How do increased levels of ammonia as

How do increased levels of ammonia as observed in clinical HE constrain synaptic plasticity? Numerous studies have investigated the modulation of signal transduction pathways activated in LTP or LTD irrespective of being necessary or sufficient for changes in synaptic efficacy (Wen et al., 2013). Yet, the final common pathway of plastic changes in glutamatergic neurotransmission, i.e. the regulation of the number and gating properties of AMPARs, which eventually determine synaptic strength, has rather been neglected. Here, we have employed a co-culture system of neurons and astroglial cells, which preserves part of their interplay during synaptic development and maturation but also allows for cell type-specific molecular and functional analysis, to investigate glutamatergic neurotransmission in conditions that model hyperammonemia in HE. We conclude from our data that chronically high concentrations of ammonia limit synaptic plasticity by compromising the number of extrasynaptic AMPARs, which are required as reserve for enhancing synaptic efficiency.
Materials and methods
SDS-PAGE and immunoblotting Crude membrane fractions were prepared from primary hippocampal neurons stratified into the respective experimental groups (DIV16–18). Cells were lysed by sonication in 20mM Tris–HCl pH7.5, 1mM iodacetamide, 1mM EDTA, 150mM NaCl supplemented with fresh proteinase inhibitors (aprotinin, leupeptin and pepstatin A, at 100μg/ml each) and centrifuged at 1000g for 5min at 4°C. The supernatant was ultra-centrifuged at 125,000g for 30min at 4°C. Pellets containing the crude membrane fraction were re-suspended in 1× Laemmli buffer and incubated for 10min at 37°C. Protein samples were resolved by 12% SDS-PAGE, electro-blotted on PVDF membrane (Millipore), and detected by immunoblot analysis. The following cck 1303 were used: rabbit anti-GluA1 (1:1000, AB1504, Millipore), mouse anti-GluA2 (1:1000, MAB397, Millipore), rabbit anti-GluA2/3 (1:1000, 07–598, Millipore), rabbit anti-GluA4 (1:1000, AB1508, Millipore), rabbit anti-stargazin (1:1000, 07–577, Millipore) and custom-made guinea pig anti-CNIH2 (1:1000, peptide epitope: DELRTDFKNPIDQGNPARARERLKNIERIC), goat anti-rabbit, anti-mouse, or anti-guinea pig secondary antibodies conjugated to horseradish peroxidase (1:15,000, Santa Cruz). Blots were developed with ECL plus reagent (GE Healthcare). Densitometric analysis was performed using ImageJ software. Data are given as mean expression level±standard error of the mean (SEM) in arbitrary units normalized to untreated control. For assessing statistically significant differences between experimental groups, the non-parametric Kruskal Wallis and Dunn's posthoc analysis were performed.
Results
Discussion In the brain, ammonia is predominantly detoxified in astrocytes by glutamine synthesis. In good agreement, astrocytes have been reported to protect neurons against ammonia toxicity (Rao et al., 2005). With increasing ammonia concentrations, glutamine will accumulate resulting in cell swelling and eventually in cytotoxic edema (Häussinger and Görg, 2010). However, it is still under debate whether glial edema is the central cause for HE symptoms, particularly in chronic HE; in contrast, there is evidence that direct alteration of neurotransmission is critically involved in the pathogenesis of cognitive and motor impairments (Felipo, 2013). For our study, we have therefore chosen a co-culture model of hippocampal neurons and astroglial cells in order to maintain neuro-glial interplay to a certain extent (Kaech and Banker, 2006). Cells were grown in sandwich orientation on different support plates holding the advantage over organotypic slice culture or in vivo models that both electrophysiological and comprehensive molecular analyses can be performed in a cell-type specific manner. Our results show that neurons stressed with relevant concentrations of ammonia in co-culture with astroglia showed a dose-dependent decrease in the expression of AMPAR mRNAs and proteins. Both the main pore-lining subunits, GluA1 and GluA2, and the most abundant auxiliary subunits of the TARP and CNIH protein families in hippocampus were affected (Monyer et al., 1991, Schwenk et al., 2012). With respect to molecular mechanisms being involved in AMPAR downregulation, we can only speculate at this point. Ammonia is a rather broadly acting pathogen. It is known that ammonia increases oxidative and nitrosative stress (Görg et al., 2013). Reactive oxygen species may oxidize RNA as has been demonstrated in astrocytes and even more pronounced in neurons in an animal model of HE (Görg et al., 2008). Oxidation of RNA could result in accelerated degradation and impair translation into protein. Also, direct effects of ammonia on gene transcription should not be excluded (Norenberg et al., 2009), which might explain the decrease in AMPAR expression.