also found exacerbating effects of binge ethanol exposure (3 g/kg, gavage) on LPS (1 mg/kg, i.p.)-induced neuroinflammation which can be ameliorated by synaptamide (5 mg/kg, i.p.) in a GPR110-dependent manner based on the expression profile of proinflammatory indicator genes including TNF, IL-6, IL-1 and NLRP3. Iba-1 immunohistochemistry as well as western blot analysis of IL-1, NLPR3 and Iba-1 also confirmed the exacerbating effects of ethanol and GP110-dependent suppressing effects of synaptamide on LPS-induced inflammatory signals at the protein level. The LPS-induced NLRP3 expression was drastically elevated in the ethanol-fed GPR110 KO animals compared to the ethanol-fed WT mice, suggesting that inflammatory responses enhanced by ethanol are further sensitized in the absence of GPR110. A binge exposure to ethanol and LPS injection significantly decreased the mRNA expression of specifically the AC8 isoform in the brain. Conversely, both ethanol exposure and LPS injection specifically increased the expression of the PDE4b phosphodiesterase isoform at both mRNA and protein levels. Synaptamide reversed the reduction of AC8 and normalized PDE4b, but these effects were not observed in the absence of GPR110. These data indicated the protective potential of ligand-induced GPR110 activation in neuroinflammation by counteracting the effect of ethanol and LPS on the cAMP system, particularly AC8 and PDE4 expression. To enhance our understanding of the newly deorphanized adhesion receptor GPR110, we continue to investigate the structural details of GPR110 in relation to its activity. GPR110 contains a SEA (sperm protein, enterokinase, and agrin) domain (AA 148-256) immediately before the conserved GPCR-autoproteolysis-inducing (GAIN) domain (AA 256-578, cleavage site HL566/T) in its extracellular region. Although it has been documented that the SEA domain of adhesive proteins contains a consensus proteolytic cleavage site (G/SVVV), the auto-proteolysis of SEA domain in GPR110 and its biological implications have not been reported. Based on the consensus sequence and the surrounding secondary structure, we first tested if the mutation in the G206/S207IVA motif prevents GPR110 cleavage. HEK cells were transfected with S207A or WT GPR110-HA and the GPR110 expression profile was evaluated by quantitative mass spectrometry and western blotting using an N-terminal targeting GPR110 antibody or a C-terminal targeting HA antibody. Each antibody detected a GPR110 fragment at 30 kDa, in addition to the intact GPR110, indicating the presence of two different cleavage products containing the N-terminal or C-terminal region. The C-terminal fragment corresponding to the GAIN domain cleavage product (AA, 567-910) was not observed in cells expressing the H565A/T567A mutant that is resistant to GAIN domain auto-proteolysis while the N-terminal fragment was diminished in the S207A-transfected cells. These data indicate that G206/S207 is the SEA domain cleavage site in GPR110. Interestingly, the S207A mutation markedly decreased the auto-proteolytic activity of the GAIN domain, similar to what is observed with the H565A/T567A mutation. These data suggest that conformational stress caused by the SEA domain is an important factor influencing the self-catalytic activity of the GAIN domain along with the known nucleophilic attack at the cleavage site of HL566/T. The ligand-induced cAMP production was independent of the SEA domain cleavage but was abolished when the SEA domain (AA, 148-252) was deleted. In summary, we identified a novel auto-proteolysis site of GPR110 in the SEA domain at G206/S207. Without SEA domain cleavage, the self-catalytic cleavage of the GAIN domain is impeded, and the presence of the SEA domain is necessary for ligand-induced activation of GPR110.