PURPOSE

The central nervous system(CNS) has a very active-oxidative-metabolism compared with other organs^1,2.Physiological levels of reactive-oxygen-species(ROS) act as cellular secondary messengers. When in excess, become neurotoxic and are involved in neurodegeneration. The presence of CNS and systemic-OS contribute to the pathogenesis of CHE³. Moreover,peripheral-inflammation and systemic-OS lead to acute worsening of CHE patients⁴.Impaired ammonium clearance by the diseased liver leads to brain glutamine(Gln) accumulation in CHE, inducing in-vitro-ROS generation associated with astrocyte deterioration^5,6.In addition,hyperammonemia impairs neutrophils function by decreasing phagocytosis and increasing spontaneous-oxidative-burst⁷. In a rat-model of CHE(bile-duct-ligated,BDL-rat) we have demonstrated by ¹H-MRS the indirect presence of OS(i.e. decrease of brain ascorbate(Asc) and glutathione(GSH) in hippocampus, together with the early increase in plasma ammonium and brain Gln,suggestive of oxidative/osmotic stress^8,9.Furthermore we have validated the OS findings using for the first time ex-vivo-ESR at a late-time-point in disease evolution¹⁰. The novelty of the present study relies on the longitudinal assessment of hippocampal and systemic-OS using a multimodal approach:in-vivo ¹H-MRS followed by ex-vivo/in-vitro-ESR detection of ROS and histological measures.

METHODS

In-vivo ¹H-MRS
A new group of Wistar male-adult-rats was BDL or sham-operated and measured before BDL(n=23) and after every 2-weeks up to 8-weeks(n=7).In-vivo ¹H-MRS (9.4T,Varian/Magnex-Scientific) was used to analyze neurometabolism including antioxidant concentrations in the hippocampus(2x2.8x2mm3) using SPECIAL-sequence¹¹(TE=2.8ms,TR=4sec;160averages).Metabolite concentrations were calculated by LCModel using water as reference.
Ex-vivo/in-vitro-ESR
ESP300E(Bruker-BioSpin,TE102-cavity) was used to measure the concentration and time-course of ROS in hippocampus and blood.The hippocampus was extracted/weighed/sliced and transferred into RPMI1640-medium(37^oC) with 10mM-CMH-cell-permeable non-toxic spin-trap(Noxygen-GmbH). Blood(1mL) was collected and CMH-spin-trap was added(10mM final-concentration).After each incubation-time, tissue-suspension/blood was transferred into quartz-capillary and measured.
Neurohistology-Golgi-Cox-method
Golgi-Cox-staining uses the principle of metallic-impregnation of neurons and it was applied to reveal detailed neuronal morphology of the hippocampus.110μm-thick brain sections were cut sagittally(25 slides/hemisphere,n=8).
Lymphocytes/Polymorphonuclear cells (PMNs) count
Sterile-blood was obtained from heart(n=20). 7ml of anticoagulated-blood was used for each isolation procedure. Acridine-Orange-assay was applied to assess the lymphocytes and PMNs concentrations.

RESULTS AND DISCUSSION

The presence of chronic liver disease was confirmed by increased plasma ammonium and bilirubin(2-weeks-post-BDL).
¹H-MRS revealed the well-known ammonium-induced increase in Gln(+67%,4-weeks-post-BDL) together with the decrease in brain osmolytes(Fig1A-B). Astrocytes are essential to detoxify ammonium¹²,while Gln increase is known to produce OS response in cells¹³.In this context,we have previously shown a global atrophy of astrocytes in BDL-rats^8,9,suggesting an astrocytic activation and synaptic depression induction¹⁴.
The main antioxidants measured by ¹H-MRS showed a decrease at 4-weeks-post-BDL:Asc -11%, GSH -10%, reaching -17% at 8-weeks(Fig1C). Of note, the previously observed Asc decrease at 8-weeks-post-BDL was absent, probably due to the smaller number of rats.The hippocampus ex-vivo-ESR revealed a significant increase of OS +68%(p<0.01) already at 2-weeks-post-BDL,thereby enhancing the time-resolution and understanding of neurometabolic changes in early CHE relative to ¹H-MRS.
ROS as signaling and stress-molecules affect synaptic-plasticity and are critical for long-term-potentiation(LTP) in the hippocampus,,a form of synaptic-plasticity for learning/memory but are also involved in aging-related-impairment the long-term-depression(LTD)^15,16.Dismutation of superoxide-anion generates H₂O₂ in hippocampal neurons, affecting LTP in complex ways.The excess of ROS attenuates LTP and synaptic neurotransmission^17,18.
In-vitro-ESR on peripheral-blood demonstrated a significant increase of systemic OS(+49%,6-weeks-post-BDL).These results are consistent with the increase in the peripheral white-blood-cells count: lymphocytes:+177%,PMNs:+78% at 2-weeks-post-BDL. Ammonia induced PMNs swelling⁷ was detected already at 2-weeks-post-BDL.This is a sign of impaired phagocytosis and increased prevalence of spontaneous-oxidative-burst⁷(Fig.2D).This finding may be in keeping with the known vulnerability to infection in cirrhosis and the consequent risk for HE decompensation.
The Golgi-Cox-staining showed a significant increase in CA1(~67%, p<0.001) and DG(~54%,p<0.001) neuronal-soma surface and a significant loss of dendritic-spines-density in CA1~49%(apical,basal,p<0.001) and DG~43%(apical,p<0.001) in hippocampal neurons(Fig.3).It’s known that different neurons have different levels of vulnerability to OS.The hippocampus,amygdala,and cerebellar granule cells have been reported as the most susceptible and therefore may the first to exhibit functional decline¹⁹.In addition NAA(showing -9% decrease by 1H-MRS) may be a minor contributor to neuronal volume-regulation which responds to hypo-osmolarity and acts as a co-transit substrate for a putative molecular-water-pump to remove excess of water from neurons²⁰. Furthermore, the hippocampus plays a critical role in several higher brain functions, such as learning/memory, spatial-encoding and is one of the few brain regions which exhibits adult-neurogenesis²¹.The alterations of hippocampus neuronal morphology due to metabolic-degenerative-disorders are often associated with cognitive impairment, suggesting that the hippocampus may be an interesting area to study in HE.

CONCLUSIONS

The present study shows the potential of multidisciplinary-approach to monitor longitudinally central and systemic-OS in a rat-model of CHE using in-vivo ¹H-MRS combined with ex-vivo/in-vitro-ESR spectroscopy. The presence of hippocampal and systemic-OS was measured early,i.e.2-weeks post-BDL. Moreover, the ¹H-MRS findings were validated by ex-vivo-ESR. Our previously depicted astroglia atrophy^8,9 may cause reduction in network connectivity, synaptic coverage and homeostatic capabilities²⁴.These results corroborate with the present Golgi-Cox-staining showing significant increase of hippocampal neuronal soma surface and dendritic-spines-density loss.Therefore, it’s tempting to extrapolate that high OS levels are associated with a loss of neurons connectivity associated with behavioral deterioration in a rat-model of CHE.

Acknowledgements

Supported by CIBM of the UNIL, UNIGE, HUG, CHUV, EPFL, the Leenaards and Jeantet Foundations and the SNSF project no 310030_173222/1

References

¹Halliwell B. J Neurochem 59: 1609–1623,1992;² Samina Salim, J Pharmacol Exp Ther 360:201–205, 2017; ³Bosoi CR, Rose CF. Metab Brain Dis. 2013 Jun;28(2):175-8;⁴Cristina R. Bosoi, Christopher F. Rose, Metabolic Brain Disease,2012; ⁵Skowronska M. and Albrecht J., Neurochemistry International, 2012;⁶Lachman V. et al., Archives of Biochemistry and Biophysics, 2013; ⁷Debbie L. Shawcross et al., HEPATOLOGY, Vol. 48, No. 4, 2008; ⁸K. Pierzchala et al., ISMRM2019 # 2279; ⁹Olivier Braissant, et al.,Journal of Hepatology 2019; ¹⁰K. Pierzchala et al., ISMRM2018 # 448; ¹¹Mlynárik et al, Magn Reson Med 2006; ¹²Gavin A. K. et al., Liver Int. 2014: 34: 1184–1191; ¹³Jan Albrecht and Michael D. Norenberg, HEPATOLOGY 2006;44:788-794; ¹⁴Rogier Min & Thomas Nevian, 2012 nature NEUROSCIENENCE; ¹⁵Lauren T. Knapp and Eric Klann, Journal of Neuroscience Research, 2002;¹⁶Cynthia A.et al., PNAS 2009 August, 106 (32) 13576-13581; ¹⁷Mol Neurobiol(2017); ¹⁸Xinkun Wang and Elias K. Michaelis,Frontiers in Aging Neuroscience, 2010; ¹⁹Wang X and Michaelis EK, Front Aging Neurosci 2:12(2010); ²⁰Brian Ross et al.,Progress in Neurobiology 81 (2007); ²¹Peter Jonas and John Lisman, Frontiers in Neural Circuits, September 2014; ²²Bruna Bellaver et al.,Mol Neurobiol,2017;²³Cecilia Hidalgo and Alejandra Arias-Cavieres PHYSIOLOGY • Volume 31 • May 2016; ²⁴Alexei Verkhratsky, José J. Rodríguez, and Luca Steardo, The Neuroscientist 2014.